Thoughts on using a Long Tail Pair PI as a Mixer

[mumford]

I'm thinking about my next build utilizing two different preamp sections with separate inputs into a long tail pair PI, something like an AC30 bright channel pre for one and something with more gain for the other.

I know using a LTPPI for a mixer is pretty common, but I don't see many people doing this kind of build, e.g., an AB763 clean preamp and a JCM800 dirty preamp in a single build.

Other than having one of the preamps out of phase with the other in case I want to use them simultaneously, does anyone have any things I need to consider/pitfalls in terms of design? I know enough to design the LTPPI, I'm just wondering if the reason people don't do these kind of builds is because X and Y cause Z to happen.

[Milkmansound]

you can think of it like a Fender amp with 2 channels, like a Twin. Maybe look at the schematic for an idea. I think the 2 channels can be fed into the PI via 220k resistors. The PI and the output don't care what the preamp is!

[John_P_WI]

I've done this before to Fenders, if you use it as a mixer the way you suggest, won't you will lose the negative feedback to the power amp? Depending on the tone / feel you want this can be a good or bad thing......

[Milkmansound]

you can still have the feedback at the output. The input of the PI on a Twin, for example is a .001 cap. Send each channel into that cap via a 220k resistor (good place for a master volume control too). The output feedback hooks in somewhere on the other side of that cap so no worries.

I have done this to a Twin before and it worked.

[John_P_WI]

I did it to a SF bassman converted to cathode biased output tubes, actually was a pretty cool blues amp - loud as hell though. If I recall I defeated the NFB.....

[jjman]

One reason to use your approach would be if you wanted to use both channels at the same time, but they were out of phase. I think it may also provide stronger signal potential entering the PI compared to the 2x220k junction approach on one side.

[Reeltarded]

oh please keep talking.. More potential at the PI? Why would I?

Imma sponge. Throw me inna bucket.

[Firestorm]

Other than having one of the preamps out of phase with the other in case I want to use them simultaneously, does anyone have any things I need to consider/pitfalls in terms of design? I know enough to design the LTPPI, I'm just wondering if the reason people don't do these kind of builds is because X and Y cause Z to happen.

I think there are a few challenges to doing this: you need to make the inverting and non-inverting gains as close to identical as you can, so you need a comparatively large tail. This will reduce the output swing and make the inverter/mixer easier to clip.

There may also be "unpredictable" volume behavior when using both inputs versus only one: when one input is active, its signal works against the 0VAC "signal" on the opposite grid, so the differential between them varies with the amplitude of the input. But with both active, being fed similar, but out of phase signals, the differential will vary by 2X the amplitude, almost as if you had an expander in the circuit.

Of course it can be done, but getting it to behave the way you want is the trick. You also do lose the conventional way to apply feedback to the power amp, so you have to design for no feedback or apply it locally.

[mumford]

you need to make the inverting and non-inverting gains as close to identical as you can, so you need a comparatively large tail. This will reduce the output swing and make the inverter/mixer easier to clip.

There may also be "unpredictable" volume behavior when using both inputs versus only one: when one input is active, its signal works against the 0VAC "signal" on the opposite grid, so the differential between them varies with the amplitude of the input. But with both active, being fed similar, but out of phase signals, the differential will vary by 2X the amplitude, almost as if you had an expander in the circuit.

Now we're getting somewhere. I hadn't thought as far as the input gains.
My thought at this point is to use a vox ac30 bright channel pre and something dirtier.

The vox at least doesn't use feedback, I guess I'll have to keep that in mind for the dirty channel as well. Feedback is a topic I feel like I have very little understanding of--I've really only read the chapter in the merlin pre book a couple of times.

[jjman]

18 Watt with trem shows this approach. But the channels are in phase so feeding them the same signal would drive the PI in an out of phase manner. Anyone ever use a phase switch on a splitter box and drive both channels of their 18 Watt?

[mumford]

18 Watt with trem shows this approach. But the channels are in phase so feeding them the same signal would drive the PI in an out of phase manner. Anyone ever use a phase switch on a splitter box and drive both channels of their 18 Watt?

So if you use the trem, you hear reverse trem because of phase cancellation!

[Andy Le Blanc]

It becomes a differential input.

I think you can still apply FB in the usual way to the tail

You can can also play with the 1M grid resistors to control volume and tone
use the local feedback of the inverter, or even simplify the inverter to just a tail

[Firestorm]

I think you can still apply FB in the usual way to the tail

Careful with that. Anything input at the tail will have the same phase on both sides, so it will be negative feedback for one input and positive feedback for the other.

[jaysg]

hmm...haven't got one, but the next time I've got a BF Fender with Reverb on the bench, I'm going to convert it and see how it sounds. I think you take the 82K and bump it to 100K for the mixer approach.