[kenreau]

I have a pair of Cayin 860 tube amp monoblocks.  I have been working my way through them upgrading various caps and resistors and am finally getting to the feedback resistor loop. 

I've read best sound is often achieved with a minimal amount feedback.  The existing feedback resistor is a 1W 8.2K part.  I imagine the value was chosen as a general compromise of a number of issues.  I am interested in optimizing this feedback resistor and was wondering what is the best route to determine this?  Is there a way to make quick work of picking the optimal value?  I don't have time to trial and error a half dozen choices and listen critically to each. 

Is the feedback resistor value in the amp influenced by what goes into it (preamp), or what it powers (the speakers)?

Maybe insert a variable pot of some kind and dial it around to see what it sounds like?

I was also going to try just bypassing the resistor and go without feedback.

Any recommendations would be appreciated.

Thanks,
Kenreau

[JoshK]

This is a loaded question.  First, one must have the fully fleshed out schematic in order to opine what any compromises might be, or what kind of circuit we are even dealing with, or one must have the ability and tools to do various measurements to determine what the best is.

You mentioned you didn't have the time to experiment empirically, so then the only way to "optimize" is to take a bunch of measurements with different values.  Otherwise there is no optimization at all, neither empirical nor measured.  You can change the value of this FB resistor and see what happens, whether you like it or not, but that isn't an optimization.

If we have a schematic, then one can makes some edjumacated guesses about what might the appropriate amount of FB be based on the experiences of others.  Many/most of the commercial offerings follow a handful of 50's designs (williamson, mullard, etc) and based on the tubes chosen there is a lot of data already out there, so there is safe guesses, which might be good enough for you. 

I guess what I am trying to say, is there is a difference between real measured/analyzed optimizing of a circuit and ad hoc changing based on rules of thumb. 

[kenreau]

Thanks Josh. 

What measurements were you refering that could be made?

I do have a pdf of the schematic.  I hesitate to post it here, but could pm to you if you may have some additional "feedback" 

Thanks,
Ken

[JoshK]

Its wise that you not post it here if it isn't in the public domain.  You can email it to me, since my interest in it is only academic. 

I cannot profess to be the authority on what and how to measure, only I know that is what needs to be done.  I'd want to test distortion and measure or compute the output impedance (damping factor) as it will predictably rise with less FB.  However, if you can also measure the spectral distribution of distortion it may be telling. 

Changing the Feedback resistor changes the gain, distortion, and dampening factor all at once.  It (lowering the fb ratio) will also make the amp more sensitive to tube imbalances (unmatching) and tube aging.  This is another reason why commercial amps generally use prodigious amounts of FB.

 Raising the FB ratio is the only thing that will put amp stability into question, lowering it will not, it will just raise distortion and lower dampening, but maybe not enough to outweigh other positive attributes.  One can always suck it and see.  Likely, your best bet is to examine your circuit and then try out what some others have determined to sound good or a range around that and see if you also like it.