Hello

I have created different buffers and want to connect these different buffer boards to one input and one output and use a switch to switch between each one to determine which circuit I like. But, what is best way to connect to a switch? Would it be best to switch the inputs or outputs of each board as to not cause any interface from any of the boards.

Thanks
Rob

Are those buffers BJT, JFET, MosFET or op-amp based..??
What do they drive..??
What are their particular input and output impedances..??

Shouldn't be much easier (for us) if you've posted some schematics..??

P.S.
I'd say it should be more easy and plain to switch between driver effects outputs to a single buffer but let's see what's in your mind..

WHY do you have buffers? Input loading? Output loading? Maybe both?

Fer example, if you want to avoid input loading, leaving two idle inputs connected to your source may be bad, unless they are infinite impedance (they never are).

If you connect two outputs together, and they do not have IDENTICAL gain, they will fight.

Do you have wire and solder? Or clipleads? Try it. Better and quicker than taking a poll of random strangers.

I have a Klon, JHS buffer, Cornish buffer, a simple JFET and IC buffer from AMZ. Here is what I was thinking of doing either on the input or output side.

I have found that having a buffer first in line I get a better tone, I loose the highs without a buffer.

Right now I just want to be able to switch between different buffers to hear any big difference between the different styles. Then I will pick the one I like and have just one buffer.

Well designed buffers shouldn't exhibit "differences"..

shouldn't and don't are different though, aren't they?

Get a DP5T switch and try it out! You could CMOS it...but it doesn't seem like it's a permanent thing, just for testing one thing, so too much bother...

But I think your true bypass like set up should work, more proper to put unused effect inputs to ground, but since you're switching both sides you won't hear anything from them anyway.

Switch both sides. Get a 2P4T switch (will probably be a rotary, so get a lever knob). For exactly three units under test, 2P3T would do, but there should be a straight-wire choice, and you may find more buffers. Actually there is a 12-contact rotary switch for $2 where 2P6T is as cheap as any other, so may as well start there. Unused positions can be ignored in the lab. On stage you might jumper to the next throw terminal so you can't switch to an open (BUZZ) position.

Try putting a buffer out on your guitar itself, so it can drive the cable capacitance without treble losses. It may make the choice of buffers moot.

The buffers you have posted are two JFET-based buffers (can work fine as long as Vgs(off) is bigger than the source's peak voltage or the gate is biased at some low positive voltage, like 1V; will result in 0.9x gain), a BJT bootstrapped buffer (can work fine, be sure to use a well-dimensioned input capacitor that limits the bandwidth to that of the guitar, eg. 15-22n for a 1Mohm impedance - bootstrapping tends to cause a low-frequency resonant peak; gain is 0.95-0.99x, depending on the transistor's hfe), two opamp based buffers (probably noisier, although this may be inconsequential - beware of possible phase inversion with the TL07x series on peaks, it might make sense to actually add some gain, like 1.5-2x, which is a slight boost that should not cause any trouble; if you use a bipolar-input transistor, the impedance won't be high enough usually).

On the average guitar amplifier, there shouldn't be any large differences (even with a low Vgs(off) JFET, that might clip negative peaks but you may not hear it). Another possible option is bootstrapping the JFET similarly to what you do with valves, that is, split the source resistor into two, wire the gate resistor to the junction of the two and add an input capacitor. Instead of clipping harshly when the gate voltage goes below Vgs(off) relative to the source, it will tend to distort more smoothly and possibly even pleasing, if at all audible (like 1-2% THD, almost only second harmonic). The downside is that you need to work out how to split the source resistor so that the bootstrapping works best (basically you need a trimmer).

Off-topic, but..

Quote from: fryingpan on October 31, 2024, 07:48:06 PMa BJT bootstrapped buffer (can work fine, be sure to use a well-dimensioned input capacitor that limits the bandwidth to that of the guitar, eg. 15-22n for a 1Mohm impedance - bootstrapping tends to cause a low-frequency resonant peak;

It's a bit more complicated..
Comparable RC time constants (input cap/bootstrapped resistor/bootstrap cap/bias configuration Thevenin resistance) might result into "some" frequency peaking (analogus to Shallen-and-Key active filter) so it's better to NOT dimension input cap according to bootstrapped impedance..
(or place a taming resistor in series with bootstrap cap)
 

Quote from: antonis on November 01, 2024, 05:27:35 AMOff-topic, but..

Quote from: fryingpan on October 31, 2024, 07:48:06 PMa BJT bootstrapped buffer (can work fine, be sure to use a well-dimensioned input capacitor that limits the bandwidth to that of the guitar, eg. 15-22n for a 1Mohm impedance - bootstrapping tends to cause a low-frequency resonant peak;

It's a bit more complicated..
Comparable RC time constants (input cap/bootstrapped resistor/bootstrap cap/bias configuration Thevenin resistance) might result into "some" frequency peaking (analogus to Shallen-and-Key active filter) so it's better to NOT dimension input cap according to bootstrapped impedance..
(or place a taming resistor in series with bootstrap cap)
 

By this I meant that instead of placing the usual large value blocking cap (eg. 1u), it makes more sense to put a small one (eg. 100n, but for guitar, smaller can work) so that the low frequency peak is minimised. Bootstrapped BJT buffers do not work all that well due to the fact that the emitter resistor will present itself as a load, so the top impedance you can get is about 2Mohms at most, and you don't have much margin to reduce the bootstrapping effect as you would with, say, an opamp.

> usual large value blocking cap (eg. 1u), it makes more sense to put a small one (eg. 100n, but for guitar, smaller can work) so that the low frequency peak is minimised.

The smaller cap will shift the frequency and Q up high.

Hehe, I misremembered 
So a big 100u cap it is.

Quote from: PRR on October 31, 2024, 07:01:27 PMActually there is a 12-contact rotary switch for $2 where 2P6T is as cheap as any other, so may as well start there. Unused positions can be ignored in the lab.

+1 this. One big rotary switch with a nice chunky knob on it, plenty of wiring, and you're done.

Let us know what you think once you've tried it, I'd be interested to hear. As has been mentioned, in theory there's no difference between well-designed buffers, so it shouldn't matter which position the switch is in. Get someone else to rotate the knob and see if you can guess which position it's in for a blind test.

I hooked up (4) different buffers using a 2P4T rotary switch. I am switching both the input and output.

I hear no difference.

Klon buffer, JHS buffer, Cornish buffer, and JFET buffer.

Quote from: caspercody on November 06, 2024, 03:39:31 PMI hooked up (4) different buffers using a 2P4T rotary switch. I am switching both the input and output.
I hear no difference.
Klon buffer, JHS buffer, Cornish buffer, and JFET buffer.

Quote from: antonis on October 31, 2024, 03:47:25 PMWell designed buffers shouldn't exhibit "differences"..

Quote from: caspercody on November 06, 2024, 03:39:31 PMI hooked up (4) different buffers using a 2P4T rotary switch. I am switching both the input and output.

I hear no difference.

Klon buffer, JHS buffer, Cornish buffer, and JFET buffer.

Hey, it's a win for theory!! Well, that's nice! 

Just goes to show they must all be decent buffers after all...

Quote from: ElectricDruid on November 06, 2024, 06:30:06 PMthey must all be decent buffers after all...

For that possibly-easy situation. Give them some real work. 500 feet of cable is a start. I've built weaklings that would slew bad in a long line, and buff buffers would drive miles of cable. Try driving a dozen tape recorder inputs (Dead tapers), a passive distribution network (often 150 Ohms).

Quote from: PRR on November 06, 2024, 10:40:16 PM

Quote from: ElectricDruid on November 06, 2024, 06:30:06 PMthey must all be decent buffers after all...

For that possibly-easy situation. Give them some real work. 500 feet of cable is a start. I've built weaklings that would slew bad in a long line, and buff buffers would drive miles of cable. Try driving a dozen tape recorder inputs (Dead tapers), a passive distribution network (often 150 Ohms).

We are NOT so malicious guys..!!