Better way to switch channels in preamp

[marcelomd]

Hi!

I designed a preamp based on the Zvex SHO and Box of Rock. Post here

Currently I'm switching channels using a simple 3PDT. Like this:


The issue is that the channel that is not active is "floating". Is that ok?

Both channels have a large anti pop resistor to ground at input and output. Is this enough to keep it silent while not in use (as to not interfere with the active channel)?

I'm trying to avoid a bunch of relays and JFETS if I can.

Thanks!

[Ben N]

'Twere an amp, you might shunt the the unused channel to ground right after the first stage, while selecting between outputs to feed the PI or power amp.

[marcelomd]

'Twere an amp, you might shunt the the unused channel to ground right after the first stage, while selecting between outputs to feed the PI or power amp.

Let's see if I understand:
- Connect input jack to both first stages at the same time;
- Use one pole to ground the output of one of the first stages;
- Use the second pole to select between channel outputs.

Correct?

[antonis]

Both channels have a large anti pop resistor to ground at input and output. Is this enough to keep it silent while not in use (as to not interfere with the active channel)?

If this was enough, they shouldn't work either on their "active" state..
(in the mean of anti-pop resistors don't kill iN-OUT signals..)

You need to wire 3PDT switches with effects IN & OUT grounded when by-passed..

[garcho]

Ground unused inputs, but do NOT ground outputs.

Or am I missing something and wrong like usual?

[Ben N]

'Twere an amp, you might shunt the the unused channel to ground right after the first stage, while selecting between outputs to feed the PI or power amp.

Let's see if I understand:
- Connect input jack to both first stages at the same time;
- Use one pole to ground the output of one of the first stages;
- Use the second pole to select between channel outputs.

Correct?

Yeah, but if you just have separate input buffers you can put the shunt there--just so long as you aren't inadvertently grounding the *on* channel. BTW, I'm not saying this is the right way to do it in a low voltage SS preamp, just how I've seen it done in HV tube preamps, and it avoids any issue of floating and kills noise.

[marcelomd]

Yeah, but if you just have separate input buffers you can put the shunt there--just so long as you aren't inadvertently grounding the *on* channel. BTW, I'm not saying this is the right way to do it in a low voltage SS preamp, just how I've seen it done in HV tube preamps, and it avoids any issue of floating and kills noise.

Any resistor between buffer output and ground?

You need to wire 3PDT switches with effects IN & OUT grounded when by-passed..

Ground unused inputs, but do NOT ground outputs.

How can I do that with one 3PDT?
- Pole 1 selects LEDs
- Pole 2 selects outputs
- How can pole 3 select one input AND ground the other?

I'm sure I'm missing something.

[antonis]

You didn't say (or draw) anything about LEDs..

In such a case, ground only Inputs (as gary said..)

[marcelomd]

You didn't say (or draw) anything about LEDs..

In such a case, ground only Inputs (as gary said..)

I forget only my wife can read my mind... =)

So leave both outputs connected and use two poles to ground the inputs?

[amptramp]

You might consider driving both channels simultaneously and using one anti-pop resistor and one RF filter for both.  Since the stages are high impedance, the extra load in negligible.  Just connect the input side of C14 to the input side of C2 and eliminate R14, R15 and C13 in the Silvia thread.  This frees up the SW1A switch section.  There is no harm in driving both preamp channels at once.

You only need to switch at the preamp output.

[marcelomd]

You might consider driving both channels simultaneously and using one anti-pop resistor and one RF filter for both.  Since the stages are high impedance, the extra load in negligible.  Just connect the input side of C14 to the input side of C2 and eliminate R14, R15 and C13 in the Silvia thread.  This frees up the SW1A switch section.  There is no harm in driving both preamp channels at once.

You only need to switch at the preamp output.

Can I expect any noise issues from, say, the higher gain channel interfering with the clean channel?

[Rob Strand]

Hi!

I designed a preamp based on the Zvex SHO and Box of Rock. Post here

Currently I'm switching channels using a simple 3PDT. Like this:


The issue is that the channel that is not active is "floating". Is that ok?

Both channels have a large anti pop resistor to ground at input and output. Is this enough to keep it silent while not in use (as to not interfere with the active channel)?

I'm trying to avoid a bunch of relays and JFETS if I can.

Thanks!

For true bypass you can do a version which shorts the effect input in bypass mode, instead of leaving the effect input open.   The common of the input switch goes to the effect in, one switch contact is wired permanently to the input socket, the other contact is wired to ground.  The trick is in bypass the output switch is also wired to the input socket.

See page 6,
https://madbeanpedals.com/tutorials/downloads/MBP_FootswitchWiring.pdf

The same trick doesn't work for channel switching.  You can get a working configuration with a 3PDT switch.

[Ben N]

So leave both outputs connected and use two poles to ground the inputs?

One pole to ground, throws connected to each channel input.

So one pole selectively grounds inputs, one pole selects output, and you have a third for the LED.

[marcelomd]

Finally back at home =)

So, grounding after first stages, selecting between outputs and LEDs. Like this?



Thanks

[Rob Strand]

Finally back at home =)

So, grounding after first stages, selecting between outputs and LEDs. Like this?

It's a good start. 

A couple of finer points:

- The input impedance of those MOSFET stages are a lot lower than what you expect
   and putting two in parallel will make that lower.   To first approximation you could
   raise the 1M gate resistors (4 off) to 2M2 but that won't fix the capacitive loading.

- The input impedance varies with gain.   It can also become non-linear or glitchy
   when the stage clips.   If for example you set one stage to full gain and the other
   to clean you don't want the glitch impedance of the overdriven stage to leak
   into the clean.   That's going to happen when the input is fed from a guitar
   plugged directly into the input, as the guitar input is high.  If a low impedance
   buffered outputs is connected the glitchyness will be reduced.

So you probably should test the set-up to make sure it is problem free.

[Ben N]

So a BJT emitter follower (or FET into BJT) input buffer, split to the two MOSFETs?

[garcho]

The MOSFET drains i.e. outputs are directly tied to ground? Ground inputs, so semiconductors amplify 0V, don't ground outputs.

[marcelomd]

So a BJT emitter follower (or FET into BJT) input buffer, split to the two MOSFETs?

Looks like I'll have to do it. Not sure how it will change the tone when overdriving the inputs though. Let's see in the next version.

The MOSFET drains i.e. outputs are directly tied to ground? Ground inputs, so semiconductors amplify 0V, don't ground outputs.

Yeah. The challenge here is to ground one of two inputs with only one switch (the other 2 switches of the 3PDT being used to select outputs and LEDs)

---

Without changing the circuit too much, I can think of:
- Using some kind of Millenium bypass LED driver to free 1 switch and use 2 switches for the 2 inputs and 1 for the output;
- Using 2 switches on the 2 inputs and having the 2 outputs connected to/mixed at the output jack.
- Daughterboard with some relays

[marcelomd]

Just want to post here that this was finally solved in this topic.