DC Servo - Direct Coupling Applications?

[exclamationmark]

Hello, it's my first post here and I come from the mystical land of guitar amplifiers. I've basically been dabbling with DC coupling within pre-amp stages, with reasonable success, except that voltage drift is starting to play on my mind.

I have multiple pre-amp stages set up in accordance to my dodgy schematic. The two resistors are fairly large (in the region of 500k) and basically form a voltage divider that allows me to set the quiescent grid voltage (and also eat half my gain ) from the negative supply . All is working well, except that the voltage supplies to each stage (each stage is biased slightly differently as I'm mixing pentodes/triodes) basically consists of a trimpot across a string of LED's to provide the variable negative supply, and is not resistant to drifting voltages in any way. Unless of course, whipping out the multimeter and a screwdriver every so often to correct the bias is considered 'resistant'.

So delving into the world of audio, I see they have this mysterious device called a DC-Servo, which alleviates this problem in many solid state devices. Basically, I want the voltage at the node in the red circle to be steady at DC (or some ridiculously low frequency), thus setting the bias accurately independant of drift. Is the DC Servo the correct device to achieve my goals? Or am I looking in completely the wrong direction (I was actually going to attempt to do this with a microcontroller, but a couple of op-amps definately sounds like the easier option ). Before anyone asks, my B+ is only 48v and -48v so the higher voltage op-amps should suffice.

Any advice would be appreciated!

[SY]

Yes, you would use a DC servo for that. In your sketch, the servo would have to be applied to the first tube.

[DavesNotHere]

I don't know what power supply design your using, but a guitar amp power supply B+ is not really that stiff and they are design to sag when signal is applied.
here is a circuit you might be interested in:

[exclamationmark]

Actually this particular project is a rack mount pre-amp powered off a 48V SMPS (laptop supply) going through a CLC filter. I'm using some oddball radio tubes (1LC5, LG5 and some triode variants) that were designed for 45V operation.

This is looking like more trouble than I thought, simply because op-amps with the needed voltage range are hard to find and reasonably expensive. Then there are logistics of hooking up multiple DC-Coupled stages. The pre-amp is actually working fine as it is, and I haven't noticed any significant change in bias since it was built (though to be fair, it was only about a month ago). I'll probably continue to monitor the performance and see if I need to actually re-visit DC servoing.

[SY]

If you do, don't worry about high voltage op amps. There's many easy ways around that.

[DavesNotHere]

the laptop power supply is probably the problem. they are made for charging a battery and have a tendency to fold back on current demands.