The AM Forum

The Surplus Radio Conversion Manual Vol 2 sates that "few amateurs have had the temerity to attempt to operate the transmitter as it was originally designed. In fact the conversion is to turn the VFO into an amplifier stage and drive it with a BC-696 Command Transmitter as a VFO. The CQ Surplus Schematics book goes further: It states things like: " The design engineers (at General Electric) responsible (for the design) almost had heart attacks when they received an ARC-5 for test comparison and found it to be a fraction of the size and weight, infinitely more stable and put out more power."

Well I have somehow avoided this thing over all these years of playing with surplus. I finally got one from Steve WA1HUD last Spring. This one was a BC191 in pretty tough shape missing tubes, the relay, connectors, many covers, tuning units and is apparently modified. Some surplus "conversion artist" had managed to replace the 211 (VT-4C) Master Oscillator with a crystal controlled 814 Pentode! This actually made sense and I think it could be a MARS or CAP conversion that was at least feasible for stabilization. Of course the tubes on all of these rigs went from worthless junk to valuable because of the audiophiles so they are always devoid of the 211's and Type 10 speech amplifier.

Faced with these facts, I decided to bite the bullet and purchase 2 Chinese 211's. These will be the modulators. But for the RF, I am going to attempt to use a pair of 814's which take the same filament voltage. The first very hard thing I did was to remove the meter panel. Wow. Over 20 screws with locknuts that spin, switches, meters and knobs all had to carefully be removed to get this panel off. But otherwise I could not see what the conversion was. I am glad I did because it was a mess. The mods were nasty and the solder was oxidized. This is going to be a process...         

That's going to be some project, but I bet it'll be fun and worthwhile once you eventually get it on the air. Good luck with it.

I bought all three of the Surplus Radio Conversion manuals back in the day, and even then I was disappointed to have learned that some of their conversions amounted to little more than destruction and/or parts salvage. An extreme example was the conversion of the MD-7.  If my memory serves me correctly the "conversion" began with stripping the chassis of all parts.

People P&M about all the nasty and otherwise conversions of mil. equipment.

However think of the thousands of hams, SWL's and budding geniuses (i i) that got their start and smarts by digging into quality rigs and parts hence obtaining the necessary education to advance the art past those GE engineers.

I cajoled, begged and waited years for a "real" electric train, a Lionel.
Parents finally relented, got a nice starter set when I was 11 and promptly ran it to death two days after Christmas.  Well I didn't dare tell dad that it lugged down and wouldn't come up to full speed.  Pretended that I was just running it slow if someone came in the room.  Took the engine apart and reassembled endless times as my knowledge of it and a lot of electric and radio stuff grew.  Finally as a young adult acquired the knowledge to remove the drivers (wheels) that hid the motor's rotor, then discovered a simple solder joint on one of the rotor winding's commutator had come loose allowing only two of the three sections to work which had caused the crippling of the engine all those years.

The line between stating an opinion on conversion techniques and P&M-ing is apparently a fine one that I seem to have stepped over.  :o

Oh no, you really haven't. I was Just referring to all the myriad conversion articles with passionate followers over the years now considered passé because of the scarcity of remaining, unmolested equipment.

Times change and the conservation impulse is certainly a noteworthy objective.
But many hams got solid, self taught educations based on otherwise unavailable or way to costly equipment that for the most part was constructed for a war planned to last eight years but ended in four.

Conversions at the least costly simply changed a rig from military assigned channels to amateur, perhaps by simple crystal substitution along with a homebrew power supply. More complicated changes, of course, ranged all the way to the bare chassis conversions mentioned earlier. ..and from the perspective of hams back then, with thousands of rigs available sometimes for the asking otherwise dumped, why not?

For another perspective of life back then, where were the conservationists when rigs were wholesale de-mil'd by Cat track crushing?  You had to live through the Cold War and nightly nuke dreams to understand the serious mindset of the military and most of the populace at large. Again, this is "you" general, hopefully not to be taken personally. I apologize for any misunderstanding. From you call, looks like you lived through that era too.(Sort of a smiley here but for the implications of what might have been.)

Each to his own. It's all a continuum. To be hard nosed about it, these items were made for military use. Putting them on the ham bands, even with zero mods could be called misuse, improper, whatever. Some years ago, I had a guy (a self-styled purist, or so he thought) give me an earful about ART-13 mods and how his was going to be "proper" with no mods and a carbon mic. When I asked if he was only going to use it in a B-17 at 10000 feet, he went QRT.  :P

I just can't get worked up about this stuff, either way. Do what you want and have some fun.

hi Steve .... some folks just can't handle the truth .... for them their perceptions are reality ... do what you want with the '375 and have a good time

Rick:

No apology is necessary because I didn't take your comments personally (NJ types like myself are not easily offended, although offending others is one of our natural talents). I'm a bit disappointed that my message didn't come across as intended, but that's my own fault for not being more careful in my writing. I'm not a purist. My initial comment had more to do with what was considered a conversion as opposed to a salvaging of parts. I modified many a military radio in my day, including ARC-5s (who didn't have one of those?), ART-13s, ARC-3s and a BC-342. I agree with the others, do what you want to your own stuff because it's your own stuff. It doesn't belong to history or future generations, it belongs to the guy who bought it. I have no difficulty getting to sleep at night knowing that by the time I wake up there might be one less bone stock ARC-5 or KWM-2 in the world.

Wow that modified oscillator was certainly more sophisticated than I imagined - I carefully studied it as peeled the onion and removed it. This is a crystal controlled electron coupled oscillator with a directly heated 814. This is a strange bird.

It was oxidized and pretty busted up but I did manage to capture the essence in a schematic. If this was a ham mod, it was a real engineer that did the deed. Take a look at this schematic and instead of discussing the ethics, let's look at the beauty. There certainly is something to be learned here. What are the two 100K resistors doing? Explain the keying method done with K1. What should be done with pin 4, the suppressor grid? What is all the tapping about on the tank? What is the Gimmick for?   

I think the gimmick is to keep the tube just on the edge of self oscillation to help kick start the crystal and the two 100 k resistors are just voltage dividers

Here is may take and I still have some questions. Yup the two resistors are a voltage divider for the screen. A crystal oscillator with 1000V on the plate!!! This at least brings the screen down to 500V. I still think it would have been a crystal cracker.

The Gimmick is a typical method to add more feedback. This is not a Colpitts Oscillator, it is an old fashioned tuned Plate Tuned Grid (TPTG) type so it relies totally on inter-electrode capacitance for feedback. The Gimmick helps to give more encouragement to oscillation.

Notice how the 1000V return is connected to the 200K resistor, not ground. This trick sets up an artificial negative cutoff bias for the tube equal to 1/5 of 1000V or -200V. When you close the key (and relay K1), this is shorted out and the 68K provides normal operational bias. You do not require a separate bias supply. We call this grid block keying. Pretty sophisticated for the time (GE used this method on the original Harley oscillator as well).

I am not going to give everything away: What should be done with pin 4, the suppressor grid? What is all the tapping about on the tank?

Mike WU2D

Oops, was typing almost the same thing as you.

Well it's not an ECO. It's a simple crystal oscillator; note the 68 k good resistor is RF grounded by .01uf. If it were an ECO some sort of RF isolation choke would be in the cathode circuit with split feedback bridge in the grid circuit making the grid /cathode the oscillator and the grid/screen/plate the amplifier.

Interesting divider in the keying circuit.  Ground the suppressor grid. It's not grounded or left floating.  Can't tell about the tapped plate coil until I see the rest of the circuit, how it's coupled to following stage.  If the lower tap is very low impedance, perhaps it is intended as a direct sampling feed. If really low, say 50 to 300 ohms, could have been used as a low power transmitter.

One things for sure, a modern xtal, even a Ft243 would probably crack.

Why the one Meg. R from B+ to ground? Serves no purpose but week bleed of HV, but don't know the T/r switching so if the Osc. Stage can be used as a single stage xmitter and the B+ untied to later stages, then, yes, a local bleed might be helpful, but then again, for CW , major, regulation bleed must be required at the power supply anyway.  Got me ?

Oh, you mean this is a quiz?
Ok, I'll bite.
Hook that low impedance coil tap to the suppressor grid.
That'll light things up !  ;D

Yes the 1M is 4 250K resistors in series as a bleeder. I have not studied the dynamotor yet so I do not know about a heavier bleed.

The suppressor grid should be grounded as you said. With DC on the filament this works great. But with AC, the engineer did something interesting. He connected the suppressor grid to the filament with a pair of 47 ohm 2W resistor from SupGrid to each side of the supply. This trick is normally done to create a virtual centertap for keying a directly heated tube at the filament. I would imagine that you would not want residual noise or hum on the suppressor grid of a tube specifically designed for suppressor grid modulation. This approach was interesting because I have not found it anywhere else in the literature. 

The BC191/375 is a true MOPA. It uses a Hartley power Oscillator much like the pre-war BC-AR-230/430 and later ARC-5's, but it directly feeds the PA grid rather than using a secondary link and tapped link neutralization. So the off center tap to the 814 sets a center RF ground zero phase. The lower tap is the feed to the 211 Final's grid. The top tap is to the neutralization cap to the plate circuit. So it is "direct feed" to a triode final. All of the tubes in the original set including the 211 modulators share a common 1000V supply. Thus the set is infamous for FM-ing and critical on neutralization.

Well here is the final schematic showing the suppressor grid approach. I have ripped this XTAL circuit completely out of the set and have installed plates to cover the old 211 Johnson socket holes. These have been outfitted with 814 sockets. I am ready to start rebuilding. I began to trace the circuit starting with the tube filaments. The modulator checks out. But I already have an open circuit on the RF filament I have to investigate and I see old electrical tape splices - UUUGH! Yes some more mods seem to be hiding in there.

Oh and look a the the picture and you will see the whole relay assembly seems to be missing.

This is going to be fun.

All the parts and goodies you could want...down South, too. :D

73DG

http://antiqueradios.com/forums/viewtopic.php?f=15&t=261432&sid=a06bcec38247656d7434c50beecab721

Well time for an update on the BC-191 project. I discovered that 4 of the banana plugs had been removed from the tuning drawer interface and I had to replace those. You might remember that they had modified the rig for XTAL operation with an 814 oscillator in place of the first 211. I made the banana replacements from standard banana plugs sawed to length, drilled and tapped and stuck 6-32 stock in and sweat soldered them. They look OK.

Getting the Filaments wired back to some semblance of normal was a chore.

Timmy's idea of using an 814 in the final in place of the 211 and sticking another triode connected 814 in for the oscillator will be interesting if it works. One benefit of using a pentode in the final is the reduced drive requirements and the promise of more stability. In the end it still is a triode power oscillator driving a PA in pure MOPA mode. My goal is to be able to use the tuning units as designed with no mods and get 814s to provide a low cost solution to replacing the costly 211's. And of course I want to cure the normal 191/375 maladies of drift and FM-ing and make it completely usable on 40 Meters.


There are gotchas. First of all as K4KYV will tell you that triode connecting a pentode with a low screen voltage rating is a bust. So you can no longer run 1000V on the oscillator if you do that. But as Timmy says, running 1000V on a free running oscillator is also nuts. They have three 2500 Ohm resistors in series feeding the oscillator off the 1000V.


So all I have to figure out how to get about 250 - 300V for the oscillator which should be a nice voltage range for stability in a conventional Hartley. Maybe I can put in a resistor to ground as a tapped bleeder. And by doing this simple trick, the voltage variation due to AM modulation should be 4X less for starters. If this voltage is too low for proper drive, I could chance raising the resistor to form a 500V tap as long as I limit the current with a resistor between the plate and screen. Comments? The ICAS rating on the screen of an 814 is 400 VDC.

The screen of the 814 PA will come off the modulation transformer through a screen dropping resistor for the conventional AM modulation hookup. Comments here?

Next I have to figure out what the tangle of wires chopped off where the relay once lived is all about...

Mike

  

Mike,

   I follow this thread with interest. I once had a BC-223 in a similar state as yours. When the 801 tubes disappeared in a move to Texas, I let the rig go after cannibalizing the main 7 gang tuning capacitor.

   I once used an 814 in a Viking I after bridging the power supply to 1500 volts. It worked well until the tube shorted internally (g1 to g2 I think). After replacing it with another, the same thing happened. I ended up wiring an 813 into the 814 socket, and that worked without failure for several years. Could run 200 watts input, and with an external plate modulator, that V1 rocked.

  Thinking about your situation, if it were me with the equipment in my shack, I'd set the oscillator stage as a pentode amplifier, with variable screen voltage to set the drive to the output 211. Then drive that 814 with a Knight V44 self contained VFO. This has a high level 80M output, so perhaps you could get 80-40 out of the rig using the 814 as a doubler on 40m. This would make a MO-Buffer-Buffer/doubler-PA (4 tubes in RF path). I guess that would be cheating now wouldn't it?  :D

Jim
Wd5JKO

Well I have the two 814 sockets wired and it is time to tackle the relay circuit. Before I attempt to replace the front panel and meters, I really do want to try to see if I can get the oscillator running and verify the bias situation. Now I am going to slow it down a bit.

As far as handling the screen grid on the PA, I did try to figure a max of 20 mA on the screen and dropping 700V, because the max screen is 300V and 400V ICAS on the 814. This should limit the screen dissipation to 6 Watts or less. This was going to take a screen resistor of around 35K to be safe. The question was if I simply wanted to tie the screen resistor below the modulation transformer (screen self modulation) or tie it above the transformer (Conventional combined plate and screen modulation).

I ended up using the guidelines of Dean McGorill. http://www.amwindow.org/tech/htm/scrnmod.htm

So split it between two resistors of around 56K and 100K and added the two caps, 0.0015 screen to ground and .005 across the 56K to the screen. This is silly for a transmitter of this type of course but there was room and we shall see what happens.

As you can see I am adding parasitic suppression on the plates and even on the grids. The 814's will have a lot of gain compared to the 211's. I also added two 0.01 bypass caps right at the fils on the sockets. Is this overkill? Likely yes. 

The relays were tough to wire. I had no relay and 10 cut off wires poking out at me. Wow was it tough to trace these! I ended up using a 4 pole relay combined with a larger 2 pole relay in parallel to give me more than enough flexibility to replace the original, at least until I find the real thing. Truthfully the pair of 12V relays are likely a lot quieter than the real thing anyway!

I was encouraged to hear some clunking when I keyed from any of the 4 jacks (2 for a key, 2 for a T-17 microphone and one built in test key switch). While I am in there, I am really starting to clean up switch, socket and jack contacts and the pots with Deoxit. Previously I had done only some cleaning and lube on the controls to free things up.

I also plugged in a VT-25 (Type 10) and an 814 into the sockets and they both lit up but that is the limit on a bench supply!

I mounted a test plate with a pair of SO-239's on it. I am missing covers and truthfully I do not expect covers to go on for some time anyway and I need a way to test.

Next I will focus on getting the oscillator happy with the triode-connected 814 in place of the 211 MO. I have a tapped resistor voltage setup but to start I will try to run the oscillator on a simple bench supply so I can see what I have and what voltage it likes. Timmy seemed to think that 500V or half voltage would work , but I am thinking less since I have a real Beam Pentode final setup and should not need as much drive.

Mike,
From the CQ Technical series 1960:
BC-375, BC-191
The BC-191, built by GE, building 89, 1 River Road, Schenectady, NY is an archaic behemoth whose design was finalized in 1935 and was produced in tremendous quantities for war destruction in planes and other expendable vehicles.  The design engineers responsible almost had heart attacks when they received an ARC-5 for test comparison and found it to be a fraction of the size and weight, infinitely more stable and put out more power.  The BC-191 for 14 volts the BC-375 for 28 volts otherwise they are almost identical.  There are a few moderately usable parts and the tuning units make nice cabinets.

I have been following your build and can't wait to hear your restomod on the (shameless net plug here) Old Military Radio CW Net on Sunday night 3570KC at 2100 Eastern.

Great work.

UPDATE! Well I have both sockets in the RF path converted over to the 814's now and I popped in the Chinese 211's in the MOD sockets and a Type 801A in the speech and they all lit up nicely. I also have confirmed that the triode connected 814 oscillates in the Hartley oscillator. The grid blocked keying worked at the modest 200 VDC I was testing with.

The 40 Meter plug in looked great but don't let that fool you; I can assure you that these things need a complete clean up inside with lubrication on all shafts and verniers and deoxit on the contacts to make the them anywhere near stable.

Next I will do the fairly complex filament adjustment from the manual. The 12VDC from my homebrew power supply is just adequate to handle the fils. Then we should be close to a smoke test with 1000 VDC on both stages to see if we have some power out!

This weekend I will try to get down to Steve, WA1HUD to look for an 80M plug in and a couple of missing covers and that elusive RELAY! My hambone dual relay solution is just a quick way to test the thing.

Mike - do you have pic of the unit with the front cover on?  I want to do a show 'n tell

Al

Not sure what you are looking for Al but here are a few pics. The BC-375 was the 28V version for planes. The BC-191 was also put in planes with 12 but it really was used in ground stations and trucks and command cars and even in jeeps.

Nice progress.  Do you have any add'l pictures of that jeep?  Very few of those WWII Willys or Fords were produced with 12V electricals in place of the standard 12V.

Wow Tom! I have to assume that you are correct in surmising that 99% of AAF unit jeeps would logically use the same equipment as the planes and it would be 28VDC?

THE HORROR THE HORROR...

A visit to Steve's basement yesterday to attempt to find an 80M tuning drawer and some metal (The elusive Relay, a top and the right ANT side and a front to finish the project was exciting, dangerous and surreal all at once. Imagine being buried alive in radio parts, freezing to death and being found as a ham radio mummy some thousands of years later.

When I took this project on, I had some goals beyond just restoring the beast.
1. I wanted to see if the now expensive 211 (VT4-C) tubes could be replaced by 814's.

2. I wanted to see if the thing could do AM and CW on the modern ham bands, at least on 80M.

3. I wanted to see if all of the harsh talk about this radio was true.

4. I had a goal of achieving 40M CW and AM operation.

I can report that the 814's are able to do 60 and 80 Watts out respectively with 80M and and 40M tuning drawers with 1000V. That is commensurate with the 211s. No problem - same filament- good drive - it works.

But - I am stunned. After several days of exploring how this thing loads up, I am mostly shocked by the horrifically bad architecture and neutralization scheme. Yes, a triode connected 814 will not require as much capacitance as a 211, I thought, and it should not be so far off that I need to change the variable or tap point in the tuning drawers- or so I thought...

In frustration I tapped up the 80M oscillator coil - down the coil, put capacitors in series to lower the neutralization feedback. But every control influences every other. Ultimately you can not properly neutralize the transmitter. An un-neutralized MOPA is no proper transmitter.

Was this thing in the same war as the TCS, ARC-5, BC-610 or BC-653 or 654?

 

Yes, I think it was in the same war, and if you remember, we won that war.  Do you know why we won that war?  Because our radio sets were far better than the other guys.

Fred

I was curious if the BC-191 was flown or if it was simply for ground communications. Turns out it was. It was the main liaison transmitter on the C-47 in a setup known as the SCR-187-A

So there were a lot them were manufactured for planes. The C47 was equipped with:
SCR-211-N Frequency Meter
SCR-AJ-183 Radio Set (Command) Pre-ARC-5
SCR- 187-A Radio Set (Liaison) including the BC-191 and BC-223
SCR -265-G Automatic Radio Compass
(or SCR-280 Radio Compass)
SCR-578-A Sea Rescue Transmitter
RC-32 Filter Equipment
RC-39 Marker Beacon
RC􀅪􀀩5 Inte rphone Eq uip ment
T-17 Microphone

The damage so Far: And I said that I was going to lay off any modifications. Well it turns out that the choice of 814's to replace 211's while sounding good, was really a bigger stretch than anticipated. You see, a 1924 something Western Electric low MU triode beast is not the same as a "modern" 1935 beam power pentode. Laws no. Even though this combo promised an easier to get and cheap solution to the tubes and the filament requirements were identical, these tubes could not be more different. The triode connection in the series fed Hartley worked out OK, but the voltage restriction based on the 814 screen would mean I should not run it any higher than 300-400 VDC rather than the 1000V on the 211. Actually that worked out because who in their right mind would operate a free running master oscillator at 1000V ?

But the PA final change out was really the issue. The 814 has a lot of gain and much lower inter-electrode capacitance. Technically it would not need neutralization to be stable according to the book, but not in this box! Taming it in such an open box with very long wires and meager bypassing, and with a plug in drawer approach, proved difficult indeed. There is no way that the 1925 vintage Rice direct grid neutralization on the MO tank in the drawer was going to work with an 814, heck it barely works with a 211. Grudgingly I removed the TU-6-B 80M drawer and prepared to blast. Somebody had already shot a couple of holes in it and it did not have much paint left so no foul so far. My idea was to emulate an ARC-5 version of the rice circuit which I know works with Pentodes (like the 1625). But when I saw that oscillator tank assembly I was not sure how I could get a tapped secondary over (or inside) the MO coil.

But then I spotted the PA tank. That looked more open. My idea was to convert to Standard Plate Hazeltine. With Hazeltine, you get the negative feedback by tapping the PA tank B+Feed which is bypassed, and connecting the neutralization cap on the opposite side of the tank as the tube plate side. This almost worked. I had way too much feedback for a pentode. This connection would be perfect for a 211 however! So if you do not like FM on your AM, think about this approach.  Turns out I needed to put the B+ bypassed "center tap" only 2 TURNS up from the cold side of the tank! Now I am neutralized - way good. Along the way I had to add parasitic suppression and more bypassing. Here is the before and after in simplified form... 

I still have a ways to go before I have a clean CW note, but it is a lot better.  Mike

Mike, on the jeeps bit.
The vast majority of jeeps went to Army contracts.  These were 6V. 
There were a number of other contracts for the Navy/USMC but the details escape me.  Again, most were 6V.  There were again some jeeps produced in 12V including the MZ radio Jeep with a TCS set.  A couple of these exist and have a separate PTO driven 12 volt system.
USAAF - I don't have any info.
While there may have been 28 volt vehicles produced during WWII I am not familiar with any.  The mil migration to 24 volt vehicular systems began with the M-series vehicles of the 50's.

I put covers on it today and it was doing 60 Watts into a dummy load and sounded passable on CW. So I put the beast on the air tonight for the maiden voyage on the Old Military Radio Net CW Edition on 3570 kHz. Oh they heard me all right. I was LOUD but the note was bad. She jumped frequency a couple of times too, and it chirped. Not well behaved for having so much work put into her. Wow this is going to be a tough nut. I do not think I have ever put so much work into a radio with so little to show.

I found the pedigree of the beast. It was a 1930-31 design by GE for the new Air Mail planes, the RT-76-A. The topology remained though military development in the 1930's.

Peeling the onion, I was wondering why the 211 modulators were idling very high at close to 50 mA each instead of the Class B value of around 10 mA per tube. To my horror I discovered that they derived bias for the modulators and the speech amp from the PA grid leak! Well the 814 pentode PA was not acting like a 211 in yet another important way. Geeze. I tried introducing a 10K resistor to the bottom of the series, but only got a little improvement. I measured -17 VDC on the modulators. I do not want to run A2.  I think I will throw in the towel and tie in some external fixed bias. Between the lack of regulation and decoupling on the B+, the direct stage coupling with questionable neutralization and this bias trick, you might start to figure out why these things FM. 

I made a small bias supply.  With -90VDC connected to the top of the bias pots, I can now properly adjust the Class B 211 modulators to proper resting current. They want 10 mA each or a total of 20 mA for Class B. Bias for the modulator and speech amp is no longer dependent on the RF Grid leak bias on the final. I have a total of 15K of grid leak for the final amplifier. This may be too high, but it is a good place to start. I still want the HV cutoff grid leak bias active so all tubes draw zero key up so I kept the returns the same at the bleeder tap point.

The speech bias pot seems scratchy and may be bad. I am modulating! The gain is very low and I can only load to 35 Watts out and get decent modulation.

So I will work on the speech amp tomorrow.

I am going to try an FM test later this week with my ICOM receiver. I will put a tone into the thing and tune for minimum FM as a critical neutralization test.

Seems R-614 is also part of the grid leak resistance.  The sa bias must be bias for the speech amp.  15K ohms seems a good starting point for the grid leak.  You can figure the resistance by knowing what negative grid voltage is required for class C and knowing what the grid current should be for the tube you're using.

Fred

Read some of the earlier posts,  FMing very often is caused by an oscillator that is running on the same frequency as the output stage.  An oscillator running on 1/2 the output freq and doubling up to the output freq will eliminate FMing.

Not sure if this is completely true with a crystal oscillator.  I think I saw a schematic for a 814 crystal oscillator which I think you're using.

Perhaps the authors of the CQ Surplus Conversion Manual deserve a little more credit than they are given today?  Within the time period and with the demands of producing something the BC-191 may have been useful at the beginning of the war but more modern sets like the ART-13 and the use of VHF pushed the  BC-191/375 aside into the trash heap of history. WW2 began with things like the BC-191/375, GO-9, TBW and TRF receivers and ended with useful transmitters like the ART-13 that stayed in service until the sixties and seventies. I would doubt that no one in their right mind would have been using a BC-191 in an aircraft after seeing an ART-13.  Not much that was used at the beginning of the war was considered useful by its end except things like the BC-348 receivers and the BC-610 transmitters, but this is just my speculation.
It is interesting to see that with your level of skill and expertise the amount of problems you’re having with this challenge, I don’t think I would try the BC-191 challenge.

Yes R614 is part of the grid leak but it is a low value parasitic suppression resistor.

They used 10K for the Grid leak for Class C bias on the old 211 PA and my gut feel of 15K for the 814 turns out to be right on purely by chance. Actually I kept the 4K 10W that was already there and added a 10K, 10W so 14K. Take a peek at the 814 conditions at 1000V with 10 mA on the grid and a goal of -150V bias.

Calculating you simply get 15K/10 mA = 150V. Not much to that math! 

I can not get 100% modulation over about 35 Watts out. That is disappointing. I do want to try another Final. It does not look too bad on the scope with pretty good symmetry. You may remember that I selected a 39K screen resistor and this also agrees with the spec., so I think I have the Final set up correctly. Drive may be another story, but the 814 should be a lot easier to drive than the original 211! I am tapping lower on the Harley coil for less chirp on CW.

 I adjusted the bias on the speech amp dynamically with a sine wave and went for least distortion at high level modulation. The 801A that I am using in the Type 10 socket is a beautiful triode with impressive power out specs.

Next I have to get back on my power supply and make sure I have enough C in my filament circuit. I am getting a little shy on voltage now that I am lighting 5 tubes.

I am getting close to an on air test. The Covers are staying on. Next I will try the dynamic neutralization test with an ICOM RX in the FM mode.

I have one of these BEASTS, it is on my to do list when I retire. We have a number of guys here in California/Nevada who have these on the air, all sound bad to you don't want to know. The main reason I have one, with a few spare 211's, is that this is what my father would have used as the Bombardier/Navigator/Radio Operator in WWII on a B-24.

    A few years ago we had a BC-375 Shoot Out on the Sunday Morning Vintage Military Radio Net, we had 4 BC-191's and 3 BC-375's check in. Here is an article I wrote that ran in Electric Radio and a nice New England Fall BC-375 picture.

Maiden voyage this morning on the Old Military Radio Net. Lucky that there were 3 other BC-191/375 transmitters on for an unbiased  comparison. All FM and none are anywhere as good as a stock Apache mind you. You basically must have at least 8 kHz of BW to "hear" them properly.

I started with a carbon microphone, switched to a carbon mic replacement which is a T-17 with an electret and Darlington and finally a low Z amplifier with a dynamic microphone.

No Soap. They heard me but my 191 was by far the worst performer in terms of audio quality and FMing. This nut is not cracked.  5 weeks of hard work and using modern tubes have produced a worse situation than the stock units running 211's!

After eliminating neutralization, coupling and bias PULLING as the major FM causes, it is time to investigate the oscillator regulation  which is related to frequency PUSHING being the FM source. My approach will be drastic. I will attempt to separately regulate the oscillator to investigate.




 

Mike,

Just out of curiosity, is the FM'ing at the audio frequencies or is it "modulation envelope warbling"?

By the way, since my father was a ham,  as kids in the late 50's and early sixties we had some of the tuning units to play with.  In the winter our picnic table was stored in the basement and we would throw a large tarp over the table and set up the tuning units and other junk underneath.  This was our space capsule.

wow!   this one is being tough  it would be interesting to change the mo to a buffer and excite separately to see iffin it would settle down ... I like the lockout / tagout idea .... we used them in the nuclear biz .... massive woe to anyone who disregarded the tag !

Your suggestion of external drive is exactly what was suggested by the CQ conversion handbook VOL II. They basically turn the VFO into a neutralized buffer. But that is cheating.

It was quite disheartening to have to remove the hundred or so screws all covers and strip it down to where I started. Removing the metering panel was again - much fun.

One (other) problem on these is the "family" of TU's.

Many have stated the TX and the set of TU's were married at the factory, and swapping them was an invitation to problems.  I suppose this was due to the N+1 rubric of handbuilding them.

Complete matched sets were much easier to get going according to the blurbs in the '46 to '50 ham rags.  I am sure how they were indicated, maybe by serial #s.

73DG

Two weeks later and another update on the beast. Let's just say that many long held myths are being dispelled as I go through the transmitter to try to find the source of the characteristic FM-ing.

1. It is the B+ 1000V power source. This has always been a great theory. After all they tie all of the B+ on the BC-375/191 to the same point.
2. It is the Filament Source, especially on AC.
3. It is the tapped HV Bleeder below ground bias system for grid block keying.
4. It is the direct MO tank feed with no buffering.
5. It is the High unregulated Voltage on the MO stage.
6. It is the high Grid Drive requirements of the 211A PA.
7. It is the weak Neutralization scheme.
8. It is the funky PA Grid leak that is used to bias the Modulators.
9. In general lack of proper bypassing.


All of the above, while showing the limits of late 1920's design practice and all having minor effects, have been systematically mitigated or replaced with proper design practice. None of the above is the root cause. But let me tell you - I am getting close to understanding what is going on, and you are not going to like it.

Oh it gets worse - much worse, but I will get to that.

On the positive side I finally have a working RF section now with a steady CW signal and a passable note on 80M CW.

And the FM situation is solved.

This was accomplished by converting the Hartley power oscillator to an electron coupled oscillator in the form of an 837 (12 V Fil) driving the 814 final. The screen of the 837 HAD to be regulated and I added a 150 VR tube. The neutralization is tapped plate on the 814. The keying is standard grid blocked on both stages. The tuning unit had to be modified extensively for this of course but very few new components are involved. It is just really a rearrangement of connections in the drawer. So the transmitter is now solid on 80M at 70W out and with 3 mA on the grid of the 814 and a very conventional late 1930's MOPA design. Is this still a BC-191? Hell no. 

So my attention moved to the 211 modulators. I am running fixed bias now and made the mistake of not keying the modulator protective bias with the grid block on the RF stages (keeping the modulators at operating Class B bias at all times). The problem here as with any high power audio amplifier, is that you now have the potential to operate the amplifier without a load (the transmitter when transmitting presents a safe impedance load). Well as you can guess I achieved a situation where there was too much swing with no load and it arced over likely taking out the modulation transformer. Replacing a modulation transformer in a BC375/191 is way up on the scale of crazy. I have changed on modulation transformers on ART-13s and they are cake compared to what I am about to try. 

Very nice rig and progress Walt.
Be careful setting modulator bias and -tone pots next to the now very open HV plate line on that 814.  A long fibre screwdriver might be in order.

Rick, you can not see it but the plate lead is sheathed with clear tubing.  Mike

Oops, and your Mike too.
Thanks. I looked closer and saw that even with the 811's , etc. some of the adjustments look to be directly behind the tubes, particularly on the left of the partition. Guess you need a double right angle driver to adjust them.  ;D

Congratulations Mike.  I can't wait to hear you on the air.

An Archaic Behomoth perhaps, but the first exposure of Signal Corp students to transmitters and whose basic MOPA configuration today has gotten new life by Mike's infusion of the appropriate new cylinders and ignition systems for years ahead performance pleasure. Nice to see a systematic approach to objectively assessing  the nature and idiosynchrosies of the system and make the appropriate changes within the parameters for the basic configuration. As to the note regarding care around  the high voltage, use of the BC-191 functional training aid shown in the attached picture was an experience all by itself. Looking forward for the completion of the saga and establishment of the BC-191U , or updated configuration.

Anthony, W1ADR 

Did a little  cleanup on one of the pics.

The 191 sounded very good on the air today Mike.
It started out a bit rough and FM'ing but cleared up after you did something.
It sounded quite good, just the audio level was a bit low maybe, but little to no distortion and the audio even had some nice low end.
A very nice qso.

Well I submitted the whole saga in an article including pictures and schismatics to Ray and we shall see if he decides to publish it!