Amplification
can be defined as: The process of increasing the magnitude of a variable
quantity--especially the magnitude of a voltage and/or current, without
substantially altering any other quality.
The (pure) sine wave is the only periodic waveform that contains no harmonic
energy. All of the energy in a sine wave is contained in the fundamental
frequency. In other words, a pure sine wave is coherent. Maintaining the quality
of sine waves during amplification is a major concern in the design and
operation of amplifiers.
Even though the following discussion pertains to the operation of gridded
electron tubes, some of the concepts apply to power FETs because both are
voltage-driven amplifying devices.
In the discourse that follows: if a voltage is said to be negative-going, that
does not necessarily mean that the voltage is negative. It could be a positive
voltage that is moving in the negative direction. If a voltage is
positive-going, or moving in the positive direction, it could be a negative
voltage or a positive voltage that is moving in the positive direction.
In an electron tube, the cathode emits a cloud of electrons. Since electrons
carry a negative charge, and unlike charges attract, electrons are strongly
attracted by the positive voltage that is applied to the anode. Unless something
is placed in the way, much current flows between the cathode and the anode.
The term "grid" describes appearance--not function. A grid is made from a number
of closely spaced wires or bars--like a bird cage. The grid is placed close to
the cathode. As a result, the grid has more influence over the cathode's
electrons than does the more distant anode. Thus, a small change in grid-voltage
produces a large change in the flow of electrons. Because the voltage applied to
the grid controls the flow of electrons from the cathode to the anode, in
function, the grid acts like a valve or gate. The grid requires virtually zero
current to perform its job. As a result, the power gain of a gridded electron
tube is theoretically high.
Since like charges repel, a sufficiently negative grid can stop electrons from
traveling to the anode. As the grid is made less-negative, the flow of electrons
to the anode steadily increases. In other words, a positive-going, albeit
negative polarity, grid-voltage causes an increasing flow of electrons between
the cathode and the anode. The reverse is also true: a negative-going
grid-voltage causes a decreasing flow of electrons from the cathode to the
anode. As long as the grid remains negative with respect to the cathode, the
relationship between grid-to-cathode voltage and anode-current is fairly linear.
When an appropriate load resistor is connected between the anode and its
positive voltage source, the changes in anode-current produced by changes in
grid-voltage create a proportional, typically much larger, voltage change across
the resistor. The ratio of changes in voltage across the load resistor to
changes in grid-voltage is the voltage amplification factor. It is designated by
the Greek letter Mu. Since Mu is higher at high anode-voltages than it is at low
anode-voltages, average Mu is a more meaningful number than maximum Mu. Average
Mu ratings vary from about 2 to 240. Mu is partly determined by the spacing
between the grid wires and the distance between the grid and the cathode..
The duration of
anode-current conduction per cycle determines the class of amplifier operation.
A conduction angle of 360 degrees means that the anode is conducting current
during 100% of the input sine wave cycle. A conduction angle of 90 degrees means
that the anode is conducting current during 25% of the input sine wave cycle.
Long conduction angles produce a more linear representation of the input sine
wave. Short conduction angles produce more efficiency--and less linearity.
Class A is defined as a conduction angle of 360 degrees. Class B is defined as a
conduction angle of 180 degrees. Class C is defined as a conduction angle of
less than 180 degrees. The subscript 1 indicates that no grid-current flows. The
subscript 2 indicates that grid-current flows--the result of driving the grid
into the positive voltage region.
When the conduction angle is less than 360 degrees, the missing part of the sine
wave must somehow be filled in. One way of filling in the missing part of the
sine wave is by utilizing the flywheel effect of an output tank circuit. Another
way to produce a smooth sine wave is to use a push-pull configuration. If each
device in a push-pull circuit conducts for at least 180 degrees, a smooth sine
wave can be produced.
Class A is the
most linear class of amplifier operation. Class A amplifiers produce only about
1/100,000 part, or minus 50dB, distortion. The theoretical efficiency of a Class
A amplifier is 50%. The practical efficiency is slightly lower. Class A is used
mainly in low level amplifiers--where efficiency is not much of an issue. Since
Class A operates with continuous (360 degrees) conduction, no tank circuit is
needed to complete the sine wave. Class A is ideal for wide band amplification.
The zero-signal anode-current [ZSAC] in Class A is set to roughly half of the
electron tube's maximum anode-current rating. Although the meter-indicated
anode-current remains constant from zero signal to maximum signal, the
instantaneous anode-current typically varies from just above zero to many times
the meter-indicated anode-current.
The maximum available power in Class A is roughly equal to the anode-dissipation
rating of the electron tube.
The Class A amplifier can be compared with a gas turbine engine. Both have a
smooth, continuous power stroke--and neither one is very efficient.
Class AB1
amplifiers are roughly 60% efficient. The trade-off for increased efficiency is
slightly more distortion--roughly 1/10,000 part, or minus 40db. Since most
transceivers produce more than minus 36db of IMD, such an amplifier would not
add significant distortion.
The anode-current in Class AB1 varies in proportion to the grid-voltage for
about 60% of the input sine wave cycle. Thus, the anode-current is off for about
40% of each input sine wave cycle. The missing 40% in the output sine wave is
filled in by the flywheel effect of the output tank circuit.
In Class A or Class AB1 a somewhat unusual relationship exists between the work
being performed by the tube and the grid voltage. The grid is operated in the
zero to negative voltage region. Maximum instantaneous anode-current, maximum
anode voltage swing and maximum peak power output coincide with an instantaneous
grid-voltage of zero--i.e., maximum stoke equals zero grid volts.
The grid must not be allowed to become positive. If the grid became positive,
electrons from the cathode would begin flowing into the grid. Whenever
grid-current flows, the linear relationship between grid-voltage and
anode-current deteriorates.
Since there is zero grid-current in Class A or Class AB1 operation--and any
voltage multiplied by zero amperes is zero watts--the driving power is usually
stated as zero on the tube manufacturer's technical information sheet. However,
because charging and discharging a capacitor requires current flow, in the real
world of conductor RF resistance, charging/discharging the grid capacitance at
an RF rate consumes some power. In a typical HF amplifier, the drive power
required for Class AB1 MF/HF operation is roughly 1% to 2% of the output power.
Thus, the typical power gain is roughly 50 to 100. As frequency increases,
conductor resistance increases due to skin-effect. More R causes more I^2 R
loss. As frequency increases, the amount of current needed to charge and
discharge the grid capacitance also increases--causing even more I^2 R loss.
These losses can only be compensated for by adding more drive power.
Drivers (usually a transceiver) require a resistive load--which the capacitive
grid does not provide--so a suitable resistance must be connected from the grid
to RF-ground..
The zero-signal anode-current [ZSAC] in a Class AB1 amplifier is normally set to
about 20% of the maximum-signal single-tone anode-current.
Tubes that are designed for Class A and Class AB1 service produce high peak
anode-current when the instantaneous grid-voltage is zero. Typically, the peak
anode-current is about three times the maximum rated (average) anode-current.
Most of the tubes that are used in Class A and Class AB1 RF amplifier service
are tetrodes and pentodes--devices that have the advantage of grid-to-screen
amplification. Triodes are seldom used because the only grid designs that can
produce high anode-current with zero grid volts are those that have a Mu of 2 to
5. Low Mu triodes require much more driving voltage than a comparable tube with
a screen requires. Since imposing a high RF voltage across the capacitive grid
is difficult, low Mu triode Class A and Class AB1 power amplifiers are only
practical up to a few hundred kHz.
The most common configuration for Class AB1 operation is grid-driven. Since grid amplification as well as grid-to-screen amplification takes place, the resulting power gain is high. Class AB1 amplifiers can also be cathode-driven if the tube is a tetrode or pentode. The grid is tied to the cathode. Thus, the grid-voltage is always 0V--so no grid-current flows. The screen is grounded. The input signal is applied to the cathode/grid. Because input signal voltage is applied between the grid and the screen, grid-to-screen amplification takes place. However, since the grid is tied to the cathode, no grid amplification takes place. Although the power gain is relatively low, linearity is excellent. The Collins 30S-1 is an example of a cathode-driven Class AB1 amplifier.
The Class AB1 amplifier is like a 2-cycle, single-cylinder engine. The power stroke is roughly half of each crankshaft revolution--it has a flywheel (the tank circuit) that supplies power between power strokes--and it is more efficient than a gas turbine (Class A).
Class AB2 is similar to Class AB1 except that the grid is driven into the positive voltage region during a part of the anode conduction period. A Class AB2 amplifier can be grid-driven or cathode-driven.
When the grid is
driven positive, it attracts and accelerates the cathode's electrons. Some of
the electrons stick to the grid, resulting in grid-current. Electrons that miss
the grid travel to the anode. The accelerated head start causes a sharp increase
in instantaneous anode-current--and a sharp decrease in linearity. The
distortion products from a single-ended, Class AB2 grid-driven amplifier are
roughly 1/100 part [minus 20db]. In SSB service, this level of distortion is
virtually certain to cause interference to other stations using adjacent
frequencies. However, by limiting grid-current and by adding an unbypassed,
low-L cathode feedback resistor (to develop an out-of-phase {negative} feedback
voltage) it is possible to achieve acceptable linearity in grid-driven Class AB2
operation--but only if the grid current produced is small.
An unbypassed cathode resistor is also useful for improving linearity in Class
AB1. For example, the 4CX250B has a somewhat objectionable distortion level in
Class AB1, SSB service. Adding a 25 Ohm resistor between the unbypassed cathode
and chassis ground improves linearity. The trade-off is that slightly more grid
drive voltage is needed to achieve the same output level. Cathode
negative-feedback is also useful with TV sweep tubes--devices that were
originally designed for switching--the opposite of linear amplification. When an
appropriate cathode resistor is used with a sweep tube, reasonable linearity can
be achieved.
Even though
Class AB2 cathode-driven/grounded-grid operation produces grid-current, it is
never the less fairly linear due to the laundering effect of negative-feedback.
This is the result of the input and output signals being in series with each
other and out of phase. Due to the negative-feedback, the distortion level in
Class AB2 grounded-grid service is low--typically about 40db below PEP.
High-Mu triodes work well in Class AB2 grounded-grid operation. Medium-Mu
triodes can be used, but they have less power gain. Tetrodes and pentodes
usually work well in grounded-grid operation. Since tetrodes and pentodes
typically have a grid-to-screen amplification factor of about 5, its easy to
assume that they offer an advantage over triodes in Class AB2 grounded-grid
operation. However, RF-grounding the grid and the screen stops grid-to-screen
amplification. Applying DC screen-voltage does NOT increase gain because
grid-to-screen amplification can not take place unless input signal voltage is
applied between the grid and the screen.
The maximum available power in Class AB2 is roughly double the anode-dissipation
rating.
Class B is defined as a conduction angle of 180 degrees. Class B RF amplifiers produce unacceptable distortion in SSB operation.
Class C is
defined as an anode conduction angle of less than 180 degrees. In Class C, the
amplifying device is deliberately not operated linearly. Instead, it is operated
as a switch in order to reduce resistance loss. The anode conduction angle in
Class C operation is usually made as short as is possible. In effect, the tank
circuit makes the RF output sine wave--like a bell that is struck at a constant
rate by a hammer. This is similar to the principal behind the spark transmitter.
The efficiency of a typical Class C amplifier is high. When compared to a Class
AB1 or Class AB2 amplifier operating at the same power input, a Class C
amplifier will deliver a received signal increase of about 1db--in other words,
1/6 of 1 S-unit. However, significant trade-offs are required to achieve that
1/6 of 1 S-unit. As is the case with Class B operation, the distortion from
Class C operation is so high that SSB operation is precluded. Only CW, FM or FSK
operation is practical. The harmonic output level from a Class C amplifier is
substantial. Extra filtering is usually needed to control harmonic radiation.
The maximum available power in Class C operation is roughly three to four times
the anode-dissipation rating of the electron tube.
Class D is used up to about 1.6MHz--mostly in AM broadcast service. In Class D operation, the amplifying device rapidly switches on and off at a fixed rate--like a switching power supply--except that the output voltage varies at an RF rate. The amplitude of the RF is controlled by varying the on period of the switch. Smoothing is accomplished by a complicated filter that converts some of the odd-harmonic energy from the rectangular waves back to fundamental frequency energy. Class D is highly efficient--but it is limited in frequency capability and frequency agility.
Arriving at an
amplifier design that will give years of surprise-free service involves many
considerations. Merely copying circuits from published amplifier designs or
commercial amplifiers is not necessarily the best approach. Doing so may result
in copying someone's mistakes. The best approach is to learn what you can about
each section of an amplifier, discuss it with others--then reach your own
conclusions.
Basic prerequisites for getting a handle on what's going on inside an amplifier
are an understanding of Ohm's Law, inductive and capacitive reactance,
impedance, resonance, how gridded electron tubes function, and some knowledge of
L and pi networks.
A useful book on amplifier design is Eimac®'s Care and Feeding of Power
Grid Tubes. I will minimize the discussion of topics that are covered
adequately in this book.
When the first
RF power FETs were introduced, it was commonly thought that FETs would
eventually replace bipolar transistors and gridded electron tubes in HF power
amplifiers. Since RF power FETs work better at 50V than they do at 12V, FETs
have not replaced bipolar transistors in 12V mobile applications. Another
difficulty with FETs is cost. A pair of FETs that can produce 1200W PEP at 29MHz
cost about six times more than an electron tube, or tubes, that can do the same
job. The FETs' input power requirements are 50V at 50A, i.e., 2500W--so there's
considerable heat to dispose of. Meeting the cooling requirement is not nearly
as easy as it is with tubes because tubes operate quite happily at surface
temperatures that destroy silicon devices.
In low power applications at room-temperature, solid-state devices can last 100
years. However, at the junction temperatures encountered in high power
applications, the P and N doped layers slowly diffuse into each other--thereby
steadily eroding the device's amplifying ability. A relatively-large rigorous
cooling system is needed to achieve a reasonable operating life from high power
solid-state RF amplifying devices.
Another difficulty with solid-state high power RF amplifiers is their power
supply requirements. Tubes are quite tolerant of moderate variations in their
anode supply voltage. Transistors, however, are fatally sensitive to
over-voltage. It is much easier to build a 3000V, 0.8A unregulated supply for a
tube than to build a regulated 50V, 50A power supply with over-voltage,
over-current and over-temperature protection circuitry for high-power
solid-state devices.
The bottom-line is that 1500W, HF, gridded electron tube amplifiers are more
efficient, more forgiving, easier to cool, more compact, weigh less, are more
tolerant of high SWR and are less costly than 1500W HF semiconductor amplifiers.
For instance, a pair of legal-limit FETs cost about $800 from Motorola®. The
efficiency is about 10% less that what one can achieve with gridded electron
tubes.
For at least the
last three decades, the vast majority of amateur radio amplifier designs have
been Class AB2 cathode-driven--a.k.a. 'grounded-grid'. One reason for this is
simplicity--or at least the appearance of simplicity. Ground the grid(s), drive
the cathode. Only three supplies are needed--the T-R switching supply, the
filament supply and the anode supply. Neutralization is theoretically not needed
because the grounded grid(s) shield the output element, the anode, from the
input element, the cathode. This theory works almost perfectly.
Grounded-grid amplifiers are virtually always stable at the operating frequency
because the reactance of the feedback C is too high at HF to allow regeneration.
This is fortunate because there is no way of neutralizing a single ended
grounded-grid amplifier. Another advantage is flexibility. Almost any tetrode,
pentode, or high-Mu triode from the junk box will work. Linearity is usually
good and the typical power gain--10db to 14db--is acceptable. So far, so good.
Now for the trade-offs.
What goes on inside a grounded-grid amplifier is not as simple as it looks. The
AC component of the anode-current and the grid-current, i.e., the RF cathode
current, passes entirely through the cathode coupling capacitor and the
tuned-input circuit--so the input circuit is in series with (and out of phase
with) the output circuit. The components in the tuned circuit must be able to
handle a substantial amount of RF current. Manufacturers of tubes that are
designed for grounded-grid operation typically recommend using a tuned input
pi-network with a Q of 2 to 5. To maintain an acceptable SWR and Q when the
operating frequency changes appreciably, all three reactances in the tuned input
must change proportionally. However, if Q is allowed to change, L can be left as
is providing that C1 and C2 are retuned. [For more information on this problem,
see the section titled "Tuned Input Circuits."]
Even though HF grounded-grid amplifiers are stable at their operating frequency,
at VHF the grid looses its ability to shield the input from the output. HF
grounded-grid amplifiers have a less-than-pristine reputation for VHF stability.
For wide frequency coverage, the Class AB1 grid-driven amplifier requires a much
simpler tuned input than a grounded-grid amplifier requires. Typically,
grid-driven amplifiers have more power gain than grounded-grid amplifiers. One
Class AB1 grid-driven amplifier has about as much gain as two Class AB2
grounded-grid amplifiers in series. The trade-off is two additional DC
supplies--a grid bias supply and a screen supply. Both of these supplies need to
be adjustable. HV power FETs make this task easy.
There are two
types of cathodes--directly-heated and indirectly-heated. In a directly-heated
cathode, a ditungsten carbide layer on the hot (c.1800 degrees K) tungsten,
alloyed with about 1.5% thorium--a.k.a. 'thoriated-tungsten', filament wire
emits electrons. In an indirectly-heated cathode, the filament (a.k.a. heater)
heats a metal cylinder that is coated with strontium oxide and barium oxide.
This coating is relatively frangible--but highly emissive.
Ditungsten carbide is commonly formed by heating tungsten in an atmosphere of
acetylene (C2H2) gas. Carbon atoms in the gas break their
electron bonds with hydrogen atoms and bond with tungsten atoms to form
ditungsten carbide on the surface of the filament wire. Since it is atomically
linked to the underlying tungsten, the ditungsten carbide layer is very durable.
During use, the process reverses. Ditungsten carbide gradually looses carbon and
changes back to tungsten. Extra heat exponentially accelerates this process. A
cathode is worn out when the carbon is mostly used up.
After their cathodes grow tired of emitting electrons, large external-anode
amplifier tubes are commonly "recarburized" with acetylene, vacuum-pumped and
resealed. This restores full emission. Although it is possible to recarburize a
3-500Z, doing so is not economically feasible. The smallest tube that is
currently being recarburized is the 3CX1000A7.
Each type of cathode has advantages and disadvantages. Indirectly-heated
cylinder [8877] and planar [3CX100A5] cathodes have much less inductance than a
directly-heated cathode made from wires. Thus, indirectly-heated cathodes are
more frequency-capable. Some indirectly-heated cathode tubes can perform
satisfactorily at 2500MHz. The 3CX100A5 is an example.
Directly-heated/thoriated-tungsten cathodes are more resistant to damage from
electrons that bounce off the anode. It's possible to use up to 22kV with the
larger thoriated-tungsten cathode tubes. Electrons that have been accelerated by
such voltages move at very high velocities. When they strike the anode, they
produce X-rays.
A thoriated-tungsten cathode typically warms up in one second, while few
indirectly-heated cathodes can warm up safely in one minute--and three to five
minutes is not uncommon. For HF operation, indirectly-heated cathode tubes have
a much higher cost to watt ratio than thoriated-tungsten cathode tubes. For VHF
and especially for UHF operation, indirectly-heated cathode tubes are often the
only choice. For super-power HF operation, thoriated-tungsten cathode tubes are
the only choice.
Cathodes deserve respect. Filament-voltage and filament inrush current are the
prime areas for concern.
For optimum life
from a thoriated-tungsten cathode, the filament-voltage should be just above the
voltage where PEP output begins to decrease. As a thoriated-tungsten cathode
ages, filament-voltage needs to be increased incrementally to restore full PEP.
By using this technique, commercial broadcasters typically achieve an operating
life of more than 20,000 hours from thoriated-tungsten cathode tubes.
According to Eimac®'s Care and Feeding of Power Grid Tubes, "every 3%
rise in thoriated-tungsten cathode filament-voltage results in a 50% decrease in
life due to carbon loss." Each additional 3% rise in filament-voltage decreases
the life by half. Thus, cathode life is proportional to [E1/E2]^23.4 where E1 is
the lowest filament-voltage at which normal PEP output is realized--and E2 is
the increased filament-voltage. However, for heater-type oxide cathodes, if the
heater potential is allowed to fall below the specified level, the emissive
material may flake off of the cathode, and cause a cathode-grid short. On the
other hand, excessive heater potential causes barium migration to the grid -
which results in primary grid emission.
It's simple to
make the filament voltage adjustable when the filament is powered by its own
transformer. All that's needed is a small rheostat in series with the primary.
For dual voltage, dual primary transformers, a dual ganged rheostat is required.
However, when the filament is powered by a winding on the HV transformer, making
the filament-voltage adjustable is more difficult since a dual ganged, very low
resistance, high current rheostat must be connected to the low-voltage
high-current secondary winding. The typical value needed for a pair of 3-500Zs
would be (2) adjustable 0.01 Ohm @ 30A--most definitely not a common rheostat. A
reasonable substitute can be made from a double pole, c.10 position, 30A rotary
switch and short lengths of resistance wire or ribbon bridging the fixed
contacts.
An indirectly-heated cathode can be ruined by operating it below the rated
minimum filament-voltage. When operated above its maximum filament-voltage
rating, an indirectly-heated cathode boils off emissive material (principally
barium) onto the grid and other parts. This results in decreased cathode life
and undesirable grid-emission when the grid warms up during transmit. This
condition is indicated when the output power steadily drops off in AØ (max.
signal, key-down, a.k.a. NØN) operation. The decrease in power normally begins
within two seconds.
For maximum cathode life in HF communications service, an indirectly-heated
cathode should be operated at the rated minimum filament-voltage. This can be
accomplished best with a regulated DC supply.
In a typical amateur radio amplifier, the filament-voltage rises about 5% during receive due to decreased load on the electric-mains. Not only is a 5% increase in filament-voltage during receive useless, it uses up thoriated-tungsten cathode emission life about three times as fast as during transmit. However, if an appropriate resistor is placed in series with the filament transformer primary, and shorted out by a relay during transmit, the filament-voltage will not change appreciably between transmit and receive. An appropriate type of relay for shorting the resistor on transmit is a 'power' reed-relay.
Thoriated-tungsten filaments
commonly consist of two vertical intermeshing helices (coils) of tungsten wire
that are suspended by their ends. (see Sept. 1990 QST, p.15) The conductance of
tungsten at room temperature is about 8.33 times the conductance at the normal
operating temperature. Thus, the start-up current for a 15-ampere filament can
exceed 100 amperes. Needless to say, 100 amperes makes for a dandy
electromagnet.
In a high amplification triode such as the 3-500Z, the filament helices clear
the grid cage by a matter of thousandths of an inch. If the position of the
filament changes, a grid-to-filament short may result. Therefore, it is prudent
to limit filament inrush current in order to minimize thermal and magnetic
stress.
Since the grid-to-cathode clearance in an indirectly-heated cathode is not
affected by movement of the heater inside the rigid cathode cylinder,
indirectly-heated cathodes are not affected by inrush current.
For many of its smaller thoriated-tungsten cathode amplifier tubes--such as the
3-400Z and 3-500Z--Eimac® recommends that filament inrush current be limited to
no more than double the normal current. This rating is easily exceeded unless a
special current-limiting filament transformer or a step-start circuit is used.
Transmitting AØ
for long periods on 10m with the load capacitor set for maximum C would result
in very high grid-current and almost no RF output. Under such a condition it
might be possible to overheat a grid. However, since most people tune an
amplifier for maximum output--and maximum output virtually coincides with normal
grid-current--very few people are likely to overheat a grid. Thus, complex
electronic grid-protection circuits are seemingly unnecessary.
A disadvantage of electronic grid-protection circuits is that they are not
effective against the most common source of grid damage--intermittent VHF
parasitic oscillation.
During a major
problem, the anode (plate) current meter and other amplifier components can be
subjected to a large current surge as the HV filter capacitors discharge. The
peak discharge current can exceed 1000a if a series resistor is not used to
limit the short circuit current that can be delivered by the HV filter
capacitors. The current limiting resistor is placed in series with the positive
output lead from the filter capacitors. A wire wound resistor with a high length
to diameter ratio works best. A 10 ohm, 10W wire wound resistor is adequate for
up to about 3kV & 1A. For higher voltages, additional 10 ohm, 10W resistors can
be added in series to share the voltage drop during a glitch. Wire wound
resistors with a high length-to-diameter ratio are best for this type of
service. Since about 1985, Eimac® has recommended the use of a glitch protection
resistor in the anode supply circuit. Svetlana® typically recommends using a 10
to 25 ohm glitch resistor.
A HV current limiting/glitch resistor may disintegrate during a major glitch--so
it should be given a wide berth with plenty of chassis clearance. If the chassis
clearance is minimal, its a good idea to cover the chassis with electrical
insulating tape. Glass-coated (a.k.a. vitreous) wire wound resistors are the
most suitable type of resistor for this application. If a glass-coated resistor
comes apart during a major glitch, it won't be throwing chunks of shrapnel
around--like a less-expensive rectangular ceramic-cased resistor often does.
Metal-case power resistors should not be used in this application. If a
glass-coated glitch resistor is damaged during a glitch, it should be replaced
with two such resistors in series to reduce the peak V-gradient per unit of
length during a problem.
If the positive HV arcs to chassis ground--due to lint, a hapless insect, a VHF
parasitic oscillation, or moisture--the negative HV circuit will try to spike to
several kilovolts negative in the typical 1500W amplifier. In the real world,
this type of glitch is not an uncommon occurrence. Anything that gets in the way
of the negative spike may be damaged. Since the grid-current meter is normally
connected between chassis ground and the negative HV circuit, the meter can be
exposed to kilovolts at hundreds of amperes.
The easiest way to protect a current meter is to connect a silicon rectifier
diode across it, or across its shunt resistor. Usually, only one diode {cathode
band to meter negative} is needed in parallel with a DC meter. In some circuits,
it is best to use two diodes in parallel [anode to cathode] with the meter
movement to protect against positive and negative surges.
It may take more than one diode to protect a meter shunt resistor. A silicon
diode begins to conduct at a forward voltage of about 0.5V. To avoid affecting
meter accuracy, the operating voltage per glitch protection diode should not
exceed 0.5V. For example, a 1 ohm shunt, at a reading of 1A full-scale, has 1V
across it. Thus, two protection diodes in series would be needed to preserve
meter accuracy. Similarly, if the shunt resistor for a 1A full-scale meter is
1.5 ohm, the maximum shunt voltage is 1.5V--so three diodes are needed.
Glitch protection diodes should not be petite. Big, ugly diodes with a peak
current rating of 200a or more are best. Smaller diodes--and the meter they were
supposed to be protecting--can be destroyed during a glitch. Suitable glitch
protection diodes are 1N5400 (50PIV) to 1N5408 (1000PIV). In this application,
PIV is not important. The 1N5400 family of diodes is rated at 200a for 8.3mS.
During an extremely high current surge, a glitch protection diode may short
out--and by so doing protect the precious parts. Replacing a shorted protection
diode instead of a kaput meter is almost fun.
To reduce the chance of the negative HV circuit spiking to several kilovolts,
connect a string of glitch protection diodes from the negative terminal on the
HV filter capacitor to chassis. At 200a, each diode will limit the surge voltage
across it to about 1.5v. Typically, three diodes are needed--thusly limiting the
negative spike to about 4.5 volts. Diode polarity is: cathode band toward the
negative HV. With one simple wiring change, the same string of diodes can also
protect the grid I meter and the anode I meter. This dual protection technique
is incorporated into the Adjustable Electronic Cathode Bias Switch on Figure 7.
A HV arc can
destroy an indirectly-heated cathode tube. Here's how it happens: In some
amplifiers, one side of the filament/heater is grounded. The cathode is
connected to the negative HV circuit. If the negative HV spikes to several
kilovolts, the cathode will often arc to the grounded heater. At a minimum, this
breaks down the insulation between the heater and the cathode. Sometimes the
heater wire burns out--and sometimes the cathode arcs to the grounded grid.
Either way the tube is kaput.
Grounding one side of the heater is an invitation for cathode-to-filament
breakdown. Instead, let both heater wires float. If the heater is fed through an
c.40micro H bifilar RFC, one side of the heater can be wired to the cathode.
Even though this arrangement can not protect against cathode-to-grid breakdown,
it assures that the voltage between the filament and the cathode is unlikely to
rise to dangerous levels.
Manufactured
amplifiers typically use a safety device to automatically short the +HV supply
to chassis ground when the output section cover is removed. If the cover is
removed before the HV filter capacitors have discharged, the resulting positive
HV to ground short can damage the amplifier. In most g-g amplifiers, the only DC
current path between the negative HV circuit and chassis ground is the
grid-current meter and its shunt resistor. Even if the remaining charge in the
HV filter capacitors is only 200V when the short from positive to ground occurs,
without glitch protection diodes, the entire 200V appears across the
grid-current meter shunt and the grid-current meter. Many potentially-fatal
amperes can flow into the grid meter as the HV filter capacitors finish
discharging. If the amplifier is accidentally switched on with the cover
removed, rectifier diodes are a common casualty.
Automatic-shorting safety devices are not only dangerous to amplifier
components, they can be dangerous to operators. It is dangerous to assume that
an amplifier is safe to work on because it contains a safety device. Even though
an amplifier's HV supply is shorted, if the amplifier is plugged in, its
electric-mains circuitry is still alive and potentially fatal. Amplifiers are
inherently dangerous. They should not be worked on casually--even if they have
so-called safety devices.
The safest quick method of discharging HV filter capacitors is through a
paralleled pair of wire wound resistors. The resistors limit the discharge
current to a safe amount. In the unlikely event that one resistor opens, the
remaining resistor will do the job. For the average 1500W amplifier, a
paralleled pair of non-adjustable 1k ohm to 5k ohm, 50W resistors will do the
job. Its always a good idea to check the anode supply voltmeter before
putting your hands inside an amplifier.
Fuses have
current and voltage ratings. For a fuse, the real test is opening safely--not
operating without opening during normal operation. A fuse's maximum voltage
rating is important. In some amplifiers, ordinary 250V 3AG fuses are casually
used in circuits where they may be required to interrupt several kilovolts.
Examples are the anode or cathode circuit. All's well until a problem occurs.
When a 250V fuse attempts to interrupt a potentially-damaging flow of current in
such circuits, the frangible link inside the fuse parts as it should. However,
due to the available voltage, when the link melts, a metal vapour arc forms in
its place. Metal vapour arcs typically have a voltage drop of around 20V--so the
unsafe current will continue to flow in the circuit. At some point, the fuse
will eventually explode--usually after serious damage has been done to other
components.
During a glitch, circuits which normally carry low voltages can spike to several
kilovolts. For example, cathode circuits normally see a maximum voltage of 30V
to 100V. Thus, it might seem appropriate to use a 250V fuse to protect the
cathode from excessive current. However, when a glitch occurs, several kilovolts
can appear across anything that attempts to interrupt the flow of cathode
current. The safest place to use ordinary 250V fuses is in the primaries of
transformers.
There are
basically two types of DC filters: inductor-input / capacitor-output, and
capacitor. Each type of filter has advantages and trade-offs.
Capacitor filters have good transient response. Since no inductor and
resonating capacitor are used, the capacitor filter is simple to build, compact,
cost-effective, requires no tuning and it is lightweight. The main disadvantage
of a capacitor filter is that the capacitor is charged only during a small
fraction of the waveform supplied by the transformer. No charging current flows
until the instantaneous output voltage from the rectifiers exceeds the
instantaneous voltage on the filter capacitor. This means that the transformer
is either not loaded or severely loaded at different times during each cycle.
For example, with a electric-mains frequency of 60Hz, the duration of a half
cycle is 8.333mS. Under load, using a capacitor filter, the capacitor charging
time per half-cycle is typically only about 1mS out of the 8.333mS. This means
that the ratio between output current and peak charging current can be 8 to 1.
To combat I^2 R loss in a capacitor filter power supply, the transformer, all
circuitry in the primary (including the electric-mains) and the filter capacitor
should have low resistance. Capacitor filters are not appropriate for use with
older-design transformers that were intended for use with inductor filters.
Typically, such transformers have high winding resistance.
Inductor-input / capacitor filters can be of the resonated type or the
non-resonated type. A non-resonated inductor tries to maintain a constant
DC-current despite changes in the load current. This is the nature of any
inductor. It always tries to maintain constant current by temporarily increasing
the output voltage when the load current decreases suddenly, or by decreasing
the output voltage when the load current increases suddenly.
When a conventional voltmeter is used to monitor the output voltage from a
non-resonated inductor / capacitor filter power supply, the transient
unregulation characteristic will usually not be detected because of the damped
response in the meter movement. If a DC oscilloscope is used to monitor the
output voltage while a string of caround 5 WPM CW dashes are sent, the
instantaneous output voltage swings can be easily observed. On make, the output
voltage spikes downward. On break, the output voltage spikes upward. The
amplitude and width of the spike depends on how much filter capacitance is used
after the inductor and on the change in current. Upward and downward voltage
spikes of more than ±50% are possible during a sudden load change on a
non-resonated inductor / capacitor filter power supply.
Transient unregulation is probably not an important consideration unless SSB operation is used. With SSB, the PEP output and the linearity of the amplifier would be adversely affected by a non-resonant inductor DC filter.
The resonant DC
filter maintains a fairly constant output voltage during rapid or slow changes
in current demand--provided that a minimum current passes through the inductor.
This minimum current can be the zero-signal anode-current [a.k.a. 'idling
current'] of the tube itself. The inductor is resonated with a parallel
capacitor. In actual practice, the value of capacitance used--as well as the
bleeder resistance--is that which produces satisfactory voltage regulation. The
resulting resonant frequency is usually slightly higher than double the
frequency of the electric-mains. Resonant L/C pairs are available from Peter W.
Dahl, Inc.
DC filter inductors come in two types, fixed-inductance and swinging-inductance.
A swinging-inductor changes its inductance according to the current that is
passing through it. Obviously, a swinging inductor can not stay tuned correctly
with changes in current. Therefore, resonated-inductor filters can only
use a fixed inductor.
The disadvantages of a resonant inductor filter are:
The resonating capacitor must have a DC-working voltage rating of about three times the DC output voltage of the supply Typical values are 0.1 to 0.15 micro F @7.5kV to 15kV.
To maintain voltage regulation during standby, a minimum 'bleeder' current must flow in the inductor. Typically, the bleeder current is 10%. Considerable heat is dissipated by the bleeder resistor(s). However, if the filter capacitor can withstand the approximately. 50% increase in voltage during receive, the 10% standby bleeder current requirement can be reduced to 0.5%.
The inductor is heavy and costly.
A resonant inductor filter usually makes an audible noise--unless the inductor has been potted in plastic resin.
The advantages of a resonant-inductor DC filter are:
Excellent voltage regulation.
Greatly reduced peak current demand on the transformer and the electric-mains. Most importantly, this reduces transformer heating.
Transformers have about double the output current capability when a resonant inductor filter is used instead of a capacitor filter. However, the output V from a capacitor filter is higher.
The resonant filter is used extensively by commercial and military amplifier manufacturers. Since a resonant filter demands much less peak power from the electric-mains than a capacitor filter demands, for 120V operation, where available power is typically much more limited than with 240V operation, a resonant filter is clearly the best choice. The resonant filter is also the best choice for high duty-cycle modes such as RTTY, FM or AM.
Half wave. ----Advantages: may be used where one side of the AC input is grounded. Disadvantages: requires highest filter C; causes DC current to flow in the transformer; poor voltage regulation; transient-voltage protection is needed to protect the rectifier from reverse voltage spikes.
Fullwave-centertap---- The fullwave-centertap rectifier circuit was used in ancient times when tube-type rectifiers were the only game in town. Only one rectifier filament-winding was needed to produce full wave rectification, so the center-tapped secondary winding was the norm in older transformers. If a more efficient fullwave-bridge rectifier circuit had been used, three rectifier filament windings would have been needed. Advantage: if needed, reduces output voltage to one-half or to approximately. one-fourth of the DC voltage that would be obtained with a full wave bridge or full wave voltage-doubler. Disadvantage: inefficiently utilizes only half of the secondary winding at any instant--resulting in less than optimum transformer efficiency.
Full wave bridge---- Advantages: full utilization of the transformer's capability; may be used with a resonant filter. Disadvantage: requires twice as many transformer secondary turns as the full wave voltage-doubler requires. This means more layers of insulating paper--and that takes up winding space--so smaller wire must be used.
Full wave voltage doubler---- Advantages: full utilization of the transformer's capability; to achieve a specific DC voltage, only half as many transformer secondary turns are needed compared to a fullwave-bridge circuit. This means that the power transformer secondary will have a higher ratio of copper to paper. If switched secondary taps are used to control the DC output-voltage, the voltage stress on the switch is only half of what it would be with full wave bridge rectification. Full wave doubler supplies typically have a remarkably low ripple content. This is because one half of the filter capacitance is being charged at the same time the other half is being discharged. Since the charging sawtooth waveform is similar and opposite to the discharging sawtooth waveform, the result is a fairly smooth DC output. Disadvantages: To achieve acceptable voltage regulation, the full wave doubler requires twice as much filter capacitance as a fullwave-bridge. This is the case because each of the two filter capacitor sections is charged once per cycle versus being charged twice per cycle with the fullwave-bridge circuit. Thus, in a full wave doubler, the filter-capacitors must be able to hold their charge twice as long--so twice as much filter capacitance is needed. The capacitance requirement is easily met with modern aluminum electrolytic capacitors. They provide a large amount of capacitance in a small space at a reasonable cost. Full wave voltage doublers are not practical for use with a resonant inductor filter
Transformers are
available in two basic types: E-I (conventional) core and toroidal core. The E-I
core is made from a stack of thin E-shaped and I-shaped iron plates. When placed
together they form a rectangle with two windows for the windings. A stack of E-I
rectangles make the completed core. The toroidal core is made from a continuous
tape of grain-oriented material that contains iron and silicon plus other
elements that increase the permeability of the core and decrease loss. This core
material is known as Hipersil. Westinghouse Corp. was the original patent and
copyright holder. Their patent expired decades ago. There are different grades
of Hipersil tape. Grade 5 has the highest performance. Grade 22 has the lowest
performance.
Higher permeability means that fewer turns are needed to achieve the required
inductance in each winding. This means that larger diameter wire can be used.
The end result is a transformer with low resistance and high efficiency. The
Hipersil core is so efficient that the principal loss factor is the resistive
loss in the copper wire. Hipersil core transformers are capable of producing
extremely high peak currents. Thus, the Hipersil core transformer is ideally
suited for capacitor filter power supplies.
It is difficult and time-consuming to thread a continuous tape core through the
completed transformer windings--so someone came up with a faster way of uniting
the core with the windings. Here's how it's done: The tape is wound on a form of
the appropriate dimensions. The tape is spot welded together, removed from the
form, and annealed at about 700 degrees C to relieve internal stresses. After
cooling, the core is varnished and dried. Then the core is cut in half with a
machine that makes a precise square cut. The faces of the cut are then polished
flat. Thus, the halves of the core can slip into the completed windings, contact
each other closely--restoring nearly perfect magnetic coupling between the
halves of the core. The matched halves of the core are marked so that they can
not be inadvertently mixed up with other core halves. The reunited halves of the
core are held together tightly by steel bands like those used for binding heavy
cartons and crates. If future access to the primary and secondary windings is
needed, a Hipersil transformer can be disassembled by cutting the steel bands
and removing the core halves.
Hipersil is no longer the most efficient type of core material. The new
amorphous core transformer is starting to come into use by electric utilities.
An amorphous core transformer is so efficient that if the secondary is unloaded
and the primary is disconnected from the electric-mains, the collapsing magnetic
field generates a voltage spike that can destroy the transformer. To avoid this
problem, one winding is paralleled with a suitable voltage surge absorber.
Transformers are
commonly rated in maximum "volt-amperes" [VA]. Maximum VA are roughly equal to
maximum RMS watts when the rated RMS current is flowing in each transformer
winding and the transformer is operated from the rated input voltage at the
design frequency. If the electric-mains voltage is reduced, the VA capability of
the transformer decreases.
For SSB and CW operation, a lighter transformer may do the job just as well as a
much heavier and more costly transformer. Manufactured 1500W amplifiers
typically use a HV transformer with a continuous capability of roughly 600W--or
VA. Such transformers are completely satisfactory for normal SSB operation. Such
transformers are also capable of handling brief FM and RTTY
transmissions--provided that the lower voltage tap is used.
If a power supply's DC output voltage drops more than about 10% under
modulation, it's a fairly safe assumption that a more capable transformer is
needed. Of course, not using enough filter capacitance or excessive
electric-mains resistance can also cause poor regulation.
Increasing the current in any
conductor causes a square-law increase in the amount of power dissipated in the
conductor. Since P=I^2 x R, doubling the current causes a 4 times increase in
dissipation. This is an especially important consideration with transformers
because they have considerable difficulty dissipating the heat that is generated
deep inside their windings. This problem is compounded because copper has a
positive, resistance versus temperature, coefficient. Thus, as the copper heats
up, its resistance increases--which increases the dissipation--which increases
the resistance, et cetera. This can lead to thermal runaway and transformer
failure.
If a transformer has a secondary rating of 1A RMS, it means 1A with a resistive
load. If connected to a rectifier and DC filter, the 1A rating does not
necessarily apply. For example, if a fullwave-bridge rectifier, resonant filter
circuit is used, the RMS current rating can be multiplied by at least 1.2. The
DC output voltage will be about 0.85 times the RMS voltage. If a fullwave-bridge
rectifier, capacitor-filter circuit is used, the loaded DC output voltage will
be about 1.3 times the RMS voltage. A 30% increase in voltage sounds good, but
obviously you don't get something for nothing. The trade-off for the increase in
voltage is a decrease in current capability. The high peak current demanded by
the capacitor filter translates into a substantive current capability decrease.
.
Any formula for converting a transformer's RMS current rating to a DC output
current rating is bound to be problematic due to the large number of variables.
Here's a rule-of-thumb that is fairly accurate. If, after about an hour of
typical operation, the outside of the transformer is uncomfortably hot for one's
thumb, the internal parts of the transformer are probably deteriorating.
Reducing the average load current slightly will greatly reduce transformer
heating because of the square-law relationship between current and power
dissipation. For example, reducing the current by 30% will reduce winding
dissipation by about 50%.
There is a simple, reasonably accurate, 2-step approximation for determining the
safe SSB, maximum current rating for a specific transformer for use with a
capacitor filter and a full wave bridge rectifier. A slightly different
approximation is used for a full wave voltage-doubler. These approximations are
based on the DC resistance and the AC-voltage of the transformer's secondary
winding. These approximations are useful when shopping around surplus stores or
swap meets. All that's needed is an ohm-meter and a clip-lead. The clip-lead is
used to short the primary of the transformer. This dampens the inductive voltage
spike that occurs when the ohm-meter is disconnected.
The fullwave-bridge, capacitor filter approximations are: Multiply the secondary
winding resistance by 70 to find the minimum intermittent load resistance that
can be placed on the power supply. To find the DC output voltage under load,
multiply the secondary RMS-voltage by 1.3. To find the safe intermittent current
rating for SSB service, use Ohm's law and divide the output voltage by the
minimum load resistance. For a more accurate evaluation, use the appropriate
graphs in this book.
For example, a 2000V RMS secondary winding has a DC-resistance of 60 ohms. A
full wave bridge rectifier, capacitor filter, circuit will be used. The safe,
minimum, intermittent load resistance is approximately 70 x 60 ohm = 4200 ohms.
The approximate voltage delivered under load would be 1.3 x 2000V = 2600V DC.
Thus, the maximum intermittent load current is 2600V ÷ 4200 ohms = 0.62a.
Another approximation can be used to find the amount of filter capacitance
needed. The approximation is 50,000 divided by the minimum load resistance. In
the above example this is 50,000 ÷ 4200 = 12micro F.
For a full wave voltage-doubler, capacitor filter, power supply, the SSB-service
approximations are: Minimum intermittent DC-load resistance equals 300 times the
winding resistance; DC-output voltage, under load, equals 2.5 times the
secondary RMS voltage.
For example: A 1000VRMS transformer has a winding resistance of 10 ohms, the
minimum load resistance for full wave doubler operation would be 300 x 10
ohms=3000 ohms and the output voltage would be 2.5 x 1000V=2500V. The maximum
intermittent load current is 2500V ÷ 3000 ohms = 0.83A.
The amount of filter capacitance needed for each half of the full wave voltage-doubler
circuit is approximately 200,000 divided by the minimum load resistance. In the
above example each of the two capacitors should have a minimum of 200,000 ÷ 3000
= 67micro F.
There is more to transformer performance than secondary resistance. If a
Hipersil® core is used, core loss is minimal and the maximum intermittent power
capability increases. Primary resistance is another factor to consider since it
is effectively in series with the electric-mains resistance. Electric-mains
resistance can cause a voltage drop problem if the amplifier is a fair distance
from the service entrance box and you are using a capacitor filter power supply.
One solution is to use larger diameter wires than the electric code requires.
Another solution is to install the power supply near the service box and bring
the HV DC to the amplifier.
It's nice to
have the ability to reduce the output power from an amplifier. One way to do
this is to reduce the anode-voltage and anode-current simultaneously so that the
output load-R of the amplifier tube or tubes does not change appreciably. This
allows the tank circuit to function at its design Q for both high and low power.
Switching primary taps is not an efficient method of reducing output voltage
because in order to do so extra turns must added to the primary. To make room
for the extra turns, the primary's wire diameter must be decreased--and that
increases R. An efficient method of reducing the DC output voltage in a HV power
supply is by switching secondary taps on the transformer. If a fairly ordinary
ceramic rotary switch is insulated from the chassis, it can easily perform this
job. The taps should not be switched under load.
If no secondary tap is provided on a transformer, it is possible to lower the
output voltage 50% by switching from fullwave doubler to fullwave bridge
rectification. All that's needed is a suitable SPST vacuum relay, or
well-insulated ceramic switch, two filter capacitors and four strings of
rectifiers. For example, a power supply that produces 4000V for SSB could be
operated at 2000V for RTTY, CW, or FM. The DC output current capability doubles
when the output voltage is halved during fullwave bridge operation--just what's
needed for FM's and RTTY's much higher duty-cycle. When switching the voltage
output, it is best to temporarily switch the power supply off and then restart.
The output voltage may be switched down without switching the supply
off--provided that the amplifier is in standby.
On paper, the
variable auto-transformer, a.k.a. Variac® or Powerstat®, looks good. Variacs/Powerstats
are intended to be used with resistive loads. When a Variac is set at or near
100% of the input voltage, it adds only a small amount of series R. However,
when a Variac is set to produce a fraction of the input voltage it adds more
series R. This is of little consequence with resistive loads. However, when the
load is a capacitor filter DC supply, due the demand for high peak current,
additional series R is most unwelcome. Although Variacs perform acceptably with
resonant choke filter power supplies, using a Variac to control the output
voltage from a capacitor filter supply is not good engineering practice.
A Variac can be used in place of a step-start relay. Provided that the operator
always remembers to set the Variac to near-zero before switching the amplifier
on, all will be well. A step-start relay offers some advantages: it is cheaper,
mistake-proof, saves many kilograms, and it adds substantially less series R.
There is, however, an appropriate step-start application for a Variac. Eimac®
recommends using a motor-driven Variac, feeding the filament transformer
primary, to bring up and bring down the filament-voltage (over a period of two
minutes each) on its tubes which incorporate water-cooled filament supports. An
example is the 8973 tetrode--just what you need if you are building a 600kW
linear amplifier.
Most transformers use paper to separate and insulate each layer of windings. Paper is hygroscopic--i.e., it absorbs water vapour from the air. The presence of water reduces the insulating ability of the paper. In time, insulation breakdown is likely. The solution is to pot the windings. Plastic resins are best. Petroleum tar is next best. Since potting fills up the air spaces in the windings--and air is a poor heat conductor--potting also improves heat transfer--thereby reducing internal temperature and increasing MTBF. Potting adds very little to the initial cost of a transformer and subtracts substantially from the long-term cost. Some custom transformer manufacturers offer potting as an extra-cost option. Peter Dahl Co. has a potting option.
Commercial
transformer potting is normally done in a vacuum chamber to facilitate the
evacuation of air bubbles. However, with a little patience, it is possible to
pot transformers satisfactorily without special equipment. Bake the transformer
in an oven at a temperature of about 175 degrees F/80 degrees C. Bake for two to
three hours per pound. Baking drives out internal moisture. After baking, place
the transformer on a table covered with a thick layer of newspapers. Position
the transformer so that the leads or lugs are down. Using masking tape, seal the
end of the transformer windings opposite the leads/lugs so that liquid can not
escape easily when the transformer is inverted.
Polyester fiberglass laminating resin is designed to flow into small spaces and
expel air bubbles. It can be used for potting transformers.
In a clean tin-can, pour in a quantity of laminating resin that will fill up the
air spaces in the bottom 5% of the transformer's windings. Using pliers, bend
the rim to facilitate pouring the resin. Mix in about 5 drops of catalyst per
ounce of resin. Depending on the ambient temperature and humidity, this amount
of catalyst will result in a moderately fast gel time. Pour the resin slowly
into the windings. Resin pouring should be done steadily and from only one area
of the windings to avoid trapping air bubbles. Any leaks from the bottom can be
patched by forcing raw silicone rubber into the area of the leak. When the resin
gels, it forms a thin bottom plug. The bottom plug need not be more than about
5mm thick.
Pour an amount of resin into the can that will fill the remainder of the
windings. For a several-kVA transformer, use about 1.5 drops of catalyst per
ounce of resin. For smaller transformers, slightly more catalyst is needed. The
resin must not gel before the air bubbles have had a chance to escape--so it is
better to err on the light side for the amount of catalyst.
Heat increases the fluidity of the resin--hastening the exit of bubbles.
However, heat tends to decrease gel-time. Internal transformer heating is
accomplished by forcing current through the windings with a Variac. Connect the
Variac to the highest voltage winding. Short the highest current winding with an
AC ampmeter. Increase the voltage until the ampmeter indicates the rated winding
current. At this level fairly normal internal heating results. As soon as the
resin begins to gel, stop the current and direct a cooling fan at the
transformer. Resin-gelling is an exothermic reaction.
The most
frequent failure mechanism for HV power supply rectifiers is too much reverse
current. This problem can be virtually eliminated in 50Hz/60Hz, fullwave bridge
and fullwave doubler, capacitor filter circuits if the total PIV in each string
of diodes exceeds the no-load DC output voltage by at least 50%. For operation
in high-temperature environments, a 100% factor may be needed.
Modern solid-state rectifiers are made differently than they were 30 years ago.
In those ancient times, same-type rectifiers did not have uniform reverse
characteristics. Rectifier failure was common. In an attempt to compensate for
the inherent weaknesses in early solid-state devices, rectifier protection
schemes were used. Resistors and capacitors were paralleled with series
rectifiers--probably a take-off on the practice of using equalizer resistors on
electrolytic filter capacitors. However, in any series circuit, the currents in
all of the elements are exactly equal. Thus, when rectifiers are in series, the
reverse current burden is exactly the same for each rectifier--provided that no
parallel resistors are used. Manufacturers of series rectifier units long ago
abandoned the practice of using parallel resistors and capacitors. The 1995/6/7
Radio Amateur's Handbook explains why rectifier 'equalization' is prone to cause
premature rectifier failure.[page 11-9, middle column, top]
Series-connected rectifiers should be of the same type. Mixing rectifiers types
in the same series string could cause a problem during the reverse half-cycle.
When a rectifier has been conducting, it takes a finite amount of recovery time
for the rectifier to stop conducting after the source of forward current reaches
zero. It is important that a rectifier not be conducting when the reverse
voltage arrives. This can be a problem when rectifying high frequency AC or when
rectifying square waves. Paralleling a capacitor with each rectifier may help
the rectifiers to stop conducting sooner. If you need to rectify high frequency
AC, one solution is to use fast-recovery epitaxial rectifiers. 1000PIV, 1A, 70
nanosecond recovery time units are currently priced at about 50¢ each in
quantities of 100.
Rectifiers that have a rating above 1kV PIV are typically made from a series of individual rectifiers that are entombed in an epoxy package. This arrangement makes for a neat-appearing installation--but there is a trade-off. Epoxy is a poor conductor of heat. Individual series-connected diodes mounted on perfboard and exposed to open air dispose of heat much more efficiently than do multi-diode packages.
Filter
capacitors usually have a ripple-current rating. The ripple-current rating
should be at least equal to the maximum DC output current capability of the
supply. Quality filter capacitors are designed to minimize equivalent series
resistance [ESR]. Low ESR ohms translates into a high ripple-current rating.
Oil-filled capacitors are available in two types: filter service, for use in
power supplies, and flash service for use in photography or pulsed laser
applications. The flash capacitor is designed for maximum capacitance per unit
volume. To reduce volume, very thin metal foil is used to make the plates of a
flash capacitor. Thin plates have more ESR--so they dissipate more I-squared
times R power when they are subjected to ripple-current.
For longest life in high duty-cycle applications, cool air should be allowed to
circulate freely around filter capacitors.
According to some manufacturers, flash capacitors can be used in filter service
if they are operated at 60% of their rated peak volts. In intermittent duty
applications, it may be possible to use flash capacitors at more than 60% of
their rated peak-voltage. To discover how a flash-capacitor is faring in
ripple-current service, after about an hour of contesting, if the capacitor is
warm to the touch, an internal heating problem is indicated. Internal heating
causes expansion and stresses the capacitor's case--which may eventually come
apart at the seams and begin to leak dielectric oil.
If not plainly stated on its label, there is a way of determining the intended
type of service for an oil-filled capacitor. Flash-capacitors usually have a
peak voltage [PV or VP] rating. Filter-service capacitors are usually rated in
DC working volts [DCWV]. Capacitors can also be rated in AC working volts. To
convert AC-working volts to DC-working volts, multiply the AC voltage by three.
There have been instances where surplus flash capacitors were offered for sale
with altered or counterfeit markings. For example, a capacitor that was
originally marked "3.5kVP" was changed to "5kV" by erasing characters. Thus, a
capacitor that should have been de-rated to 0.6 x 3.5kV = 2100V for filter
service would appear to be good for 5kV. A practical way of determining whether
an oil-filled capacitor can withstand ripple-current is to connect it in series
with an AC-ampere meter and an AC voltage source. The voltage is adjusted until
the AC current is equal to the expected maximum output current of the power
supply. If, after an hour, the capacitor shows little or no internal heating,
you have a winner.
There is also a flash service type of aluminum electrolytic capacitor, that is
not designed to handle ripple-current.
Electrolytic filter capacitors are intolerant of reverse current and heat.
Electrolytic capacitor working voltage [WV] ratings should be treated with
respect. The WV rating is virtually the maximum voltage rating. Despite their
more delicate nature, electrolytic filter capacitors offer substantial
advantages over oil-filled filter capacitors. The main advantages are more
joules of energy storage per dollar, reduced weight and reduced volume.
When electrolytic capacitors are operated in series, they should share the
voltage equally. In order to do this, a voltage equalizer resistor is connected
across each capacitor. Equalizing resistors must have fairly equal
resistance--and their resistance should not change appreciably during aging. If
an equalizer resistor changes value appreciably, domino-effect destruction of an
entire section of filter capacitors may result.
There is no formula for determining the optimum resistance for an equalizer
resistor. Less resistance equals less bleed-down time. However, less resistance
produces more heat. A compromise is in order.
Carbon-composition resistors change resistance with age. This characteristic is
unacceptable for equalizer resistor service. High resistance, wire wound
resistors are wound with extremely fine resistance wire. They are not remarkably
reliable. Metal oxide film [MOF] resistors are more reliable. The initial
resistance of a MOF resistor is typically much closer to the labeled value--and
it will stay that way for many years. A Matsushita/Panasonic® 3W, 100k ohm MOF
resistor makes a good equalizer resistor for 450V capacitors. It produces a
reasonable bleed-down time and a reasonable amount of heat. These resistors are
available from Digi-Key.
Electrolytic filter capacitors are ruined quickly by reverse current. Reverse
current often occurs when a rectifier fails. To protect electrolytic capacitors
from reverse current, connect a >600PIV diode across each capacitor. The cathode
band of the diode connects to the capacitor's positive terminal.
When a
grounded-grid amplifier's operating bias is obtained from a single Zener diode,
there is no way to compensate for tube variation. One solution is to obtain the
operating bias from a series string of forward biased rectifier diodes. By
switching the number of diodes in and out with a rotary switch, the bias can be
changed in approximately 0.7V increments.
Traditionally, a mechanical relay has been used to switch amplifier bias between
receive and transmit. An optoisolator coupled to a transistor switch, i.e., an
electronic bias switch, can do this job faster, more reliably, sans-noise, and
cheaper.
There are principally two means of actuating electronic bias switches--RF-actuation
and coil-current actuation. Although it sounds hip, RF-actuation creates two
problems. The amplifier rapidly switches between linear bias and non-linear bias
during softly-spoken syllables of speech. This causes choppy-sounding audio and
splatter. When the electronic bias switch is controlled by the current that
passes through the RF relays' coils, it is not possible to intermittently switch
the amplifier into non-linear bias during transmit. Coil current actuated bias
switching can be accomplished with an optoisolator. The optoisolator's input LED
is driven by the coil-current. The output of the optoisolator drives the bias
switch transistor.
Since the grid draws virtually zero current, it is easy to make the bias continuously adjustable in Class AB1 operation. Typically, the cutoff bias voltage during receive will be about 50% higher than the transmit bias voltage. An optoisolator driving a HV FET can be used to switch the bias between transmit and receive. A circuit is provided.
A conventional relay switches in roughly 25mS. Such relays have traditionally been used for RF and bias switching in RF amplifiers. This was acceptable when transceivers also used conventional relays. Currently manufactured transceivers are designed for AMTOR, QSK telegraphy, and unobtrusive SSB VOX operation. Modern transceivers T/R and R/T switch quietly, and do so in as little as 5mS. Such radios typically use a high-power SPDT reed relay to switch the antenna between transmit and receive. Similar relays can be used for amplifier input RF switching. Jennings and Kilovac manufacture high speed, SPDT vacuum-relays that have a continuous rating of 7A at 32MHz [2450W into 50 ohms]. The Jennings relay is the RJ-1A. Kilovac's relay is the HC-1. When used with a speedup-circuit, either relay can switch in under 2mS. Although both manufacturers make DPDT RF vacuum-relays, none are as speedy as their fastest single pole models. Thus, separate input and output relays are usually faster than a single DPDT relay.
Another device
that can be used for high-speed RF switching is the PIN [P-Intrinsic-N] diode.
PIN diodes are similar to 1000PIV rectifier diodes--i.e., they have a wide
intrinsic region. PIN diodes are utilized extensively in radars as
transmit-receive switches.
A PIN diode is switched off by applying DC reverse voltage to widen its
intrinsic region. The PIN diode is switched on by passing DC current in the
forward direction to fill its intrinsic region with current carriers. PIN diodes
are extremely fast switches. Their lifetime is virtually unlimited as long as
the allowable PIV is not exceeded.
The typical reverse breakdown voltage rating for a PIN diode is around 1000V. A
legal-limit amateur radio amplifier produces an output voltage of about 800
volts peak-to-peak [p-p] into a 50 ohm load--so a 1000PIV PIN diode is more than
adequate. When the load Z is higher, due to a somewhat less than wonderful SWR,
the switching device may be exposed to more than 1000Vp-p. This poses no problem
for a typical high-speed vacuum-relay. Even if a vacuum-relay's breakdown
voltage is temporarily exceeded, there is little likelihood that permanent
damage to the vacuum-relay's contacts will result. However, solid-state devices
are not so forgiving. A single voltage-transient can destroy a PIN diode.
For 100 WPM computer-CW, the PIN diode is clearly the only choice. For 30 WPM CW,
AMTOR and high-speed VOX, a vacuum-relay has advantages.
Different types
of solid-state components are rated somewhat differently. Some ratings are
realistic. Some ratings are not realistic. The maximum ratings of large
transistors and large Zener diodes can not be realized unless drastic, extreme
measures are used to keep the case temperature below the maximum allowable 25
degrees C at full ratings. In the real world, operation at 30% of a published
dissipation or current rating is usually safe. Additionally, bipolar power
transistors suffer from a generic weakness called secondary-breakdown
phenomenon. For example, a "1500V, 8A, 150W" power transistor may only be able
to safely dissipate 15W at moderate collector-to-emitter voltages. T-MOS power
FETs are much more resistant to secondary-breakdown.
Wire-lead rectifier current ratings are fairly realistic when they are mounted
on perfboard and cooling air is allowed to circulate freely around individual
rectifiers.
There is some variation in the inverse breakdown voltage of solid-state rectifiers of the same type. Measuring the breakdown voltage of each rectifier diode is a good precaution. When a diode's reverse current reaches 1 to 2 uA, the voltage across the diode is the breakdown voltage. Exceeding this voltage is likely to be fatal. As operating temperature increases, breakdown voltage decreases.
A breakdown
voltage tester (a.k.a. high-pot) could be described as a variable HV Ohmmeter
that does not read directly in ohms. It is a useful tool. Breakdown testers are
essential for testing vacuum relays, vacuum capacitors, blocking capacitors,
air-variable capacitors, rectifiers, and for finding problems with insulation.
Building or troubleshooting a tube-type RF amplifier without a breakdown tester
is like crossing an ocean without a navigation instrument. For most amateur
radio applications, the highest voltage component rating commonly encountered
[with vacuum-capacitors and vacuum-relays] is 15kVDC/9kV RF peak, so a 0 to 15kV
breakdown tester should suffice.
Although commercial breakdown testers are available, they are not inexpensive. A
suitable breakdown tester can easily be constructed from mostly-surplus parts.
The main parts are a 50/60Hz low-current HV transformer, a >1A, 120V variable
transformer, a 120V incandescent bulb, some diodes, resistors, a sensitive uA
meter and two HV filter capacitors.
Commercial, low-current HV DC supplies may also be used provided that they are
connected to a Variac in series with a 120V incandescent light bulb to limit
current. The bulb limits the short-circuit current to a safe value--obviating
the need for a fuse. The wattage of the bulb is roughly proportional to the
wattage of the supply. The rated [I=P/E] bulb current should be similar to the
appropriate fuse rating for the HV supply primary. A multi megohm resistor is
used to limit the current flow into the device under test. The uA meter should
be protected with back to back 1A diodes. A circuit is provided.
Most power supplies benefit from something to soften the shock of start-up. A 10A DPST-NO or 10A DPDT relay and two approx..25 ohm 10W resistors are just about all that's needed to add a step-start circuit to the average 1500W amplifier. The step-start circuit goes in series with the mains fuses or circuit breakers. With this arrangement the filaments, the HV supply and the LV supplies enjoy the benefit of a kinder and gentler start-up.
Every HF
amplifier has at least two resonant circuits in its output circuitry. The more
obvious one is the HF-resonant tank circuit. A less obvious one is the
VHF-resonant circuit that is principally formed by the anode capacitance and the
inductance of the conductors between the tank circuit and the anode. In 1500W
amplifiers, anode resonance typically occurs around 100MHz--well within
manufacturers' ratings for "Amplifier and Oscillator Service" for the tubes that
are commonly used in such amplifiers.
The equivalent resistance of a high Q parallel resonant circuit is virtually
infinite. A low Q parallel resonant circuit has a relatively low equivalent
resistance.
The voltage gain of an amplifier tube is roughly proportional to the load
resistance. High load resistance produces more gain. Low load resistance
produces less gain.
If the conductors in the anode resonant circuit have a high VHF Q, the
equivalent load resistance presented to the anode will be high and the tube will
exhibit increased voltage gain at the VHF resonance. If the conductors in the
anode resonant circuit have a low VHF Q, the load resistance presented to the
anode will be low and the voltage gain at the VHF anode resonant frequency will
be reduced. Of course, if no VHF energy were present, it would make no
difference how much VHF gain an HF amplifier had.
When a transient current passes through a resonant circuit, the resonant circuit
rings like a bell--producing a damped sine wave signal. This is how ancient
spark transmitters produce RF--and the larger ones produced many kilowatts of
it.
Whenever the anode-current in an HF amplifier changes, a small VHF damped sine
wave signal is produced in the anode's VHF resonant circuit. This signal can be
observed with a VHF oscilloscope or a spectrum analyzer. The amplitude of the RF
voltage produced is proportional to the Q of the anode resonant circuit. If none
of this damped wave signal were fed back to the input, there would be no
problem.
In a grounded-grid amplifier, the grid appears to shield the input from the
output. In a grid-driven Class AB1 amplifier, The RF-grounded screen appears to
shield the input from the output. However, no grid and no screen is perfect--so
some of the damped wave VHF signal at the anode is capacitively fed back to the
input--and amplified.
Although it's unlikely, if the phase and amplitude of the damped wave signal
happens to be just right, oscillation at the anode's VHF resonance can occur. If
the VHF energy that is produced could find its way to a load, no danger would be
posed by a VHF oscillation. However, the HF tank circuit is a low pass filter
that effectively blocks VHF energy. Thus, the oscillator is unloaded and the
resulting grid-current is very high. The unloaded condition can cause VHF
voltage transients in the anode circuit. These transients may cause
tune-capacitor arcing and band switch arcing across open contacts. Since they
are closest to the anode resonant circuit, open tune-capacitor padder contacts,
as well as open 10m contacts are most vulnerable to parasitic-instigated arcing.
Band switch contacts can be melted and/or vapourized by such occurrences.
On page 72, the
1926 Edition of The Radio Amateur's Handbook tells us how to build
an improved VHF parasitic suppressor The logic was elementary. A suppressor is
supposed to dampen the anode circuit. Since low Q is synonymous with high
dampening, why not decrease Q by using resistance-wire? Quoting from page
72:........ "The combination of both resistance and inductance is very
effective in limiting parasitic oscillations to a negligible value of current."
After 1929, someone forgot to include this information in the Handbook. In those
days, the oversight probably didn't matter very much. Large amplifier tubes
generally had low VHF amplification, so VHF instability was not a major issue.
During the ensuing decades, people got into the habit of using parasitic
suppressors made from copper or silver-plated copper. This was an easy habit to
get into since copper and silver can be soldered more easily and cheaply than
nickel/chromium [nichrome] resistance wire. Meanwhile, the performance of
amplifier tubes kept improving. Because of these improvements, modern
high-amplification tubes appear to benefit more from 1926-vintage low VHF-Q
parasitic suppressors than 1926-vintage tubes. NOTE: In 1926, a 'high-mu' triode
had a mu of around 40.
Low VHF-Q
conductor material can be used to increase VHF loading in the anode resonant
circuit. Nickel-chromium-iron alloys are best, Nickel-chromium (nichrome) and
some types of stainless-steel are almost as good. The use of copper, aluminum,
and silver should be kept to a minimum. However, good conductors are desirable
beyond the tune capacitor, which marks the end of the anode VHF resonant circuit
and the beginning of the HF tank circuit.
Output Z: The output impedance of most tubes is a matter of
kilo-ohms--not ohms. There is no scientific reason to use "heavy duty"
conductors between the anode and the tune capacitor. If good VHF stability is a
design goal, it's best to use conductors that are no larger than is necessary to
carry the highest current present, i.e., the 10m RF circulating current between
the anode (output) capacitance and the tune capacitor. Round conductors have a
lower VHF Q than flat conductors. To increase current handling ability, or to
reduce inductance, two paralleled round conductors, separated by a wide air-gap,
are better than a flat conductor of the same overall width.
When designing the layout for an HF amplifier, locate the tune capacitor fairly close to the anode. This reduces the inductance in the anode resonant circuit--increasing the VHF resonant frequency. If the distance from the anode to the tune capacitor is grossly excessive, the 3/4 wave anode resonance may cause instability problems--especially with tubes that are UHF-rated like the 8877, 8874, and 3CX800A7.
In order to prevent a conductor from sharing the anode VHF-resonant circuit with the HF tank circuit, connect the tank inductor directly to the tune capacitor. It is best not to connect the tank inductor directly to the blocking capacitor.
The output enclosure in an HF amplifier is a high Q VHF-resonant cavity. The output enclosure can become a player in parasitic oscillations. Cavity dampening may be needed. This is a common practice in commercial high-power amplifiers. Closed loops of resistance-wire can be used to dampen cavity resonances.
In some cases, the HV-RFC contains a VHF resonance that abets parasitic oscillations. This problem is indicated by several burned turns appearing on the HV-RFC after a glitch. To isolate a problematic HV-RFC VHF resonance, place at least one VHF attenuator-rated ferrite bead on the lead between the top of the HV-RFC and the anode circuit.
The simplest
type of parasitic suppressor is a resistor. It reduces Q by adding R. This
technique is effective. However, it is mostly limited to low power applications.
The traditional staggered-resonance parasitic suppressor provides two advantages
over a resistor suppressor--it can handle more current, and it causes the VHF
resonance to work against itself.
A staggered-resonance parasitic suppressor typically consists of a coil inductor
paralleled with a low-inductance resistor. The axis of the coil inductor is
parallel to the resistor. Here's how it works: The magnetic field from current
flowing in the resistor is at a right angle to the direction of current flow.
The magnetic field from the inductor is parallel to the direction of current
flow. Because the two magnetic fields are 90 degrees apart, the inductances act
independently instead of mutually. The two independent-inductances connect to
one fixed-capacitance--i.e., the anode. Since the coil has more inductance than
the resistor, it creates a second VHF resonance that is slightly lower in
frequency than that produced by the resistor. The conflicting resonances work
against each other. This technique broadbands the anode's VHF resonance, i.e.,
it reduces the Q--very much like stagger-tuning IF transformers to widen a
receiver's band pass. Reducing VHF Q lowers the parallel equivalent VHF load
resistance on the anode. This reduces the VHF voltage gain--and that reduces an
amplifier's ability to oscillate at VHF.
Choosing the optimum amount of inductance for a suppressor inductor [Ls] is best
determined experimentally by operating the amplifier on 10m. Since 10m is almost
VHF, a device that suppresses VHF energy should get hot from 10m RF. If Ls is
too small, the suppressor resistor [Rs] will not exhibit visible signs of
heating during operation on 10m. If Ls is too large, there will be too much
voltage drop on 10m and Rs will burn out.
Every straight conductor has inductance. The amount of inductance is fairly
proportional to length. Manufactured high-power "non-inductive" resistors are
long--and therefore somewhat inductive. They are too inductive for use in VHF
suppressors. However, it's easy to make a suppressor resistor of sufficiently
low inductance by paralleling straight nichrome wires that are separated by
air-gaps.
Staggered-resonance suppressors can be built without a resistor by paralleling
two unequal-inductance nichrome wires. For example, a silver plated strap in the
anode circuit can be changed from a potential source of grief into a Q-reducing
asset by replacing it with two, parallel, nichrome wire conductors. One of the
conductors is made about 25% longer than is necessary to span the distance. Its
length is shortened by winding a small 1 to 2 turn coil. The axis of the coil is
parallel to the shorter wire. This arrangement decouples the magnetic field in
the coil from the magnetic field in the straight conductor.
For large amplifiers, 100% nichrome staggered-resonance suppressors solve the
problem of not being able to find a high-power resistor of sufficiently low
inductance for use in parasitic suppression service. For very large HF
amplifiers, all-nichrome staggered-resonance suppressors should be made from
flat nichrome conductors in order to carry the large RF circulating current
between the tune capacitor and the anode capacitance.
In two-tube amplifiers, if the two suppressors are allowed to magnetically
couple to each other, a VHF parasitic oscillation may occur. In a two-tube
amplifier, the suppressor coils should be positioned at a right angle. If the
suppressor coils are parallel to each other, the coils should be wound in
opposite directions, and separated as much as is practical.
Some amateur radio operators--and
some electronic engineers--do not believe that VHF oscillations can take place
in an HF amplifier. This is understandable because the most common and most
destructive type of VHF-parasitic-oscillation, the push-push variety, lasts only
a matter of microseconds. Push-pull VHF parasitic oscillation is obviously only
possible in multi-tube amplifiers. A steady oscillation between anodes is the
result. Push-pull parasitic oscillation is characterized by extremely high anode
dissipation, moderate grid and anode currents with zero drive, and no arcing.
Push-pull parasitic oscillations can be stopped by switching the amplifier to
standby. This is not possible with a push-push parasitic oscillation---wherein
the event is probably over before the sound of the parasitic arc reaches the
operator's ears.
VHF parasitic oscillations are not cooperative. It may take a particular
sequence of anode-current transients to initiate a parasitic oscillation. Even
though there is no concrete scientific evidence to prove it, the phrase "CQ
contest" may have a propensity to produce the key sequence of anode-current
transients--especially if the contest is one you've been waiting for, and the
local radio parts emporium has just closed for the weekend.
A major factor with parasitics is the VHF-gain of the particular amplifier-tube
or tubes that happen to be installed in the amplifier. Even among new
amplifier-tubes from the same production lot, there is some variation in VHF
gain. Tubes with below average VHF gain may never have a
parasitic-oscillation--no matter how poorly their parasitic-suppressors perform.
Thus, when below average gain tubes happen to be installed in an amplifier, it's
easy to assume that the amplifier design is perfectly stable.
Since catching a parasitic oscillation in the act is virtually impossible with
ham-type test gear, a different analytical approach must be used. It's a
reasonable assumption that a resonant circuit which supports parasitic
oscillation can be found and evaluated with a dipmeter.
To determine the parasitic frequency and evaluate the parasitic suppressor,
unplug the HF amplifier from the electric-mains and measure the anode-resonance
with a dipmeter. The best place to do this is on either side of the HV blocking
capacitor. The resonant frequency typically varies inversely with the
amplifier's power capability. 700W single-tube amplifiers typically resonate
from 100MHz to 150MHz. 1500W amplifiers typically resonate from 80MHz to 140MHz.
100kW amplifiers typically resonate from 35MHz to 45MHz. You should be able to
tune the resonance a few MHz by adjusting the tune capacitor. The VHF dip in
some amplifiers' anode-circuits is so sharp that it will "suck-out" the
oscillator in the dipmeter. If that is the case, the dipmeter must be backed
away (decoupled) from the conductor to accurately observe the dip. A broad,
smooth dip is good. A sharp dip indicates that the anode-circuit has a high
VHF-Q. This is not good news unless you happen to need a VHF oscillator.
If the suppressor design is changed to (hopefully) lower the VHF Q, check the
dip again. The frequency will usually not change appreciably but the dip should
now be smoother, more broad, and it should be necessary to couple the dipmeter
coil closer to the anode-circuit to achieve the same degree of dip.
If you make an experimental change to a parasitic suppressor, and you want to
evaluate the change more precisely, use a plastic ruler to measure the distance
from the anode circuit to the tip of the dipmeter coil that will result in a 20%
dip on the dipmeter. If the coupling distance required for the 20% dip decreases
after a change to a suppressor, a lower VHF-Q is indicated and the change was
obviously an improvement. If the dip distance increases, the VHF-Q went up and
the change in the suppressor was a step backward.
Amplifiers that exhibit a tuning vagary in the area of resonance on one or two
bands are probably in need of better parasitic suppression. A stable amplifier
usually exhibits smooth, symmetrical tuning.
ALC
Today, the
typical transceiver output is 100W to 200W. There are amplifier tubes that can
be destroyed by 100W of drive. A good example is the 3CX800A7. Driving a
3CX800A7 with 100W PEP will eventually strip flakes off of the cathode. The
flakes lodge between the cathode and the grid cage--creating a fatal short. Even
a pair of 3CX800A7s are clearly over-driven by 100W. The fix is: connect a
approx. 40 ohm cathode negative-feedback resistor in series with each 3CX800A7
cathode. As a result, the 3CX800A7s won't be driven above their maximum
ratings--and into non-linearity--by a 100W transceiver. Naturally, when cathode
negative-feedback resistors are added, the cathode driving impedance increases.
The driving impedance for a pair of 3CX800A7s is about 25 ohms. With 40 ohm
cathode resistors, the driving impedance is roughly 50 ohms.
Cathode negative-feedback resistors are better than having a matched pair of
3CX800A7s. The cathode currents automatically equalize themselves---and unlike
ALC circuits, cathode feedback resistors work instantaneously--eliminating ALC's
generic flaw--leading edge splatter on SSB. Amplifier-to-transceiver ALC works
properly only on constant signal level modes such as RTTY and FM.
The 3-500Z is rated at approx. 60 watts drive. When a single 3-500Z is driven by
100W, it "flat-tops" and produces distortion. A 25 ohm low-L cathode feedback
resistor will make a 3-500Z linear with 100W of drive. The resistor is placed in
series with the cathode RF coupling capacitor.
Capacitors that carry RF current are subject to two types of internal heating. Like ripple filter capacitors, the ESR in the capacitor's conductors generates heat that is proportional to I^2 x R. Due to skin-effect, R goes up with frequency. Another source of heat is RF dielectric loss. Since dielectric loss usually varies with frequency, the current carrying ability of capacitors changes with frequency. Typically, transmitting capacitors are current rated at three widely-spaced frequencies. It's a good idea to check the manufacturer's current ratings before using a transmitting capacitor in a specific application. Just because a capacitor is a transmitting-type does not mean that it will work reliably in all RF applications.
Most of the tuned input and tuned output circuits in HF amplifiers are pi-networks. There are a number of ways to define the Q of a pi-network. In what follows, Q is defined as the input impedance of the pi-network divided by the reactance of the input, shunt element--typically a capacitor. This definition of Q is the one used by Eimac® in Care and Feeding of Power Grid Tubes.
Even though
grounded-grid amplifier circuits look simple, they are not. The grounded-grid
amplifier's tuned input circuit is in series with and out of phase with the
anode current pulses. The RF cathode current's approx. half sine wave pulses are
the sum of the anode and grid currents. Since the driver is connected to the
other end of the tuned input, some of the RF cathode current finds its way back
to the driver. Consequently the driver interacts with the amplifier. The Q of
the amplifier's tuned input affects this interaction.
Modern solid-state output MF/HF transceivers use a broadband push-pull RF output
stage. In order to meet FCC requirements, Butterworth and/or Chebyshev pass band
filters are used to suppress spurious emissions. Such filters introduce
inductive reactance or capacitive reactance within their pass bands. In other
words, the output impedance of a modern transceiver is seldom 50 ±j0 ohms. When
driving a tuned input in a grounded-grid amplifier, filter reactance interacts
with the input reactance in the tuned input. The length of the coax between the
driver and the tuned input affects the interaction.
When tube manufacturers state the cathode driving impedance in grounded-grid
operation, they are talking about an average value. The instantaneous driving
impedance fluctuates wildly during the sine wave input signal. During most of
the positive half of the input cycle, the grounded-grid looks negative with
respect to the cathode--so the flow of current is cut-off. Since virtually no
current flows, the driving impedance is extremely high.
During the negative swing in the input cycle, the grounded-grid is relatively
positive. A positive grid accelerates electrons away from the cathode, producing
high anode-current and grid-current. Due to the large flow of current, the
input-impedance is low during the negative half of the input cycle.
Consider a pair of 3-500Zs. When the driving voltage is peaking at negative
117v, the anode-current is at its peak, and the instantaneous anode-voltage is
at its lowest point--about +250v. At this instant, the total, peak
cathode-current is 3.4a. Thus, the instantaneous cathode driving impedance is
117v/3.4a = 34.5 ohm--and the peak driving power = 117v x 3.4a = 397W.
In other words, the instantaneous driving impedance swing is from near-infinite
all the way down to 34.5 ohms. The instantaneous drive power requirement varies
from 0w at the positive peak to 397w at the negative peak of the input sine
wave. Thus, the input pi-network's job is to act as a flywheel/energy storage
system and a matching transformer. That's why a simple broadband transformer can
not adequately do the job of matching the driver impedance to the cathode
impedance in a grounded-grid amplifier.
The Q of a tuned circuit is like the mass of a flywheel. More Q makes for a
better flywheel--which does a better job of averaging the wild swings in
input-Z--thereby producing a lower input-SWR. The trade-off is that more Q means
less bandwidth. With a high Q, the input SWR may be near-perfect at the center
of the band, but unacceptable at the band edges. Thus, a compromise is in order.
Eimac® typically recommends using a pi input network Q of 2 for Class AB2
grounded-grid operation. To arrive at a Q of 2, the reactance [X] of the input
capacitor, C1, is minus j50 ohm÷2=minus j25 ohm. Using C=1÷[25(2f)],
approximately 220pF of input capacitance is needed for a Q of 2 on the 10m band.
In actual practice, however, 220pF may be far from the value that produces a
satisfactory SWR with a particular model transceiver and a particular length of
coax. It may be possible to find a length of coax that would ameliorate this
problem on 10m--but there are eight other bands to contend with below 30MHz.
Since band switching different lengths of coax is hardly practicable, it would
be useful if the input capacitors were adjustable in a grounded-grid amplifier's
tuned input circuits. Adjustable coils are also useful.
When the Q of
the output pi-network tank circuit is low, two problems can occur. The harmonic
attenuation may be inadequate to meet FCC requirements--and the load impedance
matching range decreases. In other words, when Q is low, the tank circuit may be
incapable of matching even a 50 ohm load. When the Q of the tank is too high,
efficiency decreases due to the increase in I^2 R circulating current losses. A
compromise is in order. A Q of 10 is about minimum. A Q of 20 may cause
excessive tank component heating due to high circulating current. A Q of 12 to
15 is a fair compromise.
Better tank performance can be achieved by using a pi-L tank circuit. When
compared to a simple pi, the pi-L has roughly 15db better harmonic attenuation
and it typically has a wider matching range. The trade-offs are that the pi-L
requires an extra switch section and a tapped inductor.
As frequency
increases, progressively less current flows inside a wire--so current
progressively concentrates on the surface. Since a steadily decreasing part of
the conductor is being used, resistance increases as frequency increases. For
example, a 12 gauge (copper) wire will carry 20A at 60Hz with very little
heating. At 30MHz, the RF current carrying ability of 12 gauge wire is about 5A.
Band switch contact current ratings need to be similarly de-rated as frequency
increases. Paralleling contacts is a good way of increasing the current handling
ability of a band switch. Directing a portion of an amplifier's cooling air flow
at the band switch improves the RF current handling ability of band switch
contacts.
HF tank inductors can become quite lossy unless the conductor surface area
varies in proportion to frequency. Inadequate tank conductor size is the main
reason for decreasing amplifier efficiency at the higher frequencies. A tank
inductor made from 14 gauge wire is usually more than adequate for efficient
1.8MHz operation at 1500W PEP. For efficient operation at 29MHz, approx.10 mm
o.d. copper tubing (or copper strap with an equivalent surface area) is
appropriate. However, due to normal QSB--at the receiving end, even a one-third
decrease in transmit power is virtually undetectable. Thus, squeezing out the
last percentage of efficiency on 10m is not very important.
Calculating the RF circulating current in a tank inductor is fairly complex. A
quick approximation is to multiply the maximum anode-current by Q. For example,
if the anode-current is 1.2A and the tank Q is 15, the RF circulating current in
the tank will be 1.2*15 =18A. At 29MHz, 18A is a formidable amount of current.
Compared to
copper, silver [Ag] is cosmetically more attractive and more immune to
oxidation. However, silver does not make an amplifier measurably more efficient
at frequencies below about 100MHz. Copper oxidation can be prevented by
polishing copper with extra fine steel wool and applying clear, gloss,
polyurethane varnish.
Silver is useful as a component of solder. 95% tin [Sn], 5% silver, solder has a
melting temperature of 221 degrees-C/430 degrees-F. Compared to tin-lead
electronics solder, 95/5 Sn/Ag solder is about 3.5 times stronger and it has
better wetability--especially on hard-to-solder materials. 95/5 Sn/Ag solder is
ideal for soldering tank components, band switches, surface-mount solid state
devices, loose vacuum tube pins, and low Q parasitic suppressors. When
resoldering a tin-lead solder joint with tin-silver solder, first remove as much
of the tin-lead solder as possible.
The basic
requirements are: 1. The choke must have ample reactance at the lowest operating
frequency to limit the RF current through the choke to a reasonable amount. 2.
The choke can not be self-resonant near an operating frequency. 3. The wire
gauge used must be able to carry the DC anode current plus the RF current at the
lowest operating frequency without excessive heating.
If the HV-RFC has a self-resonance on or near an operating frequency, potentials
of many times the anode supply voltage can appear on the choke. When this
occurs, a choke arc and fire is likely. Choke fires can destroy more than just
the choke because the rising plume of ionized gasses from the choke fire often
creates a conduction path to the ceiling of the RF output compartment. If an arc
occurs, pervasive damage is likely if no glitch protection resistor was used in
the HV positive circuit.
Materials:
There are two types of wire insulation materials that are suitable for use in
HV-RFCs--silicone varnish and Teflon. Modern, high-temperature electric motor
wire is insulated with a tough, silicone varnish that can handle high DC voltage
and high RF voltage. At room temperature, a twisted pair of #20 silicone
varnished wires can withstand more than 5000VDC or 1500W in a 50 ohm circuit at
29MHz. This type of wire is sold by the pound in electric motor rewinding shops.
If you want to buy some, bring your own empty spools and winding device--such as
a variable-speed electric drill, with a homemade adapter to hold the spool. Due
to its toughness, silicone varnish insulation requires a special method of
stripping. An open flame from a butane lighter causes the silicone varnish to
decompose and combust. The remaining ash residue can be removed from the copper
with steel wool.
Teflon insulated magnet wire is not common. Although ordinary Teflon insulated
hookup wire may be used, the extra insulation thickness requires that a longer
coil form be used. One potential trade-off with Teflon insulated wire is
phosgene. When Teflon burns, deadly phosgene [COCl2] gas is produced.
Due to contact with air, the current carrying ability of either type of wire is
much higher in an HV-RFC than it would be in a transformer. #28 wire will easily
carry 1A in a HV-RFC. #24 will carry several amperes with acceptable heating.
G10 or G11 epoxy-fiberglass tubing is RF-resistant, strong, and easy to work
with. It is an ideal material for building HV RF chokes. It can be obtained from
plastic supply houses. 1mm wall thickness is more than adequate. Diameters of 16
to 25 mm are typically used for building HV-RFCs. G10 tubing can be cemented to
a G10 base plate with silicone rubber adhesive or epoxy. A source of G10 tubing:
Plastifab,1425 Palomares, La Verne, CA 91750 818 967 9376.
It is probably a good idea to limit RF current in the HV-RFC to no more than 1 ampere. To calculate current in the choke, take roughtly 2/3 of the anode supply volts and divide it by the reactance in ohms at the lowest operating frequency -- a.k.a. Ohm's Law.
Power supply
components can be damaged by RF. Electrolytic filter capacitors are especially
at risk. Thus, adequate RF bypassing on the power supply side of the HV-RFC is
needed. Probably no more than 10V of RF should be allowed to appear on the +HV
supply at the lowest operating frequency. Determining just how much bypass C is
needed basically involves using ohm's Law. The amount of RF current flowing
through the choke and the amount of bypass C need to be evaluated for the lowest
operating frequency--usually 1.8MHz. For example, if the reactance of the choke
is +j2000 ohms, and the AC anode voltage is 2000Vrms, then I=2000V/2000 ohm=1A
of RF flows through the choke. In order to limit the RF voltage to 10V maximum
at 1.8MHz, 10V/1A=10 ohm of capacitive reactance is needed for an adequate
bypass. Using C=1/(Xc * 2pi * f), this equates to a HV bypass capacitance of
8842pF. Obviously, a typical 1000pF bypass C [minus j88 ohm] is not going to do
the job because it would allow approx. 88V of RF to appear across the HV supply
if 1A were flowing through the choke.
500pF 20kV TV-type doorknob capacitors are NOT designed to handle RF current--so
they do not make satisfactory HV bypass capacitors. Disk ceramic capacitors may
be used for HV bypassing. Disk ceramic capacitors are somewhat limited in the
amount of RF current they can safely handle. Manufacturers typically don't
publish RF current ratings for them. To find out how different capacitors react
to RF current, you must test them yourself. Even a 7500WVDC, 2500pF disk ceramic
capacitor becomes warm from 1A at 1.8MHz. Thus, it is often best to parallel a
number of individual bypass capacitors so that the RF current will be shared
among them.
At the lowest
operating frequency, the HV-RFC should have enough reactance to limit the RF
circulating current through the choke to a reasonable amount. Allowing a RF
current of 1A RMS through the choke usually does not create problems for the
wire-lead disc-ceramic capacitors that are typically used to bypass RF on the
power supply side of the HV-RFC. To minimize RF current through the choke, it
would seem that more inductance is the answer. However, more inductance means
more choke resonances and a greater likelihood of choke fires. A compromise is
indicated.
Over the years, various schemes have been used to minimize choke resonances.
Adding gaps at presumably esoteric positions in the winding was represented as a
means of decoupling parts of the choke winding--allegedly ameliorating the
self-resonance problem. However, when the resonances of gapped chokes are
compared to similar chokes without gaps, no real improvement is observed on a
dipmeter. This should not be surprising. Optimum decoupling between two coils
occurs when they are mounted at a right angle. Adding end-to-end spacing with
gaps is the least effective decoupling method possible. To minimize resonance
problems, instead of using a single large choke, use two smaller chokes mounted
at right angles.
The highest-L choke that can built that is free of self-resonances in the HF
spectrum is roughly 60µH. At 1.8MHz, 60µH has a reactance of about +j679 ohm.
The RMS voltage that appears across an amplifier's HV-RFC is approximately
two-thirds of the anode supply voltage. For example, an amplifier that is
powered by a 3000V supply subjects its HV-RFC to about 2000V RMS. If a 60µH
inductor was used in this amplifier, at 1.8MHz the RF current through the choke
would be 2000V/679 ohm=2.95A RMS. Adequately bypassing approx. 3A of current on
the power supply side of the choke is difficult. A typical HV disk ceramic
bypass capacitor can handle only about 1A. Another problem is that at 1.8MHz
130pF [minus j679 ohm] of extra capacitance is required from the tune capacitor
to cancel the +679 ohms of reactance in the choke. Adequately bypassing 3A at
1.8MHz requires a substantial amount of capacitance. To hold the voltage across
the bypass capacitors to less than 10V at 1.8MHz, roughly 0.026µF [minus j3.3
ohm] is indicated. To handle this amount of current, four approx. 0.0075µF HV
disc ceramic capacitors would probably be needed. All things considered, using
more inductance is indicated. Limiting the HV-RFC's RF current to a maximum of
!A would make the task of bypassing a lot easier. However, increasing the
inductance above 60µH is virtually certain to move choke resonances into the HF
range. Unless these resonances are prudently parked between operating
frequencies, a choke fire may result.
To realistically evaluate the self-resonance situation, HV-RFCs should be
checked with a dipmeter after they are installed and wired in the amplifier. If
a self-resonance is within about 5% of an operating frequency, there may be a
problem. When re-parking resonances, it is usually best to remove turns from the
choke. This will move the resonances up in frequency--and only slightly increase
the maximum RF current through the choke.
In continuous coverage amplifiers, there are obviously no safe parking places
for choke resonances. The only solution is to switch HV-RFCs with one or more HV
vacuum relays.
HV-RFCs should be single-layer solenoid wound. To minimize wire vibration during
operation, the wire should be under constant tension when winding and soldering
the ends to the solder lugs. When silicone varnish insulated wire is used to
wind a HV-RFC, the finished winding should be given a coat of gloss urethane
varnish to hold the wire in place. Since varnish will not adhere to Teflon wire,
a different method is needed to keep a Teflon winding taught. Small tensioning
springs are soldered to the ends of the wire. The springs provide constant pull
to minimize wire vibration during modulation. An S-shaped copper foil jumper
should be connected across each tensioning spring.
Blocking high
voltage DC is the least difficult part of the blocking capacitor's job. During
operation on 10m, the DC blocking capacitor must be able to carry most of the RF
circulating current in the tank. Here's why: The amplifier tube's anode
capacitance normally provides most of the tune capacitance during 10m operation.
Thus, a major portion of the tank circulating current passes through the anode
capacitance and therefore through the DC blocking capacitor. In an amateur radio
amplifier, blocking capacitor currents of 5 to 10 A RMS are not uncommon during
operation on the 10m band.
Selection of a blocking capacitor should not be guesswork. It is advisable to
select a capacitor or capacitors that is rated to carry the calculated maximum
RF current present. Merely selecting an RF-type (transmitting) capacitor is not
good enough. Some RF-type capacitors have rather unspectacular current
capabilities. The capacitance of the DC blocking capacitor is not very critical.
1000pF seems to be more than adequate for operation at 1.8MHz. 88 ohms of Xc is
relatively insigificant in comparison to the typical 1000 to 2000 ohm anode
output Z.
Vacuum capacitors and vacuum relays are ideal for use in high power RF amplifiers because they can withstand high RF voltages. vacuum capacitors are able to handle more RF current than any other type of capacitor. There are some trade-offs. Vacuum components depend on their glass-to-metal or ceramic-to-metal seals to maintain their near-perfect vacuum. If a seal leaks, air molecules enter and the vacuum component is kaput. Vacuum component seals should not be subjected to unnecessary mechanical stress.
Although vacuum
capacitors can be mounted in any position, vertical mounting places the least
stress on the soft copper plates. Vertical mounting also makes the most
efficient use of chassis space. With vertical mounting, a right-angle drive is
used to bring the 1/4" diameter tuning shaft to the front panel. Cardwell-Multronics®
makes a compact right-angle drive mechanism that is ideal for this application.
It is designed to replace the shaft-cap on a vacuum capacitor's tuning shaft.
The vacuum capacitor should be set for minimum C before the drive shaft cap's
setscrews are loosened.
A vacuum capacitor should not be used as a standoff-insulator to support heavy
components. High G force can be fatal to a vacuum capacitor. The danger is not
necessarily breakage or damage to the seals. The plates in a vacuum capacitor
consist of a series of concentric, intermeshing, soft copper cylinders that
almost touch each other. A vacuum capacitor can be shorted by an inertia force
that is capable of bending the soft copper plates.
To avoid
stressing the seals, connections to the contact terminals of vacuum relays
should be made with soft copper ribbon.
The molded-in coil terminals on vacuum relays are easily broken. Connections to
the coil terminals should be made with approx. 24 gauge stranded hookup wire.
Vacuum relays generate sharp mechanical vibrations when they switch. If one is
mounted securely to the chassis, the chassis acts like a speaker cone--coupling
the vibrations more efficiently to the air. One way of overcoming this problem
is to mount the vacuum relay on small beads of silicone rubber. To accomplish
this, drill a approx. 3mm oversize mounting hole in the chassis. Use temporary
L-shaped poster board spacers to prevent the relay from touching the chassis.
After cleaning the surfaces with acetone, apply three small beads of silicone
rubber between the relay mounting flange and the chassis. Allow the silicone
rubber to cure for 2 days. Remove the spacers. The relay should float quietly on
silicone rubber shock absorbers. The vacuum relay's body should be grounded to
the chassis with thin copper ribbon. The ribbon may be soldered to the edge of
the relay flange. To avoid overheating the seals, use a large soldering
iron--and tarry ye not.
All relay coils
have inductance. Since inductance delays a change in the flow of current,
coil-inductance tends to increase the make-time of relays. Make-time is an
important design consideration when using vacuum relays for RF switching. RF-rated
vacuum relays use copper contacts to obtain high conductivity. However, copper
is vulnerable to damage from hot-switching. For example, if an amplifier's RF
output relay contacts are not closed and finished bouncing before the RF
arrives, arcing and contact damage is likely.
Make-time can be decreased by supplying extra voltage to the coil during
start-up with what is commonly called a speed-up circuit. Jennings® and Kilovac®
recommend using them to accelerate relay closure. A speed-up circuit consists of
a resistor in series with the relay's coil and a power supply that supplies two
to three times the rated coil voltage. At turn-on, the extra voltage hastens the
flow of current in the coil. The resistor limits the steady-state coil voltage
to a safe value after the flow of current builds up in the coil.
DC relay coils are usually paralleled with a diode to absorb the reverse voltage
spike that results when current stops flowing through the coil. If no reverse
diode is used, the reverse voltage spike can exceed 20 times the rated coil
voltage. The break-time of a DC relay can be controlled by adding a resistor in
series with the diode. As R increases, the break-time decreases. R should
probably not exceed three times the coil resistance.
When a vacuum seal leaks air, the breakdown voltage decreases. This problem is easy to spot in a glass-body vacuum relay--because when electrons flow through air, blue-purple photons are emitted. With glass-body vacuum capacitors, this problem is not as obvious. In a leaky glass-body vacuum capacitor, internal ionization/arcing is often not visible since the problem usually occurs deep inside the meshed concentric plates.
It is a good idea to test all vacuum components, whether they be new or used,before constructing the amplifier.
When a vacuum
component in an amplifier becomes gassy, arcing typically occurs near the crest
of the RF sine wave--so a bad vacuum component typically reduces the peak power
output. Since many amplifiers use more than one vacuum component, finding the
bad one is difficult without individual evaluation using a breakdown tester.
Random replacement--a.k.a. "Easter-egging"--is not an efficient way to repair an
amplifier that uses vacuum components. For instance--if an amplifier's RF output
vacuum relay becomes gassy, it is virtually certain to divert high power RF into
the (usually more delicate) RF input relay. If a thusly-damaged RF input relay
is replaced, the new RF input relay may also be damaged by the gassy RF output
relay. Thus, it is desirable to be able to individually test vacuum components
with a breakdown tester.
Testing the quality of a vacuum is similar to testing the breakdown voltage of a
diode. Connect a approx. 100M ohm HV resistor and a approx. 20 microampere meter
in series with a breakdown tester. Increase the voltage until about 1 to 2 µA of
leakage is detected. This voltage is the breakdown voltage. The peak RF working
voltage of a vacuum component is roughly 60% of the DC breakdown voltage.
There are
basically two types of vacuum relays--those that are designed for hot-switching,
and those that are not. Hot-switching-capable relays have tungsten contacts.
Such relays are intended for use primarily in power supplies. Relays that are
designed for RF have copper contacts. They should never be allowed to
hot-switch. Copper-contact relays have approximately one-third the contact
resistance that similar tungsten-contact relays have. For instance, the Jennings
RJ-1A is the copper-contact version of the tungsten-contact RJ-1H. The rated
contact resistance of the RJ-1H is 30 milli-ohms. The rated contact resistance
of the RJ-1A is 10 milli-ohms. Tungsten contact relays are not rated for current
RF current. However, they should work fine for RF service if they are operated
at roughly two-thirds of the RF current rating for their copper-contact
counterparts. Tungsten contacts are extremely hard. They are capable of more
operations than copper contacts. For heavy , full break-in telegraphy use,
tungsten contacts are preferable--even though they do not have the continuous RF
current handling capability of copper contacts.
With vacuum relays, contact failure is not uncommon. Contacts suffer from
contact erosion. This condition increases contact resistance. Eventually, an
eroding contact will open completely. To test a vacuum relay, the resistance of
normally open [NO] contacts and the resistance of normally closed [NC] contacts
should be measured and compared with the manufacturer's specifications. Ordinary
ohm-meters are not suitable for detecting contact problems other than an open
circuit. The voltage drop across relay contacts should be measured with a
substantial current flowing. 1A is a reasonable current to use. Measure the mV
drop directly across the contact terminals using a DMM with test prod leads.
Most of the vacuum relays that are designed to handle RF current have a rated
contact resistance of less than15 milli ohm--so no more than 15 milli V should
appear across the terminals with 1A flowing through the contacts.
Vacuum capacitors store energy efficiently because they have virtually zero ESR [equivalent series resistance] and internal L--thus, the peak discharge current can be astronomical. When the breakdown test voltage is high enough to create more than a few microamperes of leakage, a vacuum capacitor will normally self-discharge--producing a clearly audible tick due to the large peak discharge current and commensurately large electromagnetic force. After a vacuum capacitor self-discharges, it begins charging and the process repeats. A vacuum capacitor should not be allowed to self-discharge more than a few times unless the capacitor has been in storage for many years. During long-term storage, for some as yet unexplained reason, copper atoms tend to line up, forming whiskers on the surface of the plates. Copper whiskers initially reduce the breakdown voltage. Copper whiskers can be dislodged by self-discharge. If the breakdown voltage increases after a self-discharge, another self-discharge may be beneficial. Repeated self-discharge will cause a decrease in breakdown voltage.
Linear
amplifiers are like induction motors--they are designed to run fully-loaded. If
your grounded-grid amplifier's instruction book says to reduce drive power
during tune-up--and most of them do--it is not giving you correct information.
In order to be linear, amplifiers must be tuned-up with the same peak
drive-power level that they will be driven with during actual operation.
Reducing drive power changes the output Z of the amplifying device to something
other than the tank circuit was designed to match. Thus, the tune and load
settings with low drive will be wrong when normal drive is applied.
Tune-up method #1: Set the amplifier's HV supply to the CW-Tune/low-V-tap. If
you are not sure where to preset the Load control, set it to >70% of maximum
loading [30% of C] to be safe. Apply the drive level that you intend to drive
the amplifier with during actual use. Alternately adjust the amplifier's Tune
and Load controls for maximum relative power output. The whole process should
take less than 6 seconds. It may sound brutal, but this tune-up method results
in good amplifier linearity and it won't damage the tubes if the maximum
anode-current rating is not exceeded. If the anode-current is excessive, the
resistance of the cathode, RF negative-feedback resistor needs to be increased
slightly--or the PEP adjust control in the transceiver needs to be turned down.
Tune-up method #2: [not for FM, AØ, and RTTY operation] To reduce the stress on
an amplifier during tune-up, use a reduced duty-cycle driving signal. This can
be accomplished by keying the transceiver, on CW mode, with a CW keyer, set to
send approx. 50wpm dits. CW dits have a 1/2-on, 1/2-off, or 50% duty-cycle.
Using this method, the amplifier may be tuned-up, again for maximum power
output, in its higher-voltage, SSB-mode. Keyers that have a weighting adjustment
can be set to produce a light dit that has a duty cycle of about 30% instead of
the normal 50%. Another device for reducing the duty-cycle during tune-up is a
tuning-pulser.
If you want to operate with reduced power during good band conditions, first
tune up your amplifier with normal drive power, then turn the microphone gain
down to reduce power.
Tetrodes and pentodes require a peak RF drive voltage that semi-matches up to the grid bias voltage.
My strategy is to choose a value of grid termination resistance that roughly provides the needed peak RF drive V to the grid with exciters that develop 100v-p (100w rms) to 141v-p (200w rms) across 50 ohms. In other words, the goal is to match peak RF drive volts with the needed grid bias volts from the tetrode/pentode manufacturer's technical specifications.
-- For tubes that require 50v to 70v of grid bias, like the 4CX800A and 4CX1000A, a voltage-halving bifilar stepdown transformer driven 12.5 ohm grid termination is used.
-- For tubes that require 100v to 140v of grid bias, like the 4CX1500A, a directly-driven 50 ohm grid termination would be used.
-- For tubes that require 200v to 280v of grid bias, like the 4-1000A and 8169, a voltage-doubling bifilar stepup-transformer driven, 200 ohm grid termination is used.
-- For tubes that require greater than 300v grid bias, like the 8171, a trifilar stepup-transformer driven, 450 ohm grid termination is used.
However, if the peak grid V with max. drive is still a bit much, a cathode RF negative feedback reisistor (Rk) can be added to make up the difference. However, a trade-off is that the peak V drop across Rk normally subtracts from the screen to cathode V at the critical anode current peak. A workaround is to use the circuit shown in Figure 10.
Class AB1
grid-driven amplifiers look more complex than Class AB2 grounded-grid
amplifiers. However, the tuned input circuitry for multi band Class AB1
grid-driven operation is comparatively simple.
The grid capacitance of tubes that are commonly used in Class AB1 grid-driven
amateur radio power amplifier service ranges from about 15pF to 130pF. Since the
capacitance of the grid is in parallel with the input, as frequency increases,
input SWR worsens. This problem can be corrected by connecting a variable
inductor in parallel with the grid. The inductive reactance {+j ohms} of the
inductor is adjusted to cancel the capacitive reactance {minus j ohms} of the
grid--thereby resonating the grid at the operating frequency. When the input SWR
is tuned to minimum, the grid circuit is resonant. A simplified diagram is
provided.
If the other end of the variable inductor is connected to a properly-adjusted
capacitive voltage divider (connected between the anode and chassis ground), the
amplifier is neutralized at whatever frequency the grid is tuned to. Obviously,
this type of Class AB1 input circuit is a natural for continuous HF and MF
coverage--just what's needed for operation on the 9 amateur bands below 30MHz.
The ratio of the capacitances in the capacitive voltage divider equals the ratio
of the feedback capacitance (the anode to grid capacitance) divided by the grid
input capacitance. Typical ratios are 150 to 1 ... Achieving wide frequency
coverage is not as easy in Class AB2 grounded-grid operation. A pi-network tuned
input with the recommended Q of 2 has a limited bandwidth--so many, switched,
tuned input circuits are required for wide frequency coverage.
Screen and Grid Supplies
There are many
tetrodes and pentodes to choose from that are satisfactory for Class AB1
grid-driven operation. The essential criteria is that, with zero grid volts, the
tube is capable of a peak anode-current that is at least triple its maximum
(average) current rating. In most cases, this condition can only be met if
near-maximum screen-voltage is applied. Relatively high screen-voltage is
important because peak anode-current is a function of the screen-voltage raised
to the 1.5 power.
For the best linearity, screen voltage should be regulated. For smaller tetrodes
and pentodes, a Zener diode shunt regulator offers a good solution. Typically, a
series of 10v to 30v, 5W Zeners are used. Screen voltage is adjusted by shorting
out Zener diodes with a rotary switch. For larger tubes, an adjustable
series-regulator is the best way to supply voltage to the screen. Thanks to
modern power FETs and the venerable 723 IC linear regulator, building a
reliable, regulated supply of 2kV or less is fairly simple.
Since the grid does not pass current in Class AB1 operation, there is no
necessity to regulate the bias voltage. However, the bias supply should not have
an extremely high output impedance. A maximum grid circuit R of 1k to 100k ohms
is typically recommended by tube manufacturers.
'Work-space' and
'head room' are terms that describe the range in which instantaneous
anode-voltage is free to move up and down--thereby performing work. In a tetrode,
at the maximum peak anode-current, to avoid excessive screen-current and a
decrease in linearity, the instantaneous anode-voltage should not dip much below
the screen-voltage. For example, a tetrode with a 4kV anode supply and an 700V
screen supply, the work-space is approximately 4000V minus 600V = 3400V peak
In a pentode, the instantaneous anode-voltage may dip close to the
suppressor-voltage--which is typically zero volts. In the above example with a
screen-voltage of 800V, if the tube happened to be a pentode, the work-space
would be around 3750V peak. Thus, pentodes enjoy slightly more work-space than
tetrodes. As a result, pentodes are slightly more efficient than tetrodes.
However, pentodes are more expensive than tetrodes because they are more complex
to build. Sockets with low-L suppressor and screen bypass capacitors are needed
for stable operation. Pentode sockets are not inexpensive. Another
trade-off is that there are relatively few types of pentodes to choose from. A
(if not the) suitable pentode for amateur radio Class AB1 grid-driven service is
the 5CX1500.
Pentodes typically have less feedback capacitance than tetrodes. This advantage theoretically makes pentodes more stable. Some designers do not neutralize pentodes because they feel the relatively low feedback capacitance between the anode and the grid is insignificant. However, for optimum linearity and stability, plus low input SWR, a pentode should be neutralized. This can easily be accomplished with the grid input circuit diagram [Figure 5] for Class AB1 tetrodes. To use this circuit with a pentode, DC-connect the suppressor to the cathode with a l0 or so ohm resistor. However, the suppressor must always be RF-bypassed to chassis ground to decrease feedback from anode to grid.
Every screen
type tube has a maximum screen dissipation rating in watts. If screen current
times screen voltage exceeds this rating, the tube could be destroyed. This can
easily happen with a no load or light load condition--so various protection
schemes are used. If the anode voltage disappears while screen voltage is
present, screen current will be excessive unless a means of protection is
provided. Another hazard is reverse screen current. Reverse screen current can
easily become a runaway condition. It happens virtually instantaneously. Reverse
screen current is commonly experienced in Class AB1 operation. Unless bled off
into a resistor load or into a shunt Zener voltage-regulator, reverse screen
current can quickly destroy a tube. For tubes with screen voltages in the 300V
to 800V range, a shunt regulator using a Zener diode string is a good solution.
The Zener regulator string is connected through a high-R resistor to the anode
supply. A sample circuit is provided. A suitable tube would be the 4CX1500B, or
similar types.
Advantages of Shunt Zener Screen Regulation:
Limits the maximum current that can be drawn by the screen.
Protects against reverse screen current.
If the high voltage disappears, so does the screen voltage.
However, for larger tubes with higher screen current and screen voltage requirements, a Zener shunt regulator is somewhat impractical. A continuously-adjustable series-regulator screen supply is a better solution. To protect against reverse screen current, a shunt resistor/bleeder must be connected across the screen supply. A bleeder current flow of roughly 20% of the normal screen current seems to be adequate. 25% might be safer. To protect against excessive forward screen current, a fast acting fuse or magnetic-type circuit breaker is incorporated in the primary of the screen supply power transformer. An adjustable series regulator circuit is provided.
ßAdjusting
a Class AB1 amplifier may look complicated at first, but after you have done it
a few times, and you begin to understand the reason behind each step, it gets
easier.
Neutralization: The goal of neutralization is to isolate the anode from
the grid at the operating frequency. Neutralization discourages
regeneration--oscillation. Neutralization usually needs to be adjusted only
once.
1. Disconnect the amplifier from the electric-mains.
2. Temporarily disconnect the tank circuit from the HV blocking-capacitor.
3. Substitute a low-L film resistor, with the same R as the design anode-load
[output] resistance, in place of the tank circuit. Typical values would be 1000
ohm to 4000 ohm, 2W. The resistor connects to the blocking-capacitor and to
chassis-ground. Connect an RF-voltmeter or an oscilloscope equipped with a 10 to
1 hign impedance probe across the resistor.
4. Connect the amplifier to the electric-mains and turn on the transmit-receive
relay power supply plus the grid and filament supplies. Do not turn on the
screen or anode supplies.
5. Drive the amplifier with 20m or 15m RF. Tune the grid-circuit
variable-inductor [L1] for minimum input SWR or minimum reflected power. If
necessary, adjust the DC grid-voltage so that virtually no grid-current flows.
6. Adjust the neutralizing-capacitor (C3) for minimum RF-voltage at the
anode-load resistor. If needed, readjust L1 for best input SWR followed by
readjustment of C3. This completes the neutralizing procedure.
After C3 is nulled, the amplifier is neutralized for all bands. To confirm this,
check the neutralization on another band. Readjust L1 for minimum SWR. The RF
voltage across the output load resistor should not change appreciably.
Typically, no further adjustment is necessary--even if the tube is replaced.
Remove the resistor and reconnect the tank circuit.
Tune-up.
1. Switch off the screen and HV anode supplies. Switch on the T/R relay supply,
the filament supply and the grid supply.
2. Transmit on CW-mode into the amplifier and adjust L1, the grid
roller-inductor, for minimum input reflected power. This tunes out the
grid-reactance and simultaneously neutralizes the amplifier at the operating
frequency. If you are using a transistor-output transceiver, to preclude SWR
shutdown, initially tune the grid with no more than 5W of signal.
3. Apply full drive power using an electronic keyer sending dits at about 50wpm,
or a use a tuning-pulser. Adjust the DC grid-voltage so that <0.1mA of
grid-current flows. The grid-voltage is adjusted so that the grid is on the
threshold of current flow. The grid-voltage adjustment is not used to set the
zero-signal anode-current [ZSAC]--also known as 'idling current' or 'resting
current'. Although the grid-voltage adjustment can discretely be used to make a
small adjustment in the ZSAC, in Class AB1 operation, the primary criteria for
setting the grid-voltage is that virtually NO grid-current flow with maximum
drive. ZSAC is set by adjusting the screen voltage. Switch on the
screen and HV supplies. Key the amplifier but do not apply drive power. Using
the screen voltage adjustment, set the ZSAC as recommended by the tube
manufacturer. For most tubes, the ZSAC should be about 20% of the rated anode
current.
4. If a variable tank inductor and a variable tune capacitor is used, preset the
tune capacitor and the tank inductor for the desired operating Q on the band in
use. Preset the load capacitor and inductor by calculation. It is best to error
on the side of too-little load C [heavy loading]. If too-light loading [too much
load C] is used, excessive screen-current is likely. Remember that the tune C
sets the operating Q. Most of the tuning should be done with the variable tank
inductor. Fine tuning can be done with the tune C--but the final setting should
no be very far from the setting for the correct operating Q.
5. When any amplifier is tuned-up, the anode-current must be driven to the
maximum, peak, design value so that the tube's output load resistance will meet
the design criteria for the pi output tank circuit. If a lesser current is used
without proportionately decreasing the supply voltage, the output load
resistance will be too-high and the subsequent adjustment of the tank will be
for the incorrect output load resistance.
To be both linear and deliver good power output, an amplifier tube must be
adjusted by loading it for the optimum peak anode-voltage swing. The indicated
screen-current is an accurate way of tuning up a tetrode or pentode. If the
anode-voltage swing is too great because of too-light loading, the
screen-current [and distortion] will increase. This means that the instantaneous
minimum anode-voltage is less than it should be--a situation which causes too
many electrons to stick to the screen--thereby depriving the anode of electrons.
If the screen-current is too low, the anode-voltage swing is inadequate--meaning
that the loading is too heavy. This condition causes lower power-output. When
the output tank circuit is tuned correctly, the screen-current meter peaks. This
is done by adjusting the tune capacitor or by adjusting the tank inductor. Do
not peak the screen current with the loading capacitor.
·· Thus, by using only the screen-current meter, tuning and loading can be
adjusted for good linearity and good power output.
6. Set the transceiver to CW-mode. Apply full drive power. To reduce stress
during tune-up, use an electronic keyer to send dits at about 50wpm. Standard
dits are a 50% duty-cycle waveform, so current meter indications are roughly
half of the actual value. A tuning pulser works even better. [Figure 9]
7. Peak the screen-current by tuning the tank inductor or the tune capacitor. If
the screen-current begins to become excessive, stop short of the peak, increase
loading and continue. If the screen-current is too low, lighter loading is
needed.
--The last step is to re-peak the screen-current with the tank inductor or the
tune capacitor.
Loading for slightly less screen-current increases linearity with the trade-off
of slightly less power output.
·· It is useful to keep a log of the various final settings for different
frequencies. This saves time during future tune-ups.
For thoriated-tungsten cathode tubes only: - - While sending dits at full power,
gradually reduce the filament-voltage until the relative output just begins to
decrease. Increase the filament-voltage about 2%. This is the optimum
filament-voltage. This should be rechecked every few hundred operating hours.
The same thing applies to grounded grid amplifiers. For indirectly heated
cathode tubes, like the 8877, the ideal filament voltage for communications
services is near the minimum filament voltage rating. Under no circumstance
should such a tube be operated at less than the minimum filament rating.
Perfectly linear
amplification produces nothing except a larger representation of the input
signal. Non-linear amplification produces mixing--and mixing creates distortion
products.
Inter-modulation distortion [IMD] is the result of mixing between two or more
input signals. The human voice produces many frequencies at any instant. When
voice modulation is amplified non-linearly, many mixing products are produced.
This is called "splatter" or, more descriptively, "rotten splatter." IMD is
usually measured by simultaneously applying two equal-amplitude, not
harmonically related modulation frequencies such as 2000Hz and 2200Hz. When two
or more frequencies mix they produce spurious signals at their sum and their
difference frequencies--in this case 4200Hz and 200Hz. The first level of mixing
produces what are called "third order products." Additional products are
produced by third order products mixing with the two fundamental frequencies.
For instance, 2200Hz and 4200Hz mix to produce a signal at 6400Hz.
When distortion products are inside the fundamental pass band of an AM or SSB
transmitter, audible distortion results. This gives voice modulation a rough,
unpleasant characteristic that reduces intelligibility. Odd-order distortion
products which lie outside the pass band can cause interference on adjacent
frequencies.
There are two methods of referencing IMD measurements. In method A, the IMD
power level is referenced to either one of two equal-amplitude input signals.
The power ratio of PEP to either of two equal-amplitude sine waves is four to
one [6db]. In method B, the IMD level is referenced to the PEP level. Thus, an
IMD level of minus 34db using method A equals an IMD level of minus 40db using
method B. Amateur radio operators tend to use method B because receiver S-meters
respond to PEP. In commercial radio, the military, and the FCC--where distortion
measurements are typically made with a spectrum analyzer--method A is used. When
using a spectrum analyzer, distortion can be broken down further into third
order products, fifth order products, seventh order products. However, total IMD
referenced to PEP is a more significant number.
It is possible to measure IMD without expensive laboratory equipment. All that's
needed is a receiver and some understanding of what's required to make a
meaningful measurement.
By comparing the signal strength in the transmitter's fundamental pass band
window with the signal strength in the adjacent pass band windows, IMD can be
measured fairly accurately--even over the air. The amount of receive frequency
offset is critical. If the receive pass band is too close to the transmitter's
fundamental pass band, the receiver will not be able to separate the IMD energy
from the fundamental energy. As a result of this overlap, the distortion
measurement will be higher than the actual amount. If the receive frequency
offset is too far from the fundamental pass band, the receiver's pass band will
not receive all of the IMD--and the distortion measurement will be lower than
the actual amount.
For a receiver with two, cascaded SSB filters, a receive offset of 3.6kHz is
about right--provided that the receiver is set to the same sideband as the
transmitter. For a receiver with one SSB filter, an offset of about 4.5kHz is
needed. To measure the IMD level of a LSB signal, offset LSB-receive higher in
frequency. For measuring the IMD from an USB signal, offset USB-receive lower in
frequency.
Since very few S-meters are linear, a calibration chart of S-meter readings
versus decibels is a prerequisite for making accurate measurements. A
calibration chart can be made with a step-attenuator and a signal source, or
with a signal generator/attenuator.
In order to measure IMD, at least two modulation frequencies are required. Human
speech is a good signal source for measuring IMD because, at any instant, speech
contains many fundamental frequencies and harmonics. As its name suggests,
another harmonic-rich signal source is a harmonica. By simultaneously blowing
into two or three adjacent holes at the low note end, a plethora of frequencies
can be produced that are optimal for making distortion measurements.
Before reporting
splatter, it is important to keep in mind that all SSB, DSB, and AM signals have
IMD. In other words, everybody splatters. The obvious question is how many
decibels down is the IMD? Minus 40db is excellent; minus 30db is objectionable;
minus 20db is abundantly abominable. With one exception, FCC rules allow
virtually any level of IMD inside the ham bands. The exception is when IMD
causes harmful interference to emergency communications. Splattering on
non-emergency communications is NOT considered to be harmful.
Before reporting a station's level of IMD, it is advisable to determine whether
or not the station operator is interested in hearing your report. Although most
amateur radio operators are interested in transmitting a high quality signal,
some operators deliberately misadjust their equipment to maximize IMD.
Since E-peak = E-rms x 2^0.5, and P = E^2 ÷ R, at its crest, the instantaneous peak power in a sine wave is double the RMS power. A common unit of measuring amplifier output is the PEP [peak envelope power] watt. Despite the name, peak envelope power watts are not peak watts--they are RMS watts at the crest of modulation. If an amplifier was powered by a regulated anode supply, there would be virtually no difference between PEP watts and AØ [NØN] watts. In a typical amplifier, the anode-voltage sags appreciably under AØ conditions--so PEP watts are typically about 20% higher than AØ watts. PEP need not be measured with voice modulation. PEP can also be measured by keying the driver at 30 pps with a steady string of pulses that approximates the duty-cycle of a human voice--roughly 30%.
Traditionally, amateur radio operators have taken a cavalier attitude toward tube manufacturer's ratings. While some ratings can be exceeded judiciously, exceeding other ratings can be costly. Examples of ratings which should not be exceeded for indirectly-heated cathode tubes are minimum filament-voltage and maximum anode-current. Violation of either rating can result in destruction of the delicate cathode. Directly-heated cathodes are more rugged. The maximum anode-current rating for directly-heated cathode tubes is a linearity issue--not a cathode destruction issue. One rating which should not be exceeded is maximum seal temperature. It has been said that the way to tell when the blower is too big is if it blows the tube out the socket.
Tank Q, the reactance of C1, and the optimal anode load resistance for linear operation (RL) are inter-related. Tank Q is defined as the capacitive reactance of C1 (XC1) at the frequency of operation, divided into RL ---i.e., Q = RL/XC1......and XC1= RL /Q. Note: C1 includes the anode (output) capacitance (Ca) of the amplifier tube. At 29MHz, Ca may be a sizeable fraction of C1.
RL = Esupply/2*IAn where IAn is the average anode current in amperes.
(Note: There is some variation in the constant in the denominator of the RL formula. For tubes with minimal anode-cathode potential at peak anode-current, like the 8877, a constant of 1.6 should give more accurate results. However, for tetrodes like the 8171, which use a high screen potential (reduces anode AC peak-V), a constant of 2 seems to be more accurate.
Thus, for a tube operating from 2500v @ 1A, whose anode capacitance (Ca) is 10pF:
RL = 2500v/2*1A = 1250 ohms.
Calculating C1
For a Q of 12.5, XC1=1250 ohms/12.5 = 100 ohms.
The needed tune capacitance, C1 = 1/(2*Pi*f*XC1). For 14MHz, C1 = 1/(6.28*14*106Hz*100 ohms) = 113.7pF. However, since part of C1 is comprised of Ca, the net tune C is 113.7pF -10pF = 103.7pF. At 28MHz, the tune C would be roughly: 57pF -10pF = 47pF. At 280MHz, 10pF has about 100 ohms of X, so, for a Q of 12.5, Ca furnishes 100% of C1, so no tune C can be used.
Calculating the Load Capacitance (C2) and Tank Inductance, L
--------------------------------------------------------
Improved Anode-Circuit Parasitic-Suppression For Modern Amplifier-Tubes
The traditional copper-inductor/carbon-resistor anode
[plate] parasitic-suppressor has been used in vacuum-tube amplifiers for at
least 50 years. The earliest record of an anode parasitic-suppressor that I can
locate was in a transmitter that was built in the early 1930s by the (Art)
Collins Radio Company.
(In late 1990, I was made aware of some interesting information on anode-circuit
VHF parasitic suppressors in the 1926 Edition of The Radio Amateur's Handbook.
This information was inexplicably omitted from post-1929 editions. Info provided
by Dave Newkirk, WJ1Z)
Much of the reason for Art Collins' early success can be attributed to the fact
that he, almost alone, understood that where RF is concerned there is no such
thing as a zero-potential "ground" and that any wire or strap was a
capacitor-inductor VHF tuned-circuit as well as a conductor. He understood that
an "RF-choke" acted like a short-circuit at certain frequencies and that
sometimes a resistor would make a better RF-choke than an RF-choke! Because he
understood these "RF secrets", he was the first manufacturer to build a
transmitter that: worked on all frequencies up to 14.5MHz, was stable and could
be tuned up every time with no surprises.
Anode parasitic-suppressor design has not changed during the last 50+ years
while vacuum-tube design has changed markedly. In the 1930s, 40s and 50s, a
"high-Mu triode" had a (voltage) amplification factor of 40. Today, a "high-Mu
triode" usually indicates an amplification factor of 100 to 240. A fifty+
year-old parasitic-suppressor design that was usually successful at preventing
oscillation in an amplifier-tube with an amplification of 40, may not be as
successful on a modern amplifier-tube that has much more gain.
Modern amplifier-tubes have another factor, in addition to higher voltage gain,
that makes the job of the traditional inductor/resistor VHF parasitic-suppressor
more difficult. That factor is higher frequency capability. Ancient
amplifier-tubes could barely be coaxed into amplifying at 28MHz. The 203A that
was used successfully in the Collins 150B transmitter had a full-power rating of
15MHz.
Modern Amplifier-Tube Performance:
The popular 8802/3-500Z triode has an average amplification factor of 130
(Eimac) to 200 (Amperex). The Amperex version appears to be electrically
equivalent to the 8163/3-400Z with the exception of the anode dissipation
rating. The maximum-input rating of the Eimac 3-500Z, for "radio frequency power
amplifier or oscillator service" is 110MHz. 3-500Zs work well above 110MHz if
the power is de-rated as frequency increases. Other types of modern
amplifier-tubes commonly used in HF-amplifiers have an even higher amplification
factor and a frequency rating of up to 500MHz. The 8874 is a good example of a
high gain, 500MHz triode. It has an average amplification factor of 240! This is
definitely a high-Mu triode.
Oscillators:
If an amplifier-tube can amplify at a frequency, it can usually be made to
oscillate at that frequency. This is good news for oscillator builders and bad
news for unwary amplifier builders.
In addition to frequency capability, there are some other prerequisites that
must be met before oscillation can be achieved: a feedback path between the
output and the input of the amplifier and high-"Q" resonant circuits in the
output lead and in the input lead to the amplifier-tube that are resonant near
the same frequency. The resonant circuits are essential because they act like a
flywheel and sustain the oscillation during the portion of the cycle that the
amplifier-tube is not conducting and amplifying.
The (Incomplete) Schematic Diagram:
Understanding the nature of the parasitic-oscillation problem would be much
easier if the schematic diagram of an amplifier circuit would show the
interconnecting input and output leads to the amplifier-tube for what they
actually are: inductors. These incognito inductors, combined with the
inter-electrode capacitances of the amplifier-tube, form unavoidable VHF
self-resonant circuits. See Figure 1, A and B. The typical frequency range of
these resonances is from 90MHz to 160MHz in 1500W HF amplifiers.
The Parasitic-Oscillation Seed-Voltage:
The essential question is: Where does the initial VHF voltage come from that
starts the self-resonant flywheels in motion that causes the
parasitic-oscillation to take place? Certainly, it can not come from the exciter
because all exciters have a built-in low-pass filter that is very effective at
blocking any VHF signal. This leaves only the amplifier as the source of the
seed-voltage.
The answer to that pivotal question involves Q. Q represents the "Quality" of a
tuned circuit component. More Quality should be better. An old adage says: "more
is not always better". Where amplifier design is concerned, more Q is certainly
not always better. The appropriate Q for each part of the circuit is the best
design. For example: HF tank-circuit components should have a high-Q. and, as I
will explain, anode leads should have a low-Q.
For the purpose of this discussion, the most important rule about Q is: The
RF-voltage that is developed across a resonant circuit is directly proportional
to the Q of the resonant circuit.
This principle is best illustrated by the antique spark-transmitter. In a
spark-transmitter, the transient-currents from a motor-driven rotary spark-gap
(a motorized switch) were passed through a high-Q tuned-circuit. This caused the
tuned-circuit to "ring" at its resonant frequency which produced a surprising
amount of RF voltage and power. The tuned-circuit acts like a flywheel after
each impulse. It coasts a bit after each impulse and then stops, like the
ringing of a bell. This is referred to as "flywheel-effect". Lowering the Q will
reduce the flywheel-effect.
Amplifiers are routinely subjected to numerous turn-on, switching, keying, and
voice transient-currents. These transient-currents pass through the VHF
self-resonant anode-circuit and the VHF self-resonant input-circuit. Each
transient-current causes the input and output self-resonant circuits to ring and
generate an invisible, damped-wave VHF voltage that is proportional to the VHF-Q
of these circuits This is the source of the VHF seed-voltage that initiates the
parasitic-oscillation.
Part of this seed-voltage will be fed back to the input of the amplifier by the
feedthrough/feedback capacitance inside the amplifier-tube. The VHF voltage will
then be amplified by the amplifier-tube and it will appear in the anode-circuit
where some of it will be returned to the input of the amplifier-tube by way of
the feedback-capacitance.
If the amplified VHF voltage arrives with the right phase and amplitude, an even
larger signal may be fed back to the input of the amplifier. When this occurs,
the parasitic-oscillation is off and running. This would not be a problem if the
considerable energy that is generated by the VHF parasitic-oscillation could be
safely dissipated in the load that is connected to the amplifier. Unfortunately,
the VHF energy can not reach the output connector of the amplifier because it
can not pass through the HF tank-circuit inductor. This inductor acts as an RF
choke to the VHF energy. This traps the VHF energy in the anode-circuit. With no
load, the grid-current and grid-dissipation of a high-Mu triode oscillator
becomes excessive in a matter of milliseconds. This can start a chain reaction
of events that almost simultaneously results in a loud bang and can cause severe
damage to the amplifier.
Grounded-grid Oscillators:
Making a grounded-grid amplifier oscillate is easier than it might seem: In a
grid-driven, grounded-cathode amplifier, the output and input voltages are 180
degrees out of phase. They oppose each other. Before regeneration can occur, the
output and input voltages must be made in-phase, to aid each other, by adding a
phase-shift circuit. In a grounded-grid amplifier the output and input voltages
are already in-phase and aiding each other.
For many years it was assumed that grounded-grid amplifiers were inherently
stable because the "grounded"-grid acts as a shield between the input and the
output circuits, thereby blocking regeneration and oscillation. At HF this logic
is true but at VHF, the logic is false because no matter how carefully an
amplifier-tube is designed, at some frequency the "grounded"-grid will become
self-resonant. This is due to the unavoidable, combined inductances of: the grid
structure, the internal leads, external leads, and the tube socket, resonating
with the capacitance of the grid structure. In a 3-500Z triode, the directly (as
is possible) grounded-grid will self-resonate at about 95MHz. As frequency
increases above grid self-resonance, the grid exhibits inductive reactance, and
the grid is no longer "grounded".
When the grid is not truly grounded, as is the case above its self-resonant
frequency, the assumption about the shield, that we are depending on to block
regeneration, is in serious trouble. And, to make matters worse, as the
frequency increases into the VHF region, the feedthrough capacitance from the
input [cathode] to the output [anode] of the amplifier has fewer and fewer ohms
of capacitive reactance.
In other words, As the frequency increases above the grid self-resonant
frequency , the "grounded-grid" behaves progressively less as though it were
grounded and the feedback, or regeneration, path between the input and the
output of the amplifier-tube becomes more and more conductive to RF current..
This combination is not desirable unless the designer intends to build an
oscillator.
Anti-Parasitic Techniques and Q:
Another important rule is: Q is equal to Reactance divided by Resistance, or Q=
X/R . Q can be decreased by increasing the resistance, or by decreasing the
reactance, or both.
One obvious way to lower Q is to use resistive, or low-Q, conductors.
Silver-plated copper strap has the highest VHF-Q known to science at room
temperature and yet silver-plated copper strap is commonly used for
anode-circuit wiring and for VHF "parasitic-suppressors" in HF amplifiers. A
more accurate name for a silver-plated parasitic-suppressor would be a
parasitic-supporter.
The Q of copper is about 94% of the Q of silver, so copper does not provide an
appreciable improvement in Q reduction over silver. Trying to build a low-Q
circuit with high-Q silver or copper conductors makes about as much sense as
trying to make a pencil eraser out of Teflon®.
Reducing the inductive reactance by shortening lead lengths may improve
stability IF the shortened lead places the cathode and anode-circuit
self-resonant frequencies farther apart.
Another method of improving stability is to tune out some of the inductive
reactance in the grid structure by bypassing the grid to the chassis with small
capacitors. This increases the self-resonant frequency of the grid circuit to a
point where the amplifier-tube will have less amplifying and oscillating
ability.
The first grounded-grid amplifier that I know of that used this technique used
(4) 811As and was built by the Collins Radio Company. Many currently produced
commercial grounded-grid amplifiers still use this circuit. I discussed this in
a previous article about parasitic-oscillation in grounded-grid amplifiers.
{"Grounded-Grid Amplifier Parasitics", Ham Radio Magazine, April 1986,
page 31.}
Grid-inductance cancelling capacitors are most effective when used with older
design amplifier-tubes like the 811A that have a considerable amount of internal
grid-inductance to cancel. This technique is only mildly effective at improving
amplifier stability in modern amplifier-tubes, that have inherently low
grid-inductance.
Another anti-parasitic technique that I discussed in the article was the use of
an input parasitic suppressor-resistor, to lower the VHF-Q at the self-resonant
frequency of the input (cathode) circuit. Input suppressor-resistors also reduce
intermodulation distortion (IMD) with the tradeoff of a slight increase in the
drive power requirement to the amplifier.
Input parasitic-suppressor resistors are moderately effective at stabilizing
unruly amplifiers, but they are not always 100% successful. After the article
about parasitic oscillation was published, about 5% of the follow-up letters and
phone calls I received were from people who reported that their amplifiers were
more stable with input suppressor-resistors than without, but still not
perfectly free from the foreboding signs of instability like minor arcing and
spitting at the tuning capacitor and/or bandswitch. The only area left for
improvement was the anode-circuit.
In Search Of A Better Anode Parasitic-Suppressor:
The trouble with trying to troubleshoot a parasitic-oscillation problem is that
the crazy things are not always predictable. It may be that just the right
transient or rapid sequence of transients needs to come along to get the ball
rolling. For example, you can change something like a conductor-length in a
marginally stable amplifier and it will behave for months. When you are
beginning to believe that the problem is "fixed", and you confidently put the
rest of the screws in the cabinet, it will unexpectedly arc or burn-up the
parasitic-suppressor resistor, or worse.
The perfect amplifier to experiment with would be one that had an unusually high
gain amplifier-tube or tubes that consistently exhibited instability problems
even with input suppressors installed. By a great stroke of good luck, just such
an amplifier came into the possession of NF7S [Ed], who lives in Phoenix,
Arizona. From Ed's point of view it was initially a stroke of bad luck.
The amplifier was a newly purchased model which uses a pair of 3-500Zs with
either 2200V (CW) or 3200V (SSB) on the anodes. The new amplifier made an arcing
sound, but he was not concerned since, on page 14, the instruction manual said
that this arcing sound was "normal". After a few months the "normal arcing" had
burned off some of the contacts on the output section of the bandswitch. The
missing contacts made the amplifier inoperative. This was not an isolated case
because I know of at least eleven other hams who have had to replace the output
bandswitch on the same model amplifier.
Ed contacted factory-service via an authorized dealer and described the problem.
He was told that the output bandswitch was damaged by: someone who had rapidly
switched the bandswitch while transmitting at full power. He had unpacked the
new amplifier himself from a factory-sealed carton. He knew that he had never
hot-switched the bandswitch. He immediately realized that he was talking to the
wrong people.
I have heard the same outrageous story from other competent amateur radio
operators who had talked to factory-service[?] about the same problem with this
amplifier. I do not believe that any of these people were stupid enough to try
band-switching the amplifier while transmitting.
He discussed his amplifier problem with me and questioned whether the voltage
capability of the tuning capacitor and the output bandswitch were adequate for
this application. Since the actual breakdown voltage of these components is
above 5000VDC at sea level, and the maximum RF voltage is only about 2600V-peak,
nothing should arc-over unless the amplifier was operated at an extreme altitude
that would probably cause the operator to pass-out because of anoxia. Clearly,
this was not the case in Phoenix, Arizona.
As the frequency of a specific AC voltage increases, its gas ionization ability
also increases. This effect can be seen in the manufacturer's voltage versus
frequency ratings for RF-rated relays: The rated RF peak operating voltage
always decreases as frequency increases. This is one of the reasons why the
waveguides of high-power radar transmitters are pressurized with dry nitrogen
gas.
The presence of an unwanted AC voltage, with a frequency that was much higher
than the normal 29.7MHz maximum, was indicated in Ed's amplifier. The source of
this voltage could be a VHF parasitic-oscillation.
I recommended that Ed install some input suppressor resistors consisting of a
pair of 10 ohm, 2W metal{oxide}film [MOF] resistors in series with the RF-input
connection to the 3-500Z cathodes. After replacing the original bandswitch and
adding the input suppressor-resistors, he was still noticing arcing in the
general area of the bandswitch.
He threw in the towel. He asked me to see if I could fix the unruly amplifier; I
said I would try. The amplifier and the original, damaged bandswitch, that he
had replaced, made the trip to California.
The damaged bandswitch revealed that the most severely burned/vapourized switch
parts were the anode tuning capacitor padder contacts for the 3.5MHz and 1.8MHz
positions. The next most-roasted contacts were for the 28MHz tank-coil tap. The
21MHz tank-coil tap contacts were burned less than the 28MHz contacts and the
14MHz contacts were not burned. The pattern was clear: Only the contacts that
were close to the anode were damaged. And the contacts that were closest to the
anode were damaged the most.
The voltage that did this damage had a remarkable ability to jump an air-gap and
also deteriorated very rapidly as it tried to travel through the inductance of
the tank-coil. HF energy would have no problem traveling through the inductance
in the tank-coil. The only kind of voltage that fits this profile is a
high-voltage with a frequency in the VHF range.
Before operating the amplifier, I installed a 5.1 ohm, 2W MOF resistor in series
with the HV positive lead. The resistor will act like a HV fuse and current
limiter if a full-blown parasitic-oscillation occurs. This limits the discharge
current pulse from the considerable number of joules of stored energy in the HV
filter capacitor bank. If unlimited, this current pulse can disturb the grid to
filament alignment in the amplifier-tube[s] which can cause fatal, grid to
filament shorts.
A ceramic 10 ohm, 7W to 10W wirewound resistor would provide even better
protection. A higher wattage resistor should be used only if justified by
increased anode-current demand because the resistor is supposed to burn-out
quickly during a circuit-fault and stop the flow of current. .
As a further precaution before firing-up the amplifier, I checked the 10W
cathode bias zener diode. As is often the case after a parasitic oscillation and
its accompanying large current pulse, the zener diode was found to be shorted.
The zener diode was replaced by a series string of (7) ordinary, perfboard
mounted, RF-bypassed, 1A, >50piv silicon rectifiers with the polarity arrows
pointing opposite that of the original zener. This provides about 5 volts of
cathode bias-voltage during transmit.
{The polarity is opposite because the new diodes will be operated in the forward
conducting (.75v/diode) direction instead of in the reverse, zener-breakdown
direction}
My first encounter with the unruly amplifier exceeded my wildest expectations.
Even with input suppressor-resistors installed, this amplifier would oscillate
reliably with only 2200V on the anodes on the 14MHz and 28MHz bands! With 3200V
applied, the amplifier was unstable on some additional bands as well. I was
impressed. It was an electronic "Pandora's Box". This amplifier was perfect for
anti-parasitic R and D.
This situation was amazing to me because I owned an identical model of the same
amplifier that had been stabilized by using the same input suppressor-resistor
circuit that was used in the unruly amplifier. The only difference between the
two amplifiers was the particular pair of 3-500Z tubes.
Ed's 3-500Zs had remarkably high gain. With 100W drive at 3.8MHz, they would
deliver 780v p-p [1520W PEP] into a Bird 50 termination. This does not
necessarily mean that they would have also had abnormally high VHF gain as well,
but it is probably a safe assumption after witnessing their ability to oscillate
at VHF.
I set the unruly amplifier aside for a week and discussed the problem with some
of my amplifier-builder friends. After some enlightening technical discussions
and a suggestion to have the amplifier exorcised , I was ready to proceed.
In every HF amplifier design, there is an unavoidable VHF tuned circuit formed
by the anode to ground capacitance and the total inductance of the wires or
straps between the anode and the output tuning capacitor. The resonant frequency
of this anode-circuit can be varied only slightly by adjusting the output tuning
capacitor. I measured the anode-circuit's self-resonant frequency in the unruly
amplifier, with a dip-meter coupled to the wire between the HV blocking
capacitor, and the anode-choke. I found a very sharp, high-Q dip at 130MHz.
Next, I checked the self-resonance of the center-conductor of the coax that
delivers the input signal to the cathodes. The input circuit self-resonated near
the same frequency. This was not good.
Much of the inductance that formed the resonance in the anode-circuit appeared
to be in the 50mm [2 inches] of "U"-shaped #12 copper wire that connected the HV
blocking capacitor to the top of the anode RF-choke. This innocent looking #12
wire has about 39nH of inductance. At 130MHz this inductance has a reactance of
+j32. I soldered a 5.1 ohm non-inductive MOF resistor, with "zero" lead-length,
across the "U"-shaped #12 wire to damp the Q of the tuned circuit. I "fired up"
the amplifier on the 14MHz band and applied drive power. As usual, I saw fire
and I heard a familiar bang. The fuse-resistor exploded again as did the added
5.1 ohm MOF Q damping resistor ! Thanks to the fuse resistor, the 3-500Zs
remained undamaged and unshorted after this, fifth, full-blown
parasitic-oscillation..
The 5.1 ohm Q-damping resistor's demise was amazing because it was virtually
shorted-out by less than 0.0003 DC ohms of #12 copper wire when it went kaput !
This resistor had an overload rating of 20W for 5 seconds and it had been
destroyed in milliseconds. The only thing that could have so quickly blown away
a tough, essentially DC and HF shorted resistor like that was VHF current in the
multi-ampere range.
I concluded that the anode-circuit self-resonance of 130MHz was probably the
culprit due to the 3-500Z's 110MHz+ rating and the fact that the input resonance
was tuned to almost the same frequency. If I could increase the self-resonant
frequency of the anode-circuit to a higher frequency, where the 3-500Z's
excellent amplifying ability was waning, I suspected that it might reduce the
chance for a parasitic-oscillation.
I also decided that, because of the extremely sharp dip at 130MHz, the high Q of
the anode-circuit was probably another contributing factor. This problem seemed
to be exacerbated by the fact that high VHF-Q silver-plated strap had been used
for the combination anode-suppressors/anode-leads. It did not seem logical to
use the highest Q material to build a circuit that obviously requires a low-Q to
prevent the creation of a transient- induced VHF seed-voltage that could start a
parasitic-oscillation.
Low-Q Conductors:
The obvious choice for a low-Q conductor is nichrome ribbon or wire. It has 60
times the resistance of copper or silver. Q-measurement tests on a VHF Q-meter,
confirmed that nichrome produces a much lower Q than any other commonly
available conductor material. Unfortunately, nichrome wire and, especially,
flexible nichrome ribbon, is not easy to find or inexpensive. Soft
stainless-steel makes a good second-choice because it has 10 times the R of
copper and it is commonly available.
Anode-Circuit Modifications:
The #12 copper wire was replaced with a strip of nichrome ribbon about 3mm in
width and 35mm long. A three-turn inductor, with an inside diameter of about 6mm
to 7mm, made from #18 [1mm] soft stainless-steel wire was connected in parallel
with the ribbon in order to stagger-tune the circuit. This increased the
self-resonant frequency of the anode-circuit to about 150MHz and also lowered
its apparent Q.
It is not possible to connect a VHF Q-meter to the anode-circuit of an
amplifier, but I concluded that the in-circuit VHF-Q had been reduced
appreciably. I arrived at this conclusion by judging how closely the dipmeter
had to be coupled to the anode-circuit to obtain a 10% meter dip at resonance
for each type of conductor material.
The factory-original, silver-plated, high VHF-Q L/R parasitic-supporters, were
replaced with low VHF-Q L/R suppressors made from two 100, 2W metal{oxide}film
[MOF] resistors in parallel, shunted by a 70nH inductor made from #18
stainless-steel wire. The inductor has 3-turns. A 9/32" drill-bit shank can be
used as a winding-form. To keep the circuit's VHF-Q as low as possible, #18
stainless-steel wire was also used for the the leads at the ends of the
anode-suppressor assembly. The ends of the wire leads are bent into circles for
mounting with the original screws.
Construction Notes: 1: The inductor and each MOF resistor should be parallel to
each other and separated by a cooling air gap of about 2mm. Note 2: To avoid a
short-circuit and to facilitate cooling, the inductor must not be wound on top
of the resistors because the conducting part of these resistors is on their
outside surface.
For an even lower Q and better parasitic-suppression, the conductors could be
made from nichrome wire in place of the stainless-steel wire.
If an amplifier shows signs of instability with the 3-turn suppressor inductors,
try 3 1/2 or 4-turn inductors. Caution, the inductance can not be arbitrarily
increased because too-much inductance will cause the inductor's voltage drop to
be too great for the parallel 100, 2W resistors on the 28MHz band. The reason
for this is that, on the 28MHz band, with an anode-voltage of 3KV, there is
approximately 1.8a of RF current-circulating through each 3-500Z anode lead due
to the 4.7pF anode to grid (ground) capacitance of each anode.
In amplifiers with longer anode-circuit lead lengths, two or more of these
suppressor assemblies can be connected in series with each anode lead for an
even lower Q.
Results:
The once unruly (TL-922) amplifier has shown no signs of instability since the
anode-circuit was modified with low-Q conductors - even with all of the screws
in the cabinet! The output power appears to be unchanged on a wattmeter although
it is probably about 10 watts lower at 29MHz as a result of using the low-Q
anode-circuit conductors.
The same anti-parasitic technique was used successfully on several unstable
Heathkit SB-220 amplifiers; two, Henry Radio Co. 3CX1200A7 amplifiers and also
on a notoriously unstable Viewstar amplifier that had previously destroyed a
pair of 3-500Zs and numerous, other components as the result of a
parasitic-oscillation.
A Closer Look At How And Why A Successful Parasitic-Suppressor Works:
A successful parasitic-suppressor must perform two, interrelated tasks. The
first task is to reduce the flywheel-effect of a VHF self-resonant circuit by
reducing the Q of that resonant circuit. The flywheel-effect is essential to
oscillation. Reducing the flywheel-effect will reduce the chance of a
parasitic-oscillation. The second task of a suppressor is to reduce the VHF
voltage-gain of the amplifier stage.
The voltage-gain of an amplifier is approximately proportional to the output
load-resistance (RL) placed on the amplifier-tube. High RL means high
voltage-gain and low RL means low voltage-gain. If the VHF voltage-gain of an
amplifier-tube can be made low enough, by decreasing the VHF RL , the VHF
voltage-gain of the amplifier will be so low that it can not oscillate at VHF.
If a high-Q conductor-inductor is used to connect the anode of the
amplifier-tube to the, essentially VHF-grounded, tuning capacitor, a high-Q
parallel-resonant-circuit will be formed. The capacitance in this
parallel-resonant-circuit is the output capacitance of the tube and the
inductance is the built-in inductance in the leads between the anode-connection
[plate-cap] and the tuning-capacitor. A high-Q parallel resonant circuit acts
like a very high resistance at its resonant frequency. Thus, the amplifier has a
very-high RL and a very-high voltage-gain at the VHF resonant frequency which
greatly increases the risk of a VHF parasitic-oscillation. See
Figure 1,C.
A low-Q, parallel-resonant circuit will have a relatively low-resistance at its
resonant frequency. If two, low-Q, paralleled, conductor/inductors of slightly
different inductance are connected in parallel and to the same capacitor (Cout)
a dual resonant, broadband effect and an even lower-Q will result. This is
similar to the broadbanding-effect that is achieved when the primary and
secondary of an IF-transformer are tuned to different frequencies. This
technique lowers the VHF-Q even further and decreases the VHF output RL which
further decreases the VHF voltage-gain of the amplifier. The goal of
parasitic-suppression is to reduce the net (VHF) voltage-gain, by lowering the
VHF-Q, which lowers the VHF load resistance on the amplifier-tube, so that the
amplifier-tube can not oscillate.
In a typical parasitic-suppressor, the two, low-Q paralleled conductor-inductors
are: the suppressor's resistor, which makes the lower-inductance current path,
and the nichrome inductor, which makes the higher-inductance current path. Both
of the inductances in a parasitic-suppressor can also be constructed solely out
of low-Q wire or ribbon as was the case for the low-Q replacement for the #12
copper buswire in the TL-922.
The "Bottom Line":
High-Q conductors, such as silver and copper, are the best choice for the
anode-circuit/tank-circuit conductors in a VHF amplifier or VHF oscillator.
Copper is the best material for the conductors in a HF tank- circuit or
tuned-input circuit. Silver-plating the copper will improve the appearance but
not the performance at HF.
Nichrome exhibits a very low VHF-Q. Thus, it is a suitable material to use for
anode-circuit, input-lead and suppressor conductors in an HF -amplifier. Round
conductors exhibit a lower VHF-Q than flat conductors due to skin effect.
Appropriate Conductor Sizes:
1/4 inch [6.35mm] nichrome ribbon conductor is satisfactory for anode-circuits
carrying up to about 8A of RF circulating-current. The circulating-current
through the anode-lead of a typical 1500W amplifier is usually much less than
this. The conductor width should be held to a minimum to lower the VHF-Q for
better stability. It would not be good engineering practice to use 1/4 inch
nichrome ribbon if a smaller conductor will carry the current. Bigger or wider
conductors are not appropriate unless a smaller conductor is overheating from
the RF circulating-current during 10 meter band operation.
The safe RF current carrying capacity of #18 gauge nichrome wire, in free-air,
is probably about 3 amperes at 30MHz.
Construction Tips:
Nichrome and stainless-steel can be easily soldered with an ordinary soldering
iron by using a special flux that is made for soldering nickel-chromium alloys
and 430ºF tin-silver solder. These materials are sold in hobby shops and in
welding-supply stores.
Notes:
There is no single "sure-cure" for every case of amplifier instability.
1. Taming especially unruly amplifiers may require the intelligent use of a
dip-meter, several anti-parasitic techniques, more than one L/R
parasitic-suppressor per anode-lead and a, VHF Q-lowering, 1 metalfilm [MF]
resistor in series with the L/R parasitic-suppressor.
2. In some cases, it may help to add a low-Q, series-resonant L/R/C suppressor
between the cathode and ground. The resonant frequency of this series circuit
should be at, or slightly higher than, the self-resonant frequency of the
anode-circuit. The resistor should be a 1 to 5, 2W MOF or MF-type and the
capacitor is 25pF. The inductance is controlled by adjusting the leadlengths on
the resistor and the capacitor. The resonant frequency of this circuit is
difficult to check because the cathode must be directly shorted to ground and
the resistor must be bypassed with a straight wire in order to find the dip on a
dipmeter.
In rare cases, a VHF self-resonance in the anode HV RF-choke or in the
filament-choke can become a player in a parasitic-oscillation. This problem can
be overcome in these ways: A filament-choke can be effectively isolated by
placing a VHF attenuator-rated ferrite-bead (Mu850) over each filament lead on
the filament side of the filament-choke. An anode HV RF-choke can be effectively
isolated by placing an unbypassed 10, 15W, wirewound resistor in series with
either end of the choke.
Parasitic oscillation can be one of the most vexing amplifier problems. If you
would like to discuss any part of this article or the malady in general, please
feel free to call me at [805] 386-3734.
Fixing a parasitic oscillation problem is definitely different. In the end, the
only reward you get is: no surprises. Just be sure that you put all the screws
in the cabinet before you relax.
End