Switching channel inputs with a relay at amp input?

[wizard333]

I've never seen this done exactly, so I'm curious if anyone sees a problem with it?

What I'm talking about is, on a multi channel amp, switch the input jack to one channel input or the other right up front with a relay.

Channels switching amp designs seem to
1) Share an input stages. Don't wanna on this one for all kinds of reasons.

2) Have the input to both channels run in parallel. I experimented and decided this had negative tonal affect on both channels; not entirely sure why. If the 12ax7 input impedance can be considered high enough that the input impedance is essentially the same as the input grid leak resistor, putting two in parallel shouldn't make an audible difference, but it does.

3) Input usually to one channel (generally clean) and running in parallel when channel 2 (typically gain) kicks in. Leaves clean alone when it's on, but doesn't do anything nice for gain channel, so don't wanna on this one.

Since I don't see inputs relay switched right after the input jack, I'm guessing there's a reason, but not sure what that is give you have grid leaks on both channel inputs.

Before someone mentions it, yes there is a scheme for switching channel outputs, so don't worry about that end of it.

Anyone?

[Chuck H]

Actually... People do this sort of thing all the time with A/B (or A/B/Y) switches. Just not usually in the same amp chassis. I don't see a problem. But we may see a better solution if you would include a schematic.?.

[wizard333]

I don't have one drawn up but one side is a Fender AB763 Reverb preamp (with modded 2 stage reverb instead of 3) and the other is a 3 stage gain channel I came up with while modding other things. The input stage is 100k plate 300v source from node 4, 2.7k cathode resistor bypassed by .68uf cap, .0033uf output coupler & 1M to ground feeding a series of RCs into the next stage.

[Chuck H]

Draw a schematic? You'll be glad you did.

[Gtr0]

I assume the op means something like this...



Sorry about the fact that it is drawn/photographed, it was the fastest way to do it.

I too would like to know this... I mean I assume it would work, but is there some capacitence needed before it? Since we are talking a small AC signal coming in from the guitar and to a relay straight away. I have an amp similar to an AB165 and I have a DPDT at the input instead of several input jacks and the switch makes a popping noise, so I wonder if a cap to ground or other would fix this... of course a relay should have a diode from power + to - and am I wrong to think this supresses the popping of channel switching??

[Zozobra]

Look at the recent friedman stuff. He uses a relay to bypass the first stage. Is that what you mean?

I assume the op means something like this...

[ATTACH=CONFIG]48808[/ATTACH]

Sorry about the fact that it is drawn/photographed, it was the fastest way to do it.

I too would like to know this... I mean I assume it would work, but is there some capacitence needed before it? Since we are talking a small AC signal coming in from the guitar and to a relay straight away. I have an amp similar to an AB165 and I have a DPDT at the input instead of several input jacks and the switch makes a popping noise, so I wonder if a cap to ground or other would fix this... of course a relay should have a diode from power + to - and am I wrong to think this supresses the popping of channel switching??

Each tube still needs to see a grid leak resistor, even when disconnected from the input, otherwise it'll die.

[wizard333]

Look at the recent friedman stuff. He uses a relay to bypass the first stage. Is that what you mean?



Each tube still needs to see a grid leak resistor, even when disconnected from the input, otherwise it'll die.

Right, which is why I mentioned that.

Friedman either feeds stage one of the clean channel into the gain channel (HBE or "Boost" depending on the amp) or bypasses that and goes into the gain channel, but he's using a .1uf input cap and a 2M grid leak on that gain channel when the input is in parallel with the input of the clean channel. He doesn't switch out the input to the clean channel grid when the gain channel is engaged in non-boost mode; i.e., the inputs to both grids are in parallel.

That actually brings up another question: There is a 1M grid leak on the clean channel at the input jack; if you're going direct to the gain channel (I.e. bypassing the boost stage that is stage 1 of the clean channel) there is a .1uf cap followed by a 2M grid leak for that channel. Does that .1uf cap isolate the input from the 2M giving the normal 1M input impedance, or are they effectively in parallel giving a 666k input impedance for that channel? Or is the .1uf there just to give DC isolation between the grid of channel 1 and 2?

I'm not sure I see the point; that would roll off high end going into that channel with passive pickups, maybe that's what he wanted.......I think I'd have used 2M in both places so the pickups still see the larger input impedance, and what we're used to hearing, with the 1M input, and input cap on both in case that caused any DC on the grids. A 2M input impedance on the clean channel won't hurt the tone, or even affect it in any appreciable way other than maybe a little extra sparkle due to the pickups seeing a larger input impedance. Or am I nuts?

I did an aural experiment with the amp I'm working on with exactly that setup to see how it changes the tone, and it does change the tone on the gain channel a bit (not in a good way) vs. just switching inputs with a relay.

After more thought, I'm wondering whether the miller capacitance sums when you put the two grid stages in parallel. For a JJ 12AX7, Cga = 1.7uf, so with a gain factor of ~80, you'd get 136pf, doubled 272pf, which is an audible difference. Not exactly what I'm hearing, but that would have to affect the tone slightly, mostly in the 'air' and pick attack range. Maybe the .1uf in series with that second miller capacitance is designed to eliminate that effect?

Still doesn't explain the difference I'm hearing.

[Chuck H]

More like this?

[wizard333]

More like this?

Yes sir.

[Chuck H]

Right. The only rub is the additional loading at the inputs (500k vs. 1M). Enough to shift the resonant peak of the guitar circuit just noticeably. For better or worse I suppose.

[wizard333]

Right. The only rub is the additional loading at the inputs (500k vs. 1M). Enough to shift the resonant peak of the guitar circuit just noticeably. For better or worse I suppose.

If I were setting it up that way, I'd have a 2.2M on the jack, then a 2.2M grid leak on each input grid, so when either channel was activated it would have effectively a 1.1M input impedance (or that in parallel with the guitar's volume pot anyway).

[Chuck H]

Or, try it with all 1M, all 2.2M and pick the one that sounds best. The difference in gain is not worth noting. I tend to like the sound of a slightly lower resonant peak from the guitar, but you do have to hear it in each instance to know.

[Gregg]

Look at the recent friedman stuff. He uses a relay to bypass the first stage.

And it was popping like crazy to the point he was seeking advice how to handle it.

[Chuck H]

And it was popping like crazy to the point he was seeking public advice how to handle it.

Link?

[wizard333]

And it was popping like crazy to the point he was seeking advice how to handle it.

? I haven't seen that issue with it. Maybe that's the reason for the .1uf cap into the 2nd channel, among other things. Could be that, could be blocking DC from the grid caused by using 2.2M there, who knows. In any case, it seems to work just fine.