Long tailed pair phase inverter issues

[Kay Pirinha]

Hi,

most guitar amplifiers with push-pull output stages feature Schmitt type phase inverters showing details I don't really understand. Let's have a look at the one in the following HIWATT schematic:



There are two resistors (22k&Omega and 2,2 KΩ respectively) in the common cathode path of V4. Bias is provided by the proceeding cathode follower. Other ones are biased by an additional resistor directly between the cathodes and the next resistor, but this doesn't matter so far. Negative feedback, coming form the OPT, is injected via a 10KΩ resistor into the grid of the second PI triode and, what is disturbing me, into the cathode tail. This means common mode feedback to both triodes!

My doubts, or questions, are the following:

1. What is the use of the common mode NFB? Would stability, or sound quality, be affected if NFB would be applied to the second triode grid only?

2. If we'd agree on NFB to the grid only, would it be a good advice to replace the tail resistor by a CCS? Would this affect sound quality?

3. If we'd agree to a CCS, would it be a good advice to place the CCS to the negatice bias voltage source instead of ground?

4. Do I think right that using a CCS instead of a tail resistor does no longer call for different anode resistors? Or could it be possible that different anode resistors were provided to achieve some certain sound effect?

Thanks in advice for your efforts!

[DF96]

Feedback is applied to the first triode only, via its cathode. The second triode has the same signal applied to grid and cathode, so no effect (apart from a small amount due to anode resistance).

The unequal anode resistors are because of the low resistance tail. With a CCS you would have equal resistors. However, you can't just add a CCS to this circuit as you would have to find a different way to put the feedback in.

[Kay Pirinha]

Thanks! In my mind I'm tinkering with putting the feedback solely to the grid of the second triode, which *imho* acts as the inverting input of the whole amplifier. The CCS would be placed to the negative voltage rail that also provides bias for the finals. Both LTP grids' DC potential would be equal to ground, and both anode resistors would be 100 kΩ or so. Are these good ideas?

[DF96]

I'm not a guitar man, so not the best person to advise you!

[es345]

The CCS would be placed to the negative voltage rail that also provides bias for the finals. Both LTP grids' DC potential would be equal to ground, and both anode resistors would be 100 kΩ or so. Are these good ideas?

This is somewhat comparable to your idea. I have implemented that in 200W KT120 design and is working quite well.

[Merlinb]

Originally Posted by Kay Pirinha

Hi,
1. What is the use of the common mode NFB?

It is a relic of Fender circuits, which were arranged that way purely for layout convenience. Lot's of other manufacturers copied from Fender. The (very small) amount of common mode feedback has no useful effect here.

Would stability, or sound quality, be affected if NFB would be applied to the second triode grid only?

No, it's unlikely you would notice a difference.

2. If we'd agree on NFB to the grid only, would it be a good advice to replace the tail resistor by a CCS? Would this affect sound quality?

No! Guitar amps are often overdriven, and a CCS will NOT respond nicely to overdrive, and will cut in and out, oscillate, or do other strange things. Sounds horrible!

4. Do I think right that using a CCS instead of a tail resistor does no longer call for different anode resistors?

Technically yes.

Or could it be possible that different anode resistors were provided to achieve some certain sound effect?

It was probably not the designer's intention, but it can make a tonal difference sometimes.

[Kay Pirinha]

Thanks a lot for your opinions!

But I realize that I'd better have put my questions to the Instruments and Amps section of this board. I beg your pardon!