Class B Biasing issues!

[Thorshavn]

Hello folks, I am building an amp based on D Self's blameless. One channel works perfectly, while I think I am having biasing issues on the other. The only difference I can tell between the two channels is that I switched one of them (the one that works) to be an inverting amp.

Both channels work, per se, but the one that really 'works' doesn't draw much current at all, and the one I am having problems with draws very little current with no load, I add a load and it is fine under little signal, then guzzles the current if any volume is applied. It still sounds fine, however, its just really warm. It's like it starts out in class B, then with volume it goes hard into A/B.

Adjusting the potentiometer gives me basically two settings, fuzzy static with massive current draw, and the hard A/B state I described above. On adjustment, the channel that works gives me the static with massive current draw, and the working state with little current draw.

I've gone through and replaced almost all of the small signal transistors, the only ones I haven't yet replaced are the diff. amp section. I also tested the output transistors. The Vbias transistor is mounted to a driver transistor on it's own heatsink.

Please give your thoughts, thanks!

[Richard C]

I don't understand the ''fuzzy static" bit but often when an amp suddenly starts drawing current for no apparent reason it's a sign of high-frequency oscillation.

[Thorshavn]

Yeah, in simulation I get a clean signal no matter how I set the potentiometer. So what do you think would cause the oscillation?

[Jan Dupont]

Have you checked the B C E Voltages on the bias transistor ???

[Richard C]

Unfortunately, if the amp is oscillating it's impossible to say what the cause might be. It could be a component, physical layout, a soldered joint, anything and because the amp is enclosed in a feedback loop isolating the problem is going to be difficult. An oscilloscope will be useful here.

I haven't got a schematic for this particular amp but some of Self's designs use CFP output stages and these can be prone to oscillation. If this is the case a possible cure is to fit a small cap between the base amd collector of the smaller transistor of the pair. Use the smallest value that's necessary to stop the oscillation it should be no more than a few tens of pF.

[Mark25]

Thorshavn,
i had a similar prob with my ESP p3a, (also CFP)it turned out to be a faulty miller cap around the VAS. Should have been 100pF, but when held, the temperature of my hand made the capacitance fall to zero on my meter.
Good luck and let us know how it sounds
Mark

[tennisballg]

I tried the tiny caps from B-C and it seems to work great now. I don't really have a spectrum or distortion analyzer, just an O scope, so I can't really tell how this is effecting my noise or frequency response, what do you guys think it will do? it sounds good so far, so I am not really worried about it though...

[phase_accurate]

Hello folks, I am building an amp based on D Self's blameless.

Does that mean you use exactly his design or did you do something similar (e.g. using other transistor types, different component values, different layout ?).

Does the working one also show strange behaviour (i.e. oscillation) if used non-inverted ?

Regards

Charles

[Thorshavn]

Well, actually I read one of his books, and made my own design based on the same principles. I don't feel comfortable saying it's my own, though, since it's very similar to what is outlined in the book.

Both work great now, they bridge great too. I never did try the inverting one as a non- inverting, simply because I already burned the PCB a specific way. the circuits are the same, though, except for the gain resistors (of course): a 470 ohm R1 and an 11k R2 on the non- inverting and a 450 ohm R1 and an 11k R2 on the inverting, so there is a .04 difference in gain between the two, theoretically. Probably more, but I did use 1% resistors.

Other than that, the layout is very similar for the two, they're actually mirror images on the PCB. Not exactly the same, though, so I wonder if I have crossed a track somewhere on my double sided PCB that causes the one to oscillate.