Mark 4, 20 Amp Power Supply, upgrade notes
(Mark 4.2)
Back in the early 1990's, I was asked
by the MDRC committee of the day if I would like to take on the role as
Electronic components and project officer, I didn't see any problem with this
and agreed. After being elected into the position I soon found out that a
little more time would be required than I expected, most of which was spent
answering mail. One item that kept popping up was a couple of small problems
with a project released prior to my association with the club, being the 1988
Mark 4 Power Supply Project. The following is the upgrade notes to rid the
problems without a total rehash...
UPGRADE to MARK 4.2 as
written in 1993
To see
circuit diagram page down...
For over 20 years now the
Moorabbin and District Radio Club has supplied kits to the amateur fraternity.
Our current mark 4 has been quite popular since its introduction in 1988.
However, recent constructor feedback has necessitated the need for an upgrade.
The regulator circuit of the
mark 4 employs a linear design based on the long standing industry standard
"723" regulator from Fairchild (tm). Current trends obviously lean
towards Switch Mode designs as they are inherently more efficient, physically
smaller and cheaper to produce. However many switch mode designs are often
noisy, even when displaying good regulation characteristics and it is for this
reason that we have maintained a linear design.
The Mark 4 as with earlier MDRC
supplies is equipped with over Volts protection. This prevents the unregulated
24 to 28 Volts DC from causing damage to your equipment in the event of a
regulator failure. It is with this part of the power supply circuit that
constructors are experiencing problems,
the symptoms are as follows...
Cut-out on switch on, indicated by the over volts LED.
Cut-out on the application of a load, also indicated as above.
The over volts protection
circuit is a simple resistor, zener
diode arrangement with transistor switching. In the event of a output of more than 15 volts the zener with begin to
conduct, once sufficient Voltage is developed across R16 (about 1 Volt) TR11
will be biased on. since
TR10 is connected in a feedback configuration with TR11 both transistors and
thus the relay will be ON. This condition will remain until removal of the
mains and discharging of capacitors C1-C6.
A problem with this type of
circuit is its sensitivity to spikes or noise. Spikes, caused frequently by
switch bounce, may develop sufficient volts across R15 to switch TR10 & 11
ON. This spike need only be for micro-seconds or less as the BD139 / 140
transistors can now switch in the hundreds of Megahertz range.
The following mods have been
carried out on several units without further problems...
1/ Replace R15 (1k) with a 1k2 if the secondary voltage is equal to (=) or greater than (>) 20 Volts AC.
2/ Replace C12 (originally 0.22uF) with a 2.2uF Tantalum (10 to 25V), be sure to observe polarity.
3/ Add C15 (10uF) and C16 (0.1uF) across the supply output (position not critical).
4/ Place a 470 Ohm, 1/4W resistor (see R10 on schm) across the Emitter and Base of TR10 (BD140), solder under PCB.
NOTE: Capacitors C7 & C8
were reversed in the original 1988 AR article, this error is not on the original schematic or
PCB artwork.
Mark 4.2 Circuit Diagram
To print this, change your
printer settings to Landscape...
Printed Circuit Overlay for
the Mk 4.2

Printed Circuit
Board layout for the Mk 4.2

NOTE: PCB measures
6 inches Sq
Economic construction for
lower power applications...
Some points on Heat Sinking
Depending on your application
the power dissipated in each pass element device (TR2 - TR9) can be as much as
30 Watts giving a total of 250 Watts for 8 x 2N3055. In most amateur radio type
applications we would be using the unit for SSB equipment so the duty cycle
will be much lower, say 20 - 30% and dissipation on average will be more like
10 Watts per device. A reasonable heatsink for 2 x 2N3055's for the 6 Amp situation should
have a thermal resistance of around 2 deg C per Watt. For higher powers - mounting
4 x 2N3055 on the one heat sink would require a thermal resistance of around 1
deg C per Watt. Physically smaller heatsinks can be used in conjunction with fan forced
cooling.
For information on PCB's please
contact me via Email
Regards, Chris VK3CAE
Email
VK3CAE: cja@rmit.edu.au