Superfuzz waveform problem - Pictures attached

[GaryM]

I built a copy of the Univox Superfuzz from This schematic. All voltages look good and I used a matched pair for the octave section. All transistors are 2N4400. The problem I'm seeing is the waveform on the differential driver,  transistor Q3.  Do these look normal?

Waveform on emitter of Q3:



Waveform on collector of Q3:

[digi2t]

Good thread on this;

http://www.diystompboxes.com/smfforum/index.php?topic=94633.0

[PRR]

Those waves may be correct. Q4Q5 become a low impedance above about 20mV-50mV. The emitter will pull that, the collector won't. By the time this happens, the pair is totally swamped, signal is too high for proper operation.

What is the Q4Q5 collector waveform with small and medium drive? Should be full-wave rectified with a soft curve where a true rectifier has a deep notch. This will be on the + side of the wave in this case.

[Mark Hammer]

In an issue of Electronotes from the late 70's, Bernie Hutchins notes that phase-splitters nominally require equal-value emitter and collector resistors, but actually perform better with a slightly smaller value in one of them.  I wish to heck I could remember whether he was talking about the emitter or collector resistor, and knew why he made that recommendation. 

I mention it here because we tend to work from  nominal values on schematics, often using 5% resistors (which Univox also did).  Maybe some diligent member with a scope could insert a trimpot in the relevant location and see if the phenomenon Gary notes has any relationship to it, or is entirely unrelated.

[PRR]

The posted waveforms show a REAL hit above a certain signal level. This isn't any small-percent non-optimization. This is overdrive.

Yes, cathodynes make horrible waves when overloaded. One "fix" is: don't overload. In this case I think the next stage is totally hammered before the cathodyne gets ugly.