Even More Stupid Questions about MOSFET Sidechain Driver

[PRR]

> LF442 is happily running on just about -30V on the -V terminal and ground on the +V terminal.

No, it needs to sit near ground when idle, and the op-amp won't pull all the way up to the V+ pin, so the V+ pin needs to be around +3V or +4V. V- can't be more than 36V away from that, so about -30V is fine. -25V might be fine too, because I don't think the 6BC8 ever wants more than about -20V bias to turn hard-off.

1G release resistor seems absurd. Maybe you have leakage somewhere?

 

[Swedish Chef]

[quote author="PRR"]>No, it needs to sit near ground when idle, and the op-amp won't pull all the way up to the V+ pin, so the V+ pin needs to be around +3V or +4V. ? [/quote]

Ok I'll give that a try. Does that have something to do with the odd osc then? I am fascinated to know what is physically going on here!!!

[quote author="PRR"]>I don't think the 6BC8 ever wants more than about -20V bias to turn hard-off. [/quote]

Yep. Off hand I think if it gets beyond about 16V you're almost cutting off, but I've been aiming at around 10V or a little over.

[quote author="PRR"]>1G release resistor seems absurd. Maybe you have leakage somewhere?[/quote]

Yep sorry I meant the resistors in the buffer are 1G. The release is 20M at the moment.

Cheers for the continued support mate! :thumb:

Steve

p.s. do you remember talking me through that 5 valve radio guitar amp thing?
I did some rewiring of it the other day and got the hum down almost to zero with various extra grounds and a big filter cap.
sounds great through big speaker cabs!!!! :green:
Oh and to soak it, I listened to the radio through it all day...Lovely :grin:

 

[Swedish Chef]

Well I banged my head against this thing all day and got exactly nowhere.
Whatever i do I seem to end up with a low (between about 10Hz on the smallest cap to, at a guess 0.4Hz at the highest cap value) frequency oscillation which is manifest in thumps of about -30V on the grids.
I have reverted to the original UA s/c which stopped the thumping but took me back to an attack time of around 10ms. Then I tried to graft on a 670 approximation but with all of the res x10 and caps /10. Ha ha ha ha comes to mind.
What exactly is it that is causing this? Obviously it is a feedback thing, but i don't understand the mechanics of it... :?

chef

 

[gyraf]

My guess: poorly balanced input pair. When you get over threshold, it reduces gain and makes a thump in audio (probably sub-audio, if you haven't spotted it yet), that in turn triggers the threshold again..

Try balancing out the input pair by driving both grids simultaneously with a relatively high level (as high as your CV) low-frequency source, and trim out any resulting signal over the anodes.

Is this a project with some long-term commercial scope, or is it a one-off for yourself?

Jakob E.

 

[Swedish Chef]

Hi Jakob!
Yeah i have been thinking that it may be the i/p pair again, but have trimmed them as far as I can, and as I asked Chelmer to match them before they sent them to me I was hoping that they would be pretty close.
I will have a bit more of a slog at it today, and perhaps try a larger proportion of trim pot/ load res on the anodes.
This will be a product if I can ever get it working. It has been working fairly well a couple of times and then i have torn it to bits to try and improve it more. I'll give it another day and then get on with something else :sad:

steve