Bias too hot in Traynor CS100H, how to adjust overall bias if pot won't work?

[Stratjeep]

I have a Traynor Custom Special 100 H here, the EL34's where reading up to 170 ma in a badly matched set. New JJ's come in around 70-90ma depending on the set I put in.

I changed resistor values of R182 and R49 and it did nothing really, it did change the voltage until standby was turned off and then it went high again. There is extra circuitry tagged on to the bias pot and I honestly don't understand what this is doing. I almost scrapped the bias circuit and put in something I'm familiar with but not knowing what the extra stuff does I don't want to remove it. Supposedly it will adjust for worn tubes?

Any help would be greatly appreciated !

Schematic is here http://www.traynoramps.com/downloads.../smycs100h.pdf


Cheers,

David

[R.G.]

The bias supply generates a large-ish minus bias voltage through R174 (10K) and R181 (4.7k) to the grids. This is pulled up to +15V through R49 (15K) and the bias pot. The transistor is connected to voltage divider from the screen voltage through R257, R258, and R259. If the screen supply gets too low, the transistor turns on and pulls the bias up towards +15V, turning the output tubes on hotter. Possibly this is some kind of compensation for bias shift at high drives. I'd have to think about this some more to really figure out why they did that.

But you can do a couple of things. I personally would:
- remove the output tubes and see if I could still move the grid voltages around by diddling the bias pot. If I could not get -30 to -60/70V on the grid socket pins, I would dig through and find out why. The grids should all be at the same voltage under this setup, and vary together.
- If that worked, I'd remove Q7 and test it for opens, shorts, etc. Then I'd say to myself "What the heck, I have this thing out, I'll just put a new one back in, whatever the test says. If it tested good, OK fine; same amount of work" I'd do the same thing to the diode in series with its base. A bum transistor here would under bias the outputs no matter what the bias pot says, which is the problem you're experiencing.
- If the transistor and diode were good, I'd look at the voltages on the screen divider of R257, R258, and R259. If that was bad, the transistor could still be OK, but being told to turn on all the time. An open on the path to the screen voltages would do the same thing. Or a too-low screen voltage for other reasons.

[Stratjeep]

The bias supply generates a large-ish minus bias voltage through R174 (10K) and R181 (4.7k) to the grids. This is pulled up to +15V through R49 (15K) and the bias pot. The transistor is connected to voltage divider from the screen voltage through R257, R258, and R259. If the screen supply gets too low, the transistor turns on and pulls the bias up towards +15V, turning the output tubes on hotter. Possibly this is some kind of compensation for bias shift at high drives. I'd have to think about this some more to really figure out why they did that.

But you can do a couple of things. I personally would:
- remove the output tubes and see if I could still move the grid voltages around by diddling the bias pot. If I could not get -30 to -60/70V on the grid socket pins, I would dig through and find out why. The grids should all be at the same voltage under this setup, and vary together.
- If that worked, I'd remove Q7 and test it for opens, shorts, etc. Then I'd say to myself "What the heck, I have this thing out, I'll just put a new one back in, whatever the test says. If it tested good, OK fine; same amount of work" I'd do the same thing to the diode in series with its base. A bum transistor here would under bias the outputs no matter what the bias pot says, which is the problem you're experiencing.
- If the transistor and diode were good, I'd look at the voltages on the screen divider of R257, R258, and R259. If that was bad, the transistor could still be OK, but being told to turn on all the time. An open on the path to the screen voltages would do the same thing. Or a too-low screen voltage for other reasons.

Thanks RG this really helps and gives some where to go. I'll let you know how it works out.

Appreciatte it!

David

[Stratjeep]

Replaced the transistor and diode and all is well.

Thanks!

David

[g1]

Good to hear you got it sorted.
Just a couple of comments:
a) Ads for that model mention "auto-matching" for the output tubes so the transistor has to do with that.
b) The schematic appears to have a typo, the 220mV listed for the bias should be 22mV. This seems quite cold but it is what is shown on the schematic for the H2 version. 220mV would make pretty short work of the output tubes as far as I can see.