On Building A 1937 Modulator

[wa2fxm]

Looking for advice and comments on a plan to build a modulator for my 1937 Utah Radio UAT-1 CW exciter. The exciter output is a pair of parallel 6L6's rated at 80 watts RF input. The original modulator was Kit No.2 which when paired with the UAT-1 was marketed as a 75 watt phone transmitter. The only information I've found about the modulator is from the Oct 1937 issue of SW&TV which describes it thusly:

- 50 watt high fidelity speech amplifier and modulator
- Class AB fixed bias push-pull 6L6's
- Adjustable output impedance and universal input
- Audio line-up: 6J7 resistance coupled to 6N7, transformer coupled to a pair of 6C5's which drive the 6L6's.
- Power supply on a separate chassis uses a 5Z3 and an 80 for separate bias voltage

Building a modulator will require a new level of knowledge for me so I'm reading all that I can find about it. The 1937 ARRL manual shows a very similar looking modulator. I am thinking of using this as the basis for recreating the Utah modulator. Assuming that this is the right direction to be heading in my first question is about obtaining the transformers, particularly the modulation transformer. I'm assuming you don't just go on eBay and peruse all the Thordarson 8470's and then pick out the one you can afford? Any helpful thoughts about taking on this project will be greatly appreciated.

Mark - WA2FXM

[DMOD]

Can you post a schematic of the Utah Radio UAT-1 because we will need to know the impedance of the Final stage.

AC0OB

[wa2fxm]

Can you post a schematic of the Utah Radio UAT-1 because we will need to know the impedance of the Final stage.

AC0OB

My unit differed from original schematics printed in Radio News magazine. This is what I traced out.

[w9jsw]

Just noticed this ad for a 8470 mod trans...

https://swap.qth.com/view_ad.php?counter=1610314

[wa2fxm]

Wow. Amazing. Thanks for that jsw.

Mark - WA2FXM

[W3SLK]

Ok something to help you out with your search for parts. Those are old Thordarson P/N's. I was able to cross them to a more 'modern' number. Your T-2 T-8459 crosses to an 84D59 which crosses to a T-20D78 4:1 driver transformer. T-1 crosses (I think but it looks to be true) to a T-57A41 which crosses to a T-20A23. I can help you out with the T-20A23. I have to look closer but I didn't see a T-20D78 in my stash but I have a Halldorsen HVTS-3 that is listed P-P plates to P-P grids @35mA. I'll look again and see if it is listed at 4:1.

[wa2fxm]

I've started work on the modulator. Ugly bread boarding of the mic and intermediate amp seems to be working although with a bit of AC hum on the signal. Tons of thanks to Mike W3SLK for volunteering his 20A23 transformer. It will be put to good use. That listing for the Thordarson modulation transformer was old and it was long gone. But I found a UTC-20 universal modulation transformer that I believe (hope?) will work. It has wiring for a 3800 primary (same as the Thordarson) and multiple output choices for the secondary, 400 up to 30K.

So now I need my first lesson. Plate impedance for the UAT-1 6L6 finals for modulation parameters is Plate V \ Plate I, correct? With limited CW use I was measuring 350V @ 170mA at keydown on the plate = 2060. The 1940 ARRL Handbook lists 6L6 typical operation for Class-C Plate Modulation as 325 V @ 90mA, so with the 2 tubes in parallel that's 180mA = 1806. I assume the 2000 ohm connection on the S-20 secondary will handle this?

Finally, for now, I'm still looking for the Thordarson T-8459 driver transformer. I also found a UTC Transformer list where they say their UTC S-10 driver transformer is purpose built for PP-6C5 to PP-6L6, AB2, 60 watt modulator service. If anyone has one of these and wants to deal\trade\barter\volunteer please let me know. Thanks.

Mark - WA2FXM

[W3SLK]

Let me see if I have a T-20D78 in my pile. The S-series were like an economy type of transformer. I don't think you will get the frequency response you want.

[W7TFO]

Yes, the 'mystique' of the UTC "S" series is wholly unwarranted.

Running them even close to max ratings will speed up your failure time.

Audio BW is miserable, and the adage "UTC = you take chances" certainly applies with the "S" series.

73DG

[K9MB]

DJ has forgotten more about AM than I ever knew and my impression of the “S” series is the same. I probably have an S20 and maybe an S21 too. The fact that they do not weigh much is a bad sign I think . The total inductance will be low, I think.
If you cut the lows down it might help on these things to get more power, but it will not sound like good iron.
I like these Thordarson trannys better. The link below is a T11M76 good for 75-110 watts. Might work for you.
DJ has a lot if stuff- ask him since he stuck his head up. He will point you in the right direction.

Here are the S-specs.
https://www.technicalaudio.com/pdf/UTC_United_Transformer_Corp+TRW+OPT/UTC_United_Transformer_Corp_Type_S_terminal_arrangements_1954.pdf

Hi Dennis!

[W3SLK]

Looking at that transmitter it looks like you will only have about 40~45 watts out. So you don't need a humongous modulation transformer. Something around 30 watts should be more than sufficient. I think you will be pushing it with 'shove-yank(tip of the hat to Slab Bacon), 6L6's at 40 watts as stated in the 1942 Editors and Engineers radio handbook.

Sorry, I don't have that transformer.

[wa2fxm]

Ok All,

Thanks for the tip on avoiding UTC transformers, although it's what I have right now. Mike, thanks for looking again through your pile of iron. There are 2 Thordarson modulation transformers on eBay right now, the mentioned T11M76 and a T19M16. Understood that these are probably overkill for the UAT-1. It's not clear to me from the attached spec sheet that these would even work with the PP-6L6's and parallel 6L6's.

Mark - WA2FXM

[K9MB]

I believe that it can match the loads ok and bigger is always fine if you have the space.

There are other more compact options like a Stancor A3893 or A3894 but I do not see either right now.

You could also look at a mod tranny from a Viking II or DX100 though they are fixed ratio trannys.

Below is a listing for a Viking II transformer. It fed 807s and 6146s

[wa2fxm]

Still searching for a driver transformer to start this modulator project. And still mindful of the learning curve I'm facing dealing with trandformer specs. And understanding the limitations of the UTC S series. Would the UTC S-10 driver transformer be appropriate for the anticipated 6C5 to 6L6 tube lineup? It's not the Thordarson 20D78 but it's the only thing I've found that comes close, I think.

Mark - WA2FXM

[Carl WA1KPD]

Mark,
Be careful of the modulator in the Handbook. I built one from either the Handbook, or QST using 6L6's and had all sorts of trouble getting it to work. Ultimately discovered the pinout on the tubes was wrong. Later found a "Strays" correcting the error so it is possible it was fixed for the Handbook.
Hopefully with the various sources you are using you are in the clear.
 

[wa2fxm]

It's time to finally start on this UAT Modulator project. I've collected most of the parts to recreate the modulator deck which will follow the 1937 ARRL Handbook 50 Watt 6L6 Modulator. I'm flying blind as I've still not found any UAT-2 Modulator schematics anywhere. The ARRL design matches the UAT-2 tube lineup as described in the period literature except for the 2nd audio amp which was originally a 6N7.

I've also lucked out by acquiring an original UAT-2 power supply and cabinet. Again with no schematics to go by I've traced out the power supply wiring I can see without unsoldering anything yet. Studying other late 1930's modulator designs and doing a lot of guessing here's my first stab at what is in there.

1. I'm not sure exactly how the bias supply is created. Maybe from a tap somewhere on the main HV secondary?

2. What is the exact function of the relay? It appears to be tied to the Send-Receive switch, with an output going through the rusted screw-in tip-sleeve plug but the connections do not make any sense to me.

3. I'm assuming low level audio voltages will come from the modulator screen voltage tap?

Please chime in with big picture suggestions and detailed caveats. The details of AM modulation is all new territory for me so all input will be welcome.

Mark - WA2FXM

[wa2fxm]

The traced out wiring.

[N2CQR]

I know you are not working on the same circuit (Shep built a Heising modulator) but the time frame is similar.  I thought the group might take solace from Jean Shepherd's tale of modulator woe:
https://soldersmoke.blogspot.com/2015/03/shep-on-parasitics-and-troubleshooting.html
EXCELSIOR! 73  Bill N2CQR

[WQ9E]

WRL's UM-1 (Universal Modulator) is a good starting point for a twin 6L6 design.  I have one paired with a Globe Chief 90.  UM-1 manual/schematic:  http://bama.edebris.com/manuals/wrl/um1/

As to UTC S series, while I wouldn't go out of my way to buy those I picked up a homebrew rack rig using those transformers and it has survived and gotten good on the air reports and probably had quite a few years on it before I acquired and restored it.

With fewer modulation transformers and supporting cast available each year, we really can't afford to discard or turn up our nose at everything that is less than optimal.  I agree that operating S series at its maximum ratings will shorten component life but that is true of pretty much every piece of iron that wasn't very conservatively rated by its manufacturer.

It is sort of like sweep tubes used as finals; treated properly they will give many years of service but abuse them and treat them like it was still the 1960s where you could buy them cheaply at the neighborhood grocery or drug store and it will get expensive in a hurry.  No matter how good the iron was in 1940 or 1950, it makes sense to treat it like the senior citizen that it is now and not abuse or overload it.

Rodger WQ9E

[wa2fxm]

I know you are not working on the same circuit (Shep built a Heising modulator) but the time frame is similar.  I thought the group might take solace from Jean Shepherd's tale of modulator woe:
https://soldersmoke.blogspot.com/2015/03/shep-on-parasitics-and-troubleshooting.html
EXCELSIOR! 73  Bill N2CQR

Yes. Every night. 10:15 pm. WOR 710 AM in my upstairs bedroom out on Long Island. This is classic Shepherd. Thanks for the inspiration. And the memories.
FLICK LIVES. 73 de Mark WA2FXM

[wa2fxm]

First detail question:

Large wire-wound, tapped, dropping resistor measures 2 Meg out of circuit across the large tapped section and only 2.2K across the small section. I've never come across a resistance as large as this in a dropping resistor. Is this normal?

Mark - WA2FXM

[W7TFO]

Not really.  Sounds like it may have been wet at one time that left mineral deposits across an open section.

73DG

[W3SLK]

WA2XFM said:

Large wire-wound, tapped, dropping resistor measures 2 Meg out of circuit across the large tapped section and only 2.2K across the small section. I've never come across a resistance as large as this in a dropping resistor. Is this normal?

I have learned, by recent past experience, that these ceramic wire wounds suffer from age and open up rather easily. I had a 1937 RCA PA that had its dropping resistor opened up. When I went to my local parts jobber, he said he had one close to the value needed but it had been sitting on his shelf for 30 years! Unfortunately, it was bad. When we tried to combine values in an effort to make what I needed, the ones from Mexico, (only 5 years old), disintegrated in my hand. I haven't even attempted to catalog the ones I have from my parts pile to see if they still work or not.

[wa2fxm]

Cleaned up the power supply deck a bit, re-soldered connections, replaced the 2 bias filter caps and the power cord. Getting decent voltages with a kludged low power 60K double tapped wirewound.

[wa2fxm]

Here's my first real stab at the modulator. The chassis was going to be temporary but it may wind up being the real thing unless I can find a wider one with 1" height to fit the big Thordarson modulator transformer.

All component values are within 20% of the ARRL schematic above. I've wired up the 6J7 mic amp, 6C5 intermediate amp through the audio transformer, the 6C5 PP drivers, the driver transformer inputs and the outputs to the 6L6 sockets. I'm using +325VDC from the Heathkit power supply for B+ and a separate 6.3VCT filament transformer. The B+ is drawing about 20mA and the 4 tubes 70mA. With a 20mv 400Hz input I'm getting a decent 30VPP at the 6L6 socket grid pins with volume control about 1/3 the way up. Above that waveform starts to flatten out at the peaks.

I'm getting a 60 cycle hum from the filament transformer that increases when I move it around. With volume control down it's .1VPP at minimum and up to .5 VPP when it's close to the audio transformer. With the volume control at max it's about 10VPP and decreases to 6VPP if I touch the chassis. I'm assuming that once I start using the real power supply which will sit on a separate chassis on the lower shelf of the UAT-2 cabinet this will not be a problem? I'm guessing my under chassis filament wiring is also problematic and could be cleaned up?

The 6C5 drivers are hot to the touch.

Still looking for a 50K tapped power resistor for the UAT-2 power supply.

So am I on the right track so far?

Mark - WA2FXM