Phase Inverter and Reverb Phase

[Daver]

If I want to use the second input to the phase inverter for a reverb return, should that signal be in phase or out of phase with the dry signal? This would be a long tailed splitter type with no negative feedback like a Vox AC-30 or similar. I can make arguments with myself for both ways. Thanks!
Dave

[Arthur B.]

Originally Posted by Daver

If I want to use the second input to the phase inverter for a reverb return, should that signal be in phase or out of phase with the dry signal? This would be a long tailed splitter type with no negative feedback like a Vox AC-30 or similar. I can make arguments with myself for both ways. Thanks!
Dave

I don't think it really matters, since you're not using NFB.

[Bruce / Mission Amps]

I think Arthur is right here but make sure there is still a good AC path to ground for the grounded grid side and your actual reverb signal is pretty strong as the grounded grid side doesn't have a lot of gain.
Check out any Matchless amp with reverb recovery sent to the grounded grid side of the PI.

Bruce

[Merlinb]

It should be out of phase with the dry signal if you want the greatest amplitude output from the LTP.

I'm not sure why Bruce said you need a good AC path to ground, that would leak the reverb signal to ground obviously! I think Bruce has misunderstood what you're trying to do. And both sides of the LTP must have the same gain if you're using both inputs, so make the anode resistors the same.

[Daver]

OK, some slightly differing opinions. Not a bad thing. I was thinking that the two signals should be out of phase. The PI is a differential amp and amplifies the differences in the signals, right? Knowing the signal strengths would never be exact nor the tube gains, I could see how in phase could work too. The Chieftain is the only Matchless schematic I've seen that includes reverb. That one has the reverb signal out of phase with the dry. They used both halves of a 12AX7 for reverb recovery. The first stage has a low 22K plate resistor and an unbypassed cathode resistor. It seems like they were going for phase inversion and not gain with that setup.
I've wondered too about the 18 Watt Marshalls and clones that send each channel into a different PI input. Can you jumper the channels and not have some of the signal get cancelled? Thanks.
Dave

[Mark Lavelle]

I'm a little late to this party, but...

<flame on>
No no no no NO! For any practical purposes, a reverb return doesn't have a phase relationship with the clean signal!
<flame off>

What is a reverb tank but a mechanical way of "bouncing" your signal around? Do you even have a chance of controlling the phase of the reflections in the springs? Don't kid yourself!

Phase cancellation only occurs if two signals are substantially the same. Even if you could make any meaningful statements or predictions about the phase of the individual reflections in the tank, how would you decide which reflections should be cancelled?