The AM Forum

   The whole entourage of BC-1T iron came home with me from Nearfest. Thinking of something along the lines of a K1JJ 813 rig. The mod transformer is actually the Peter Dahl replacement, but it's got all the same terminals and windings, including the tertiary secondary winding that Gates used to "tickle" the plates of the 807 RF drivers. Now, since I'd probably be driving my 813's (or whatever) with external plastic, I won't have any driver tubes to apply that modulated voltage to. Could I, instead, use that tertiary winding to apply modulated screen voltage to my final PA's?
   Those two 807's in the BC-1T must generate at least 100 watts to drive the grids of the 833's, so that tertiary must be good for maybe 200 ma. My 813's, I figure, will pull maybe 30 or 40 ma of screen current each. According to the Gates literature, that winding modulates the 807's to 20%. I suppose I could use a rheostat to determine how much of that goes to to the screens for best linearity.
   I know many people here have employed Gates iron - has anyone done this?
   

That winding was indeed for the 807 drivers to increase modulation percentage.
Your idea of modulating the screens of the 813s might also be a good idea. I would regulated the screen voltage or generate it from a choke input heavily bled supply and it might be good to make the screen voltage variable with an autotransformer in the supply primary to find the right voltage. It should be ok to feed the DC for the screen through the cold end and bypass the cold end to ground at that point….

One wonders why Gates did not do this, but those 807s were there and why not use them to modulate the grid RF drive-right?

Interesting idea.

   Well, the 833 is a triode, so that's why Gates didn't do it. I've heard that the linearity of a class C triode RF PA can be improved a bit by modulating its grid drive a little, but I suspect that Gates employed the tertiary winding more for cost cutting, since it allowed them to use inexpensive 807's to drive the 833's. I've never worked in broadcast, but I've heard that the 807's were marginal in this application. In a tetrode/pentode, of course, applying modulation to the screen isn't optional, it's required; the method of accomplishing this is the choice to be made.
   In my case, I don't see why using that tertiary winding to modulate the screens wouldn't work: with the conventional method of applying modulated screen voltage to a tetrode/pentode RF PA through a dropping resistor right off the PA side of the modulation transformer, you're dropping voltage down to whatever is appropriate for that particular tube, but that voltage will still be effectively doubling on modulation peaks. The voltage on the screen side of my tertiary winding should in theory - depending on the mod transformer ratios - effectively double as well. The only disadvantage I can see is that a separate transformer might be required for the screen source, since regulating down to 400V (assuming 813's) from my B+ of ~2600 volts will incur a poop-load of dissipation, though of course you'd dissipate that with a standard screen-resistor approach, anyway.

I believe the ART-13 used a tertiary winding for that same purpose, and that was a single 813. With the Gates, the 833s would've been asking for more drive at peak modulation in order to still be fairly linear, ideally you would just use a robust driver that could deliver more drive than you would need at all times, but that's why Gates went with the modulated 807s, you modulate them in phase with the 833s and they would be delivering peak drive right when the 833s needed it. Biggest thing i would be more concerned with is the ratios you have there with the primary to tertiary in that iron, you want to modulate the screen but not overmodulate it. I have an RCA transformer here that i seem to recall is like a 5:1 turns ratio to the tertiary winding, and 1:1 on the seondary.

Tonight I'm going to tickle the primary with a little AC from a variac, and see what appears on that tertiary. Gates' description of the circuit operation says "During operation, this
plate and screen voltage is modulated approximately 10%", so it would seem that the ratio is not 1:1.

833A tubes are low mu triodes. Low to medium mu triodes exhibit a very good square law response to plate modulation so that linearity is very good.
Tetrode tubes do not give a square law response to modulation of just their anodes, so to improve linearity. The screens are also modulated.
Also Gates was able to get a higher percentage of modulation with less audio this way and save money, I think. They apparently started making smaller mod reactors and mod trannys in these later models and were trying to keep performance up, I guess…
K8JI wrote a nice article on modulating triodes and tetrodes on his website.
Nice clear explanation.
https://www.w8ji.com/amplitude_modulation.htm

It could also be good for negative feedback if you are not using the usual resistor ladders.

That is a great idea. 😎
Much easier to tap at lower impedance of tertiary winding.

be careful - there are low power limits associated with the t windings - 2W afair

Yeah, it's not a heavy winding, but in the Gates it did supply the plates of two 807's driving the 833's at probably around 100 watts, so it couldn't be too frail.
Don, K4KYV, has a good write up on that transformer over on the AM Window site.