Adding *clean* gain stage

[EFK]

One of my old Supros has a channel which uses only half of a triode, into a volume and tone, then right into a paraphase PI. So, I have a half of a triode just sitting here. I would like to get just a bit more volume out of it without really changing the overall sound, like a volume going to eleven or twelve, so I'm unsure how to proceed. I did receive a suggestion of paralleling the triode and halving the plate/cathode resistors, but I've never done this and I've read a number of comments which seem to indicate that it doesn't really add much save reduced noise. The other option would be to take the signal off the tone and feed it back to the other half for a 2nd gain stage before the PI, but I don;t want to add any more distortion, just a little more volume. Woudl this be a good option if I stuck with something like a 82 or 100K plate resistor and a 2.7K or even higher - maybe 4 or 5K? - unbypassed cathode resistor? Any ideas? Everything in the preamp is a rats nest on tag strips, and it is tough to get in there and do neat work so I'd like to minimize the amount of soldering I have to do. Any information appreciated! Thanks.

[NorCalTuna]

how about a cathode follower to drive to the tone control? That could reduce insertion loss and make the tone more effective- and be easy to implement in your ptp setup.

[Chuck H]

+++ But the difference in impedance can change the frequency knee of any coupling caps so this needs to be accounted for. Also, the use of a cathode follower would not change the phase of the amp as it is. Some people will say it doesn't matter but I think it does. If you simply add a gain stage the amp will be 180* out of ohase from the way it plays now. This, to my ears, seems to change the harmonic structure for better or worse, depending... I might do it anyway. Add a gain stage that is. I would set it up as a virtual earth input with local NFB around the stage and trim the output with the feedback resistor. It will invert phase but it will rais gain with very little change in noise or preamp distortion content.

Chuck

[EFK]

Actually I misspoke, it's tone first then volume - I'm not sure if a cathode follower would work. This is a cathode bias amp w/ no negative feedback. The second channel does have two stages and then feeds into the PI, so the phase shouldn't be an issue here - should it? Here's a schematic: the top channel is the one I'm talking about (mines a Supro 88T, not a Gretsch, but their both Valco and with only a few minor changes, it's really the same circuit).

[GibsonLover]

The two channels are already out of phase, so if you haven't noticed, it's probably not an issue. It's no different than the BFDR, which I also find acceptable.
Separate the two inputs on top and wire the unused triode to the second jack. Use the same components as the first stage, except leave off the cathode bypass cap. There will be less gain and hopefully a little more headroom for you before it starts to break up. That would be the "minimum invasion" plan when adding to rats nest wiring. Good Luck.

[EFK]

Oh I see what you're saying - that basically adds another 'channel' so to speak, but that's not what I'm after. I like the sound of this channel, it's just that w/ half of a triode feeding straight into the PI, it's not making a lot of volume. I want to try to maintain the same tone - close to it anyway - but figure out if there is a way to use that other half of the triode to add volume without dirtying it up much more. This why I was thinking of paralleling it, or maybe cold bias stage?

[Merlinb]

Parallel stage would certainly be the easiest way to squeeze a little more voltage gain. You could also reduce or eliminate that 470k resistor before the clean gain pot, as that is throwing half your signal away.

[EFK]

Thank you Merlin. I think I'm going to try this out. Is there a point to that 470K being there in the first place?

[Merlinb]

Thank you Merlin. I think I'm going to try this out. Is there a point to that 470K being there in the first place?

Not really. Maybe the designer wanted it to stay clean at all costs at any setting and with any guitar, which is really just an aesthetic thing.
The same thing is done in the Epi Valve Junior. Dunno why they would throw away so much perfectly good gain in a two-valve channel. If it has too much just *turn down the vol pot*. It's not rocket science...

[pdf64]

There's a lot of signal being lost both on the input and output circuits of V1A. It might be that they are creating the tone you like (due to low and high frequency filtering), but it may be worth experimenting with setting up that whole stage like a standard Fender, with the 2 x 68k / 1M resistors on the switching input sockets, 1k5/22uF cathode, 100k plate.
The 470k feeding the vol pot seems just to be halving the signal, with no other function, eg mixing. If you bypass it though, the 0.001uF cap will need doubling in order to maintain the same low end roll off. Or change to cap to something larger, say 0.022uF, to get full bass response.

[EFK]

Thanks very much guys - I'm going to try this out today. I have a bunch of these old Supros and they are all a pain in the butt to do anything with, but they have a very unique sound and usually a few little changes here and there can really de-constrict them without changing things too much. Regardless of the schematics, they all use 100K input resistors and I usually swap them out for 22K which really helps. The fact that most of the caps are old wildly variable ceramic caps (which don't actually leak, but they're all 20% which is quite a spread) probably contributes to the neat tone. I'm going to parallel V1 first then possibly kill the 470K and swap in a .0022 and see how that goes. I suspect either or both will provide exactly what I want.

[Merlinb]

BTW, is the input stage's bias resistor really 22000 ohms? Surely that's a typo.

[EFK]

hahahahaha yep, that's a typo. It's actually 2.2K.

[Chevy]

One comment... taking out the 470K padding R (you can just short it with a test lead if you want) I think will possibly change the tone of the preamp stage... it will change the effective load on the first stage, which will alter the high and low-pass knees of the output of the first stage. I'm very rusty at this stuff..., and take what I say with a grain of salt... and also what I'm saying may not be too noticeable in the outcome. But I am saying don't be surprised if it doesn't sound exactly the same. Intuitively, I suspect it may be brighter sounding without the 470K... but then again this may be a wrong assumption. Just a wild ass guess.
You might try play with the first stage plate load 100K R's to see what you can get there... perhaps up one to 170K and see what happens. The .05 C in between the two 100K's will affect gain as well... see what it sounds like without it? Dunno... just taking a few WAGS here... good luck!

[Chuck H]

Spot on, and good observation. If the first stage bypass is 22uf then it's already going to add 6dB to anything audible either way. I think bypassing the 470k would be best tempered by reducing the coupling cap value. Not doing the math, but the common values are easy enough to experiment with. Try .022, .01, .0047, and .0022 and go with what sounds best to you. That is if you decide to jumper that 470k.