here's a nice screen modulator that follows the modulated plate

[Opcom]

This is from a GRT-3 transmitter unit.

Instead of a separate modulation transformer screen winding, it uses a small cathode follower to sense the modulated 900V plate supply and use that to regulate and modulate the 190V supply to the 4X150 screens. The TX had a 200-20,000Hz frequency response so this screen modulator should be good enough in that department.

So there's a 5763 cathode follower for a pair of 4X150 screens.
Next, a 4X150 cathode follower modulator for the 4-1000 screen.
Maybe not but it seems simple and scalable and that's why it is interesting.

Plus, if anyone sees one of these little (gray or od green) plug-in sub-chassis, now you know what it's for.

[kb3ouk]

This has to me about the best way to modulate the screen of a tetrode or pentode I've seen yet. Looks like all of the screen voltage and current comes from a seperate supply, that's good. One tube that I think this would be really useful for is the Russian GU-81M, a big pentode. That tube will make 400 watts with 2000 volts on the plate at 300 mA, but the screen draws 160 mA at 400 volts. That's a lot of extra current if you were getting scrren voltage from the modulated plate voltage through a dropping resistor. I've been tossing around the idea of building a transmitter that uses all Russian tubes, a GU-81M for the final modulated by either GK-71s or GU-13s (both are based on the 813, the GU-13 is an exact copy, and the GK-71 is a pentode version of the 813). A pair of GU-50s in parallel might work, but they are limited to 1000 volts on the plate, it looks like they are using 3x the screen voltage for the plate voltage, so this might not work that good.

[steve_qix]

That's actually pretty cool!

The string of VR tubes can definitely be eliminated (they're only used to drop the 850V supply down to something the 5763 can handle), and the .05uF cap across the first 330k resistor should go - this is a presence rise network that should not be used in a good quality system.

Use a separate supply for the screen - gets rid of the heat from a dropping resistor, gives you an adjustment and is better than a choke in the screen, by far.

That's a very interesting way to do it.

The circuit could also be equally accomplished (and certainly easier - no floating fil xfmer needed) using MOSFETs such as the FQA11N90, which will handle 900VDC.

Good find!

[N2DTS]

I got to wonder why they came up with this circuit.
I never had a problem with a dropping resistor from a seperate supply, and/or a choke.

My 4x150 rf deck seems to self modulate a LOT.
Maybe too much, I need to re run my tone tests, but I tested all the rigs before and got good results.

And for military stuff, hi fi was not in the picture.

Maybe this design was more for screen protection and multi mode use?
 

[VE3LYX]

Maybe it is a bad idea but I steal the screen volts for my ARC5s from the voltage drop across the neg lead installed power choke. I use it to run the ocs and the 1625 screens while the HV is saved for the plates. Got the idea from the econopower supply 1984 ARRL book however I cheated a bit to get + polarity.  Osc stays stable and screens seem happy. I have not tried modulation on these screens and that could be difficult perhaps. I use a series cathode modulator stuck in the CW key hole now or for a quick QSO a loop modulator. However screen mod is in the works so I am always looking for good ideas. Similar to what was said here I did this to get rid of the room warming, stinky, megawatt, power resistors. Been a few years now so seems good.
I am looking into your design seriously too. Very interesting. Thanks
don 

[kb3ouk]

The circuit could also be equally accomplished (and certainly easier - no floating fil xfmer needed) using MOSFETs such as the FQA11N90, which will handle 900VDC.

I thought you (or somebody) had discussed a source follower MOSFET circuit similar to this on here before. I think the circuit I'm thinking of was to be used as a screen modulator by itself, but there's no reason that getting the audio from the modulated B+ like in this circuit then feeding it into the MOSFET modulator wouldn't work either, and actually sounds like a much better idea. This method (using either a tube or a MOSFET) looks like it would certainly eliminate the problem with too much modulation that you have with a screen dropping resistor off of the modulated B+, and also eliminates having to deal with the issues that come with using a choke to self modulate the screen.

[Opcom]

While a transistor would do the job well, the first arc over is likely to pop it unless it's covered in transorbs and has fault current limiting resistors in series with it, which not a lot of people do. Maybe it's the extra capacitance? The tube can be arced over, and still used. But, things are not supposed to be arced over anyway. There's a huge range of solutions from the Frankenstein type to the surface mount type and all have their contributions to make and be enjoyed in their own way.

In the autotune/synthesized version of this set, the T-217, the screen is modulated directly with plate voltage using a voltage divider and dropping resistor.
MOD B+---33K--T--20K---GND
then from the Tap, a 15K pot to the screens.

I believe it is a space issue in the transmitter to use this simpler method because the autotune version is very densely packed having several rather high powered servos to turn the necessary screws. (I like servo controlled tube sets and want a T-217, and its R-278 receiver, and already have a couple of the MD-129 modulators)

The T-217's modulator, the MD-129, has a less elegant audio limiter/compressor than the T-282+MD-141 setup from which this screen modulator comes.
The T-217's compressor is acting on only one polarity of the voice signal instead of both, and a slightly longer release time. It has an interesting bridge circuit as the gain-adjusting element.

My take on this screen modulator circuit of the T-282 /GRT-3 is that it is intended to control the screen better when loading was imperfect and the screen current higher or lower.
Otherwise they'd have used the cheaper solution of the set of resistors. The T-217 tunes itself to preset positions and so may be less sensitive to an operator's error.

This opinion is from studying the training manuals for both sets. It is available on the same page "UHF Military Equipment" and is the direct link:
Training Book: training Book covers complete theory and repair of R-361/GR, T-282/GR, MD-141/GR, R-278B/GR, T-217/GR, and MD-129/GR."
It's 267MB but a good read.

[steve_qix]

Yes, if one were to make this with a MOSFET, the circuit would need the usual protections - which in this case, since it's slow audio, are easy to do.  In theory, the screen modulator would be virtually indestructible - at least from arc-backs and other electrical hazards.  A current limiting resistor, a zener in the gate side, possibly a diode at the output and a fuse should do it!

[Steve - K4HX]

I recall seeing a similar circuit in an old QST. I think it was for modulating a 4X150 or 4CX250.

[N2DTS]

I don't get it.
A LOT of circuits using the 4x150/4cx250 tubes use mod transformer windings or special circuits to modulate the screens, but I get a boat load of modulation on my pair with just a dropping resistor off a low voltage supply.

Running 300 watts out of a pair and modulating it with another pair in AB1, I get 1500 watts pep out of the thing, the highest peak power (ratio) of any rig I have.

But any design you see using these tubes in AM has some special setup for the screens.

I don't get it.

I recall seeing a similar circuit in an old QST. I think it was for modulating a 4X150 or 4CX250.

[Opcom]

It's probably because they are easily damaged due to small size. Look at the tube itself, the anode is some 7/16" diameter and 1" tall, and has 1500-2000V on it and plate current of 250mA. The poor screen lives between that and a very generous cathode and sensitive grid.

[N2DTS]

I do not see that as a good reason to go to all the trouble.
I see these  tubes as no different from any other wiith screens, a clamp tube, a high current cut out relay, or a current limited supply would work like any other tetrode.

For 30 years I never used them for AM RF service because of that note in the RCA tube manual, and when I did try them, I find they seem to work great.

I will do some distortion tests again and see if I can find anything wrong with them.