Negative Feedback Question

[cryptozoo]

This may seem like a crazy question -- I understand the need to keep capacitance out of an amp's output stage in order to avoid an impedance mismatch at the speaker, but what is the danger of running something like a diode there? I've gotten some fun fuzz tones by running a diode in parallel to the NFB resistor, just wondering if this is going to damage the OT. Thanks!

[Mick Bailey]

What amp?

[Chuck H]

It depends on the circuit. Can you provide a drawing. diagram or schematic??? Adding a diode to an NFB loop COULD send watts to circuit not designed to handle it. I'd need to see what's actually being done before endorsing it.

[cryptozoo]

This one is kind of a bastardized version of the Fender 5C3 schematic: http://www.davidsonamp.com/sf/images/deluxe5c3.gif

[Mick Bailey]

Any voltage above the forward voltage drop of your diode for one half of the output waveform is going to be applied to the grid of the connected output tube. Would make an interesting scope pic to see what's going on.

[cryptozoo]

Interesting... so it's potentially taxing to the power tubes? Doesn't seem like it would generate more voltage, and in fact it's noticable quieter when engaged. I guess that's because half the waveform is getting lopped off!

[Chuck H]

I don't see a diode in the above schematic so I'm still not clear about the circuit. But just speculating, a diode in the NFB shouldn't tax the power tubes because the relative impedance of the circuit is much higher than the speaker load. It may do funny things to the speaker damping due to NFB. And THAT may change the ideal load impedance. But I don't think there is any danger to the tubes.

[Mick Bailey]

Interesting... so it's potentially taxing to the power tubes? Doesn't seem like it would generate more voltage, and in fact it's noticable quieter when engaged. I guess that's because half the waveform is getting lopped off!

It's out of phase to the grid voltage, so will reduce the drive to the output much more than the standard setup (on one half of the waveform). You can work out what peak voltage is getting fed back, from the amp's output and speaker impedance. You'll not only get asymmetric distortion, but a slight shift in frequency response due to increased NFB.

I would imagine that the peaks on one half of the output waveform are severely attenuated and the 0.6v or so forward voltage drop will introduce a step in the leading and trailing edge. Bit like crossover distortion. You could make it even (...worse!) with anti-parallel diodes.

I'm in agreement that the impedance of the circuit is unlikely to have any detrimental effect on the OT.

Have you got a scope and signal generator? - I'd like to see a pic of the output.