The B&W 5100B was manufactured between 1955 and 1961 by the Barker & Williamson Company of Bristol, PA (near Philadelphia). This transmitter offered both AM radiotelephony and CW radiotelegraphy on the 80, 40, 20, 15, 11, and 10 Meter Amateur Bands. The unit had a self contained power supply. The 5100B is rated at 180 Watts DC power input to the final amplifier on CW, and 140 Watts on AM, typically putting out nearly 100w carrier. The transmitter is well known for its audio quality, its modular design (more on that later), and its heft tipping the scales at 85 pounds.

A separate phasing type sideband adapter was also made for this transmitter, the 51SB-B. It is housed in a matching chassis, and the 5100B has threaded mounting holes and cable feed-thru holes in its right side for attaching the 51SB-B. The adapter is connected via octal type, Cinch Jones, and several terminal strip connections. It is fully powered by the 5100B, as well as using the 5100B PA stage.

For AM operation the 5100B has a basic plate modulation topology of a pair of 6146's in push-pull AB2 as a modulator, plate modulating a pair of 6146's in a Class-C PA. The audio/modulator circuit consists of a 6U8A for the mic preamp (triode section) and second stage (pentode section), and a 6AQ5A audio driver which is coupled to the 6146 modulator tubes via a 4:1 interstage transformer. The outputs of the modulator tubes are fed into a modulation transformer which is directly coupled to the plates of the PA. With the exception of a Centralab couplate which resides between the second audio stage and the driver stage, the 5100B has none of the clampers, limiters or other audio limiting gimmicks found in many transmitters of the day.



One of the nice features of this design is the modular construction of this transmitter. The modulator, VFO, crystal buffer unit, and the entire multiplier/RF PA unit are removable as separate complete entities. They are connected to the chassis via various octal type and Cinch-Jones connectors which are connected through holes in the the chassis. Once removed, the only thing remaining on the chassis is the power supply components and the internal wiring. The VFO comes out complete with the housing for the circuits and the complete mechanical tuning knob and dial mechanisms attached. Most of these modules are removed simply by taking out just four screws on each unit, a few more for the VFO unit. The VFO and the Mult./PA also only require you remove the the knobs. The shafts just pull out the front panel. No fiddling or removing the front panel nor are there any sub-panels here. The shafts are not attached to it. They have there own mechanical support built into the respective module itself. For instance, if you make the proper cabling, it allows you to remove the modulator, plug in the cable and work on the modulator on the table in front of your transmitter and test the circuits while it is connected and running attached to the transmitter in situ.

This article details some modifications that can be performed to improve the overall audio quality. Near broadcast quality audio can be had after these mods are performed. There is a PDF document with a schematic documenting all of the mods mentioned here. This schematic is not a complete schematic of the entire transmitter and only includes the circuits within the scope of this article. However the circuits are shown contiguous, and not just chopped up into isolated segments.

Here's a link to this schematic:

http://www.qsl.net/w2wdx/images/Mods.pdf

A high resolution JPG of this is available here (huge file warning!):

http://www.qsl.net/w2wdx/images/Mods.jpg

A full schematic & users manual for the B&W 5100B can be found at the BAMA site. Here's a link to that document:

http://bama.edebris.com/download/b&w/5100b/5100b.djvu

You will need a program (DJView) to view the manual file. Download it at: http://windjview.sourceforge.net/

Some Simple Audio Mods

Changing the values of the coupling capacitors in the audio chain to .1µƒ, replacing the stock .001µƒ caps, is enough to fatten up the audio quality significantly. Also replacing the couplate with discrete components and changing the coupling cap to .1µƒ as well is highly recommended.

Another slick trick involves some more involved modification. This one does sacrifice the Key input in favor of using it for a line level input. It involves simply disconnecting the RF choke following the key jack input, replacing the Audio Gain pot with a similar value pot that has a SPST switch attached to the back. The wire that came from the mic amp stage to the pot is placed on one side of this switch and the other side of this switch is wired to the end of the pot where this was previously. The RF choke coming off the former key jack is then wired to the wiper of the pot. In this configuration all you have to do is turn the audio gain to zero or in this case "OFF", and the switch disconnects the mic preamp from the rest of the circuit. Plug in your line level audio chain and adjust the audio level using your outboard audio gear, preferably monitoring on a scope. If you want to simplify this you can simply replace the Key jack with a ¼" phone jack that has a normally closed contact and keep the stock gain pot. The contact disconnects the internal mic preamp when a plug is inserted. Remove the plug for stock microphone operation. (Note: This is the mod shown on the mod. schematic).

If you want to take this a step further, you can change the phone jack to one that is TRS (tip-ring-sleeve). Wire in a small 1:1 600O audio transformer that converts balanced to unbalanced. Wire the ungrounded side of the unbalanced output of the transformer to the pot switch contact and the RF choke (which leads to the wiper of the pot) to the other side. Wire the common to pin 3 of P501. See Fig.1 (The pot switch in this case must be a SPDT type). This allows you to drive the radio with a balanced line level signal when you turn the gain knob to the full "off" position. In the "on" position this switches the input to the internal radio audio driver chain with the pot controlling level from the stock mic input as usual.

One note: While Radio Shack offers a 1:1 transformer of this type, its frequency response is quite limited and may defeat the purposes of these particular mods, that being broadening the audio frequency response of this modulator. There are much better transformers, which are very tiny in size offered on the web. Jensen makes several (what I used was a Jensen JT-11SSP-6M), but these can be pricey for the more frugal HAM's. The Jensen I used is sever overkill but I had one from my audio profession I wasn't using, so I used it here. Many small high quality transformers which are much less expensive are available. Check online.

If you have an RFI problem or just want to prevent the potential of one, you could try this circuit. (See Fig 2) This circuit also has some additional changes that might better match the input of the 6U8 which could be incorporated into the original alternate circuit. Values may change if other transformer are used other than the Jensen & Coilcraft types.

These mods allow you have the original microphone input (for your beloved D104) with the gain control and have a line level input without changing the appearance or function of the radio one iota! (except for the front Key input).

Modulator Modifications

Some more sophisticated modifications of the modulator can be performed that can improve overall sound quality of the transmitted signal. Some of these that I have performed involve very minor drilling and installation of items like internal switches & connectors.

The suggestion was made to me one evening on the air by Jeff, W2NBC to connect an external drive stage directly to the grids of the 6146 modulator tubes. (Hmmm) This is the classic audio amplifier connected to a flipped around hi-fi audio output transformer to drive the grids. The method I used was specific to the audio gear I had available and other methods and gear can certainly be applied here. I had an old Dynaco ST-70 tube audio amplifier. One of the nice features of this old vintage amplifier is the use of very high fidelity output transformers. Also, the amplifier would have more than enough power output to drive the grids of the 6146's and maintain a high level of headroom in the driving audio amplifier. This amplifier uses two EL-34's in P-P.

I disassembled the ST-70 and removed one of the output transformers. Basically, I removed the entire "left" channel and all of its components, leaving it as a "monoblock". After cleaning up the chassis, repainting the transformers, and otherwise making it all nice and pretty, I rewired the amplifier and set the bias. I checked the amplifier for performance.

I took the removed output transformer and mounted it on one of those Hammond cast-aluminum project boxes, added two pair of five-way binding posts and a ceramic feedthru connector for the bias voltage for the 6146's. I wired the 8O secondary to one set of binding posts and the primary to the other binding posts next to the feedthru connector. Across the binding posts connected to the 8O secondary I soldered an 8O/50w resistor. For the mod that follows, any low power audio amplifier can be used. Output power of the amplifier should be between 8 to 25W. Any hi-fi P-P type audio output transformer can be used, flipped around and loaded with an 8O power resistor.

Now comes the mods to the 5100B that involve the drilling. (Don't fret ... it's just one tiny hole on the modulator sub-chassis). On the modulator module between the modulator tubes and near the edge of the chassis, drill an appropriate hole and mount a sub-miniature TPST switch. This switch disconnects all of the radio's circuitry before the grids of the 6146's and connects the grids and the bias voltage directly to several unused pins on the 11 pin plug (P401) that connects the modulator to the radio chassis circuitry.

In the modulator, on P401 remove the black wire on pin 2. This is a terciary output lead from the mod transformer that is not used for normal operation of the transmitter (except when used with the 51SB-B SSB Adapter ... I think, not sure). Insulate it well and leave it. Remove the ground connection from pin 10. (Don't worrry, there is already another ground connection on pin 9 and this connector can handle the current with just one ground pin. Besides the chassis is at the same potential, and these pins only come into play when you have the modulator module removed from the chassis). The lead connecting the grid switch of V403 goes to pin 2. The lead connecting the grid switch of V404 goes to pin 10. These are used to connect the switch which controls the grid input switching to the terminal board on the rear of the chassis.

On the main chassis, several terminals on the terminal board (TB501 Term. 1,2 & 5) are utilized to feed the signals from the external source to the modulator grids via its large multipin plug on the bottom of the modulator sub-chassis. To do this remove the HV wire from terminal 1 and the zero volt wire from terminal 2 and insulate them well and just leave them. Then move the wire from terminal 5 and move it to terminal 1.  Also locate the -65v negative bias voltage and wire that to terminal five. You may want to take it from the J503 connector or the wiper of R510 (Mod Bias Adjustment) mounted on the chassis near the PA section.  Disconnect the ground lead from pin 10 of the 11 pin jack (J503) and run a wire to TB501 terminal 2. The wire you moved from terminal 5 to terminal 1 on TB501, connects the other switched grid lead.

Now connect the positive lead of the flipped output transformer to terminal 2 and negative to terminal 1 and connect the center tap of the transformer (the ceramic feedthru connector) to terminal 5 on TB501 for the bias. Fire up the transmitter and run up your audio amplifier. Using a scope, monitor your transmitted output waveform for 100% modulation.

Refer to the schematic provided here to clarify the connections.

The Simple "Turbo" Mod

While the pair of 6146's in the PA can produce a 90-100W carrier output on this transmitter, you may want  punch that up by using an external linear amplifier. Unfortunately, most commercially built amplifiers will be toast quite quickly if you drive it with that carrier level. Not to mention you will have no headroom left on top of the modulation peaks, if you have any.

To reduce the carrier output on this transmitter, a very simple and effective "mod" can be performed. Simply remove the jumper connecting terminals 6 & 7 on TB501. This an external access point for the plate voltage which goes to the modulation transformer. I should point out here, the voltage across these terminals is 635V and can really smart (or even kill) if you are not careful. Across these terminals place a power resistor between 7k & 20k (depending on the amount of output you need to drive your amplifier). I used the type that has the screw terminals on it and mounted just above the terminal strip on the back of the chassis on the perforated grill. I used two ceramic standoffs to mount the resistor, and uses high voltage wire down to the terminals. The resistor should be rated at 40W minimum. No change to screen voltages or any other modifications seems to be necessary, as far as I have observed.

That's all there is to it. Works well!

The Simple Soft-Start

A nice mod that limits the inrush current when turning on the B+ (and to get rid of those loud "ka-chunks" that could at times shake the entire cabinet of the 5100B) is adding a thermistor in series with pin 4 on the Control Relay (X504). A CL-60 type thermistor is a good choice. The time constants and current rating is perfect for this transmitter. It is rated at 5 amps steady state current, 10 ohms cold resistance and about 0.2 ohms when hot. These get very hot, so take care where mounting it!

This mod came from Mark Bell, K3MSB.

Replacing the Selenium Rectifier

There is a selenium rectifier (SR501) used on the negative bias voltage supply. These are prone to failure and should be replaced. Remove the selenium rectifier from the ceramic standoff it is mounted on. Screw a terminal strip with multiple lugs onto this standoff. Watching the polarity, use a 1N4007 diode to replace the selenium. In order to match the voltage drop of the selenium rectifier, place a resistor in series after the 1N4007 to achieve the correct voltage. You should do this by trial and error until you achieve the correct voltage of -140V on the negative side of the 40µƒ/150V filter capacitor (C519) that follows the diode.
  

Some Things to Check When When You Get Your 5100B

Since this transmitter is now over 50 years old, you should of course change all the electrolytic capacitors. Most are contained on/in the main chassis, and two more are found in the modulator module. Use high quality NEW (not NOS) capacitors. The FP can type electrolytic (C520), which is a 40/40/40µƒ @450V, is still available if you want to "keep it original". CE Electronics is still manufacturing this can new, using Mallory's original equipment. But they are not cheap, costing around $34USD at retail pricing. It can be purchased online from Antique Electronic Supply (www.tubesandmore.com). I only suggest them since they usually have this can in stock all the time (as of the time of this writing). Of course you can simply use ordinary axial lead caps underneath to save money. Whatever type you choose, do this first.

I stayed with the stock values of capacitance. On the bias supply I recommend raising the voltage rating from 150V up to 450V on these capacitors, since the first filter cap in that supply is very close to the normal operating voltage of -140V. I have found the amount of capacitance in the stock design is more than sufficient, and the stock supply is fairly stiff as is. No hum or ripple was detected anywhere in the circuitry. The oil capacitors in this transmitter (one used across the HV choke & two feed-thru filter caps on the AC input), most likely do not have to replaced. But it is advisable to check them for leakage, as a rule of thumb.

One other suggestion I would like to make, is on the resistors (R520 & R521) used across the HV filter capacitors. While the original design uses  68K, 2W, 10% carbon composition resistors, I went with hand matched 2W, 5% Metal Oxide Power Resistors instead. Even distribution across these capacitors allows the filtering to work better, so the closer tolerance coupled with the better long-term reliability of these types of resistors seemed like a good choice. It is a minor thing and not important or even necessary, its is just small detail I felt could be addressed.

While you're in there, you should also check the values on the various carbon composition resistors. They do change value over time and should be checked. Replace any you find out of value.

The 5100B uses two continuous bead chains to drive the mechanical of the multiplier. These can stretch over time, causing the chains to slip. There is an "idler" wheel on each that can be adjusted to take up slack. Do not make the adjustment too tight. And whatever you do, don't break those bead chains! They are the rarest & most difficult to find components in this transmitter. They are made from pure unobtainium! If the multiplier tuning control works smoothly, it's best not to mess with these chains.

If you find you cannot reach 100% modulation you should check the 4.7O/1W filament dropping resistor (R415) for the 6U8 (V401) inside the modulator. Its value can and does change over time in this transmitter and it will reduce the modulators ability to achieve full modulation.

You may get a scare the first time you tune your "new" 5100B when smoke appears near the modulator the first time you switch it into tune mode. Inside is a very unique plate voltage primary dropping resistor (R525 "Special Resistor"). It is comprised of a disc of mica wrapped with about 20 turns of Nichrome wire, which is then sandwiched between two more discs of mica. It is mounted to two ceramic standoffs. Sometimes if the transmitter hasn't been used for a while dust and stuff may build up inside or on this resistor and burn off when the transmitter is first tuned. If you find this resistor is not functioning, it cannot be found. It has to be made. I bought sheets of mica & nichrome wire, and using an x-acto knife and alot of patience, made a new one that functions and looks exactly like the original. I used the original as a guide for cutting the mica sheets. I made two, just to have spare. If I can do it, I assure you it can be done by anyone. Since it's not mentioned in the schematic or manual, the value of the resistor is 27O at probably about 50W. Any power resistors near these values can be used, of course. The circuit seems to draw about 30-35W while used.