Design considerations for high voltage, high power, tube output section/PSU

[nsubulysses]

Feel free to have a laugh at my expense but when this YBA3 showed up it really got my mind going. If there is an elegant way to get 160W from 4 6CA7 with no strain on the power tubes I am jumping down the rabbit hole trying to find out how that's possible.........because, if I can get 160W and long tube life instead of 138W I will be very happy. There's your chance to laugh

I ran the amp originally with ONLY 1K screen grids and no shared 1K/20W in series with screen grid supply and it was awesome and put out 160W but probably after 20 hrs of playing it was down to 120W at clipping. Too harsh on the screens. Now with 1K in series and 1K to the screens it has been played full volume for about 4 months and still puts out 138W (I had to get a new quad of power tubes )

[nickb]

Exactly how are you measuring it? It sounds like you are using a DMM. In that case all bets are off. The only time I trust them is a pure sine wave AC or constant DC. I do not trust them to average the waveform.

To measure the average with DMM I suggest you make a little low pass filter across the screen resistor say use a 10K resistor and a 0.1uf cap in series across the screen grid resistor and measure across the cap. This will give you the average drop across it. I'm assuming the input is a steady 1KHz.


When I mean peak I am thinking about the highest point on screen current waveform, not the peak output. Also, I think you need to to drive it hard to get the worst case as the screen current shoots up when driven hard.

[nsubulysses]

Thanks for your response. I know I'm going on kind of a wild tangent. But I really want to understand this as well as I can since I am actually just trying to make a crushing amp that will bring the house down. It's gotta bring the house down not the power tubes.

I'll study this further and see how much closer I can get to the true, accurate answer.

Yes I am using DMM all bets are off but it's supposed to be a True RMS fluke so whatever.

Thank you!

[nickb]

Thanks for your response. I know I'm going on kind of a wild tangent. But I really want to understand this as well as I can since I am actually just trying to make a crushing amp that will bring the house down. It's gotta bring the house down not the power tubes.

I'll study this further and see how much closer I can get to the true, accurate answer.

Yes I am using DMM all bets are off but it's supposed to be a True RMS fluke so whatever.

Thank you!

I don't think you are on a wild tangent at all. You are trying to make a tricky measurement that most people ignore and hope for the best. You are to be commended. And we all want to know the answer.

[nsubulysses]

WEll you're puttting in a lot of effort to help me understand so that is very commendable in its own regard.

I always wondered how all these very knowledgable people have time to answer questions but I really appreciate it

[nsubulysses]

In case anyone is interested in this PT OT comparison. the Traynor PT dwarfs my Hammond 278CX. This is supposed to be same transformer of Traynor Custom Special. is it possible they changed the design over time to make it smaller?? 2nd pic is PTs only

Since this amp does 140-150W at clipping and even more cranked, I ditched my Hammond 1650T 120W output transformer. The amp sounded bad when I would go past 6 on master volume. I asked Heyboer to make me a 150W version of Hammond 1650T and it's so awesome. I can go to 8 or 9 on volume and it STILL sounds good. No more grainy nastiness when you go too high in volume. And that's what I need is too high volume

Funny thing is this Heyboer 150W OT is bigger than 120W Hammond 1650T, but the old Traynor 1650T is STILL bigger than both of them.

Maybe it is the same transformer set since they do the same power. I don't know

[g1]

You mentioned a 470R screen supply resistor, there was a version like that, the non-master version. Page 4 of attached schematic set.
You said something about using a resistance substitution box to set bias, is the bias trim pot gone?
One more trick in these amps, (check that it hasn't been changed) the suppressor grid (pin1) is tied to the neg. voltage of bias rather than the usual grounded configuration.

[nsubulysses]

Thanks for letting me in on the tricks G. The 470R looked stock but I second guessed it because I didn't know there was a 1K and 470R version.

Bias trimpot is there. I couldn't reach a low enough range so I set the bias trimpot in the center of its rotation and pulled R38 150K 2W resistor to access more -DC. I actually ended up placing it with............150K. Whoops I guess it drifted up in value over the years. I never measured it.

[NoelW]

nsubulysses,
What happened to this amp and did you sort it all out?

The reason I am asking is that I am currently building something similar, for a client, with 2 x 6CA7 and 500V+ HT.
I am using Hammond transformers: 290HX PT/Marshall 100W (356VCT @ 420mA) and 1750N OT/Marshall 50W (50W 3200 ohms).
The single-channel preamp is aimed at clean headroom, as a pedal platform and the back end, from the tone stack to OT is 50W Hiwatt style.
I would like to know whether you got what you wanted and whether it turned out reliable, or not, with the high HT, using the JJ 6CA7s.

Any further info may help me avoid any pitfalls along the way.
Thanks for any comments/advice, from any of you knowledgeable lads.

Cheers, Noel.

[nsubulysses]

Glad I saw this because I haven't even been able to read the forum much lately.

I basically got what I wanted.

JJ 6CA7 seems pretty rugged in my experience. My amp has about 585V B+. Screen supply has 1K 20W resistor to drop the B+ voltage down and 1K screen grid resistors so it is probably around 560V or so and it sags down to around 400V when the amp is cranked.

I have had two JJ 6CA7 fail but it's only when they were brand new. One I got did not work at all when I plugged it in and another I got a few days ago would intermittently blow the HT fuse if you tapped on it. Once the tubes are in and work decent they seem to stand up to the high voltage pretty well.

Other than that 6CA7 seems pretty reliable. My amp has killed two JJ 6L6GC but it only did it when the amp was positioned upside down with the tubes hanging underneath the chassis so I could play it and do tweaks. When I finished my tweaks I flipped the amp back over to put it in the normal position with chassis at bottom and tubes and transformers on top. When I turn the amp back on in this position after heating the tubes in the upside down position it killed two 6L6GC on two different occasions. It seems to run NOS 6L6 with no problem but the new ones sometimes fail.

When I originally made the amp I used a choke rather than 1K 20W dropping resistor. The amp did about 160W with the unsagging screen supply. But it wore out the JJ 6CA7 fast. Probably after 20 hours of playing it it would only put out 120W or so. When I crudely measured screen dissipation by measuring voltage drop over 1K screen grid resistor I was calculating about 17W screen dissipation. Max screen dissipation on the datasheet is 8W, but they did not die.........

with 1K 20W resistor in place of choke I could get the amp to do about 144W and be seemingly reliable and not that harsh on the tubes. THis amp basically only gets played full volume.