Jim, The circuit I built is in the red or yellow Bill Orr books, uses octal tubes, including the big triode pass tube (6sa7??).
I have the output metered on the front panel of the modulator, and it ALWAYS stays at what I set it at (350 volts). Crazy, but I just pulled some (used) tubes out of the junk box and plugged them in, and the circuit has been working without problems for well over 10 years now, without a single crap out! I don't know about the regulation, the meter never moves, but I don't think the 4x150 screens draw a lot of current. >From memory, the input voltage is around 500 volts, the range the regulator will work in is somewhat limited according to Bill Orr, but the circuit works great for what I do with it. There is a limitation in the filament to cathode insulation of the big triode pass tube according to Bill Orr... It was such a cool design, I had to build it, plus, its adjustable, and I put its control on the front panel, along with the bias adjustments for each tube(seperate). The 4x150/4cx250 tubes are interesting in that you can variac the plate voltage between 1000 and 2000 volts without changing any other setting (bias or screen voltage). A great way to play with mod power out and impedance.... Having the adjustments on the front panel allows experimentation with different values of bias and screen voltage. Cool tubes, 600 watts of clean audio in AB1, drive it with a 12ax7! I drive it with the audio power amp into a LS transformer, so I can use the same audio setup with all the transmitters, but you can drive the grids with almost anything. I suspect the screen voltage regulation is quite important as far as the distortion goes, and many tubes like the 4-400 can sound good if great care is taken, otherwise, they sound like crap. I could never get 6146 tubes to sound good tho, even with regulation and feedback. There are likely better ways to do it these days, but the design was so cool, I had to build it. I dislike solid state components at higher voltages, as much as I dislike running electrolytic over 400 circuit volts. Yes it works, but I don't think its as reliable as tubes or oil filled caps. Except for arcing over a feed through insulator between the mod and RF cabinets in the 813 rig, all my homebrew stuff has been 100% reliable, well over 10 years for the 813 pair and mod decks for that, about 2 years for the new push pull 812 deck. Not even a single tube has gone bad in anything in the shack, the home brew or the commercial stuff (Collins, Scott, Hallicrafters, R390a, Gonset G76, etc) There are very few solid state components in my stuff, only diodes really...those HV brick diodes and so on. Brett N2DTS > > > Brett, > > Yes I know that circuit. I looked a lot at some of > Bill's circuits when I > did my first regulator, and what I used has parts in it > stolen from Bill. I > found that I got tighter regulation with the 6AH6 error amp > tube over the > 6AU6. I also had the problem where every time I turned it on, > the regulated > B+ was at a different voltage (within a 5 volt range). Once > on the voltage > was solid until cycling power, and then it would jump to a > new stable value. > Digging through some reference material, I read that the OA2, > OB2 are famous > for this, and that is why more precise regulators of that > vintage used a > 5651 reference tube. I experimented some, and the capacitor I > placed across > the OA2 seemed to fix the problem completely, and that > regulator holds the > voltage like glue now. There is a maximum value of > capacitance that these VR > tubes can handle before they start acting like a RC oscillator. > Another issue is which side of the regulator do you > pull current from to > light up the VR tube? I used the output side as it seemed to > provide better > regulation (load and line), and lower noise output. The > drawback of course > is that the regulator series pass tube must give up some > percentage of it's > available current to support the VR tube. I had about a 2X margin of > available current to needed output current, so this was a > good compromise in > my instance. > I also played with sampling Vin and Vout at a certain > ratio to feed the > bias voltage to G1 of the series pass tube. There seems to be > a point where > the ripple rejection of the regulator goes through a peak, > and the output > noise nulls. Anyone ever play with that? > I sure like the 6W6 as a series pass tube. I got 5 of > them NOS in box for > $9.99 from a Ebay auction. Look at the plate curves in triode > connected > configuration. They are very stout tubes for the size, and > money (little > demand for these so they are cheap!). > Brett in your case I wonder what your Vin, and Vout > voltages are, and what > is the minimum Vin - Vout at Maximum load current that the circuit can > accommodate before going open loop (out of regulation)? > > Regards, > Jim Candela > WD5JKO
