How Does This Weird Reverb Arrangment Work?! (Tube Amp)

[Paul Marossy]

I've been collecting information on Gregory amps since I picked up a Mark X about six years ago (http://www.diyguitarist.com/GuitarAmps/GregoryAmps.htm). An amp tech sketched up schematics for the preamp (tube) and reverb driver section (solid state) for the "Reverb 1500" model and contributed it to my website. Anyway, I was looking at this reverb circuit which appears to have been used in the vast majority of their amps which had reverb. I am really baffled about how this even works - the reverb circuit is powered by the cathode of the first triode stage and uses PNP transistors. And it uses that same cathode to carry the reverb effected signal into the signal path. This is very weird stuff to me. Can anyone explain how the heck this works?! I am confused by how it gets power to even work and how the cathode can carry a signal while powering something else. Never seen either of these things in an amp before.

Reverb schematic - http://www.diyguitarist.com/Schematics/1500ReverbCircuit.pdf
Preamp schematic - http://www.diyguitarist.com/Schematics/1500PreampCircuit.pdf
The amp tech's analysis of the amp - http://www.diyguitarist.com/Schematics/1500CircuitAnalysis.pdf

I also was wondering if this driver circuit would be any good for a DIY spring reverb project. It seems like it wouldn't, but I thought I'd throw it out there...

[igerup]

Seems you have read the schematic wrong. The reverb gets the signal from the "To Reverb Drive" junction between the two 68k resistors before V1. Also the 7189 pentode only has one cathode and that's where the output from the reverb returns to the preamp. A slightly different method than most tube amps but if it worked for Gregory Amps then I see no reason why it shouldn't for you.

[Paul Marossy]

Seems you have read the schematic wrong. The reverb gets the signal from the "To Reverb Drive" junction between the two 68k resistors before V1. Also the 7189 pentode only has one cathode and that's where the output from the reverb returns to the preamp.

No, I didn't read it wrong. I know where the input to the reverb driver circuit is coming from. The reverb schematic is wrong as the preamp schematic shows the reverb return going to the cathode of the first triode stage, V1A, not the cathode of the 7189 power tubes.

Anyway, I don't understand that cathode reverb return part and how the circuit is also being powered by that same cathode. Is the circuit current driven and at a negative voltage?

[CynicalMan]

IMO....

The reverb driver's power comes from the cathodes of the 7189 (the power tubes, similar to el84). The power is filtered (5k6 resistor and 500u cap), then powers a transistor amplifier and buffer. The mystery component (probably a transistor) drives the reverb spring tank, which outputs to the cathode of V1A, a 12AX7.

Edit: do you have the power amp schematic?

[Paul Marossy]

One of those schematics is wrong. It's either being powered by the 7189s or the cathode of V1A. I don't know for sure, I don't have the amp and neither does the guy who drew the schematic.

[CynicalMan]

Why do you think one of them is wrong? The schematics say that the cathodes of V3 and V4 power the driver circuitry, which drives the driver transducer in the spring reverb tank. The output transducer in the reverb tank outputs to the cathode of V1A. The schematics aren't contradictory.

Edit: 600th post 

[Paul Marossy]

Why do you think one of them is wrong? The schematics say that the cathodes of V3 and V4 power the driver circuitry, which drives the driver transducer in the spring reverb tank. The output transducer in the reverb tank outputs to the cathode of V1A. The schematics aren't contradictory.

Oh, yeah, I see what you are saying. That makes more sense actually. Duh! Then the preamp schematic has omitted where exactly that comes from on the cathode(s) of the 7189s. Minor detail.

OK, I get that part. But how can the signal return on the cathode of V1A while the plate is amplifying the dry input signal? It's almost kind of like a split load phase inverter setup. I've never seen anything like that in a tube preamp section before. It's just a little weird...

[CynicalMan]

That's the weird bit. I guess it would get added because the tube amplifies Vgk, but I've never seen mixing done like that. Also why are there two 10k pots for reverb level? And wouldn't turning the reverb down all the way remove the bias on V1A and cause clipping? I might try simulating it.

[Paul Marossy]

That's the weird bit. I guess it would get added because the tube amplifies Vgk, but I've never seen mixing done like that. Also why are there two 10k pots for reverb level? And wouldn't turning the reverb down all the way remove the bias on V1A and cause clipping? I might try simulating it.

Yeah, it's a weird creature, isn't it? That dual pot is also kind of strange. I was also wondering how that whole arrangement affects the bias on V1A. One is obviously the reverb output control, but I wonder why they have a 10K pot on that cathode. Maybe they work together in parallel somehow? The guy that drew the schematic (tech with 20 yrs of experience) says that there is between 8 and 500 ohms DC resistance on the reverb pot as you vary the setting.

[gmoon]

That's different. Using the cathode bias voltage on the power tubes to power the revb driver transistors is really cool. Would be fun to add a "Mesa Booger" style equalizer like that...

I wonder about the amp tech's analysis. A 10 Meg grid resistor, along with a grounded cathode on V1A sounds like "contact bias". Although I see a plate resistor of 3 Meg (1.5M + 1.5M), not 1.5M as the tech writes.  I've seen as large as 1M, but that's really unusual. 3M is nuts in any case. And I'm not sure what a variable cathode resistor would do to contact bias, though.

The amp tech also writes that the V1A stage has a high output impedance (true). But then he writes:

Of course, this means the tube can hardly drive anything, so they have to up the impedance of the NEXT stage by
adding a 560k resistor, making the input impedance of V1B about 900k Ohms.

And that's flat wrong. The grid resistor of V1B is 330K; if the designers wanted a higher input impedance for that stage they'd have used a 1M or higher. As is, the 560K just forms a voltage divider with that 330K Rg. This drops the output voltage of V1A, rather than raising the input impedance of V1B...

Paul, do you have one of these? It would be cool to verify a few things. Like the V1A cathode resistor...it would be very different if the POT ends were wired as the cathode resistor, and the wiper injected the mix.

[Paul Marossy]

The grid resistor of V1B is 330K; if the designers wanted a higher input impedance for that stage they'd have used a 1M or higher. As is, the 560K just forms a voltage divider with that 330K Rg. This drops the output voltage of V1A, rather than raising the input impedance of V1B...

That sounds reasonable.

Paul, do you have one of these? It would be cool to verify a few things. Like the V1A cathode resistor...it would be very different if the POT ends were wired as the cathode resistor, and the wiper injected the mix.

I don't have one. Otherwise, I probably would have drawn up that schematic myself. I'll have to take that amp tech's word on the schematic as far as that reverb control is concerned. It really is a very odd arrangement if it is correct. It seems to be consistent with other whacky things I have seen in other Gregory amps, though. There's some very unusual things shown on most of the schematics I have been able to round up so far.