Silent Bypass Switch

  Hello, I'm interested in understanding how to make a popless bypass switch.
Apparently a micro-controller is often used.

Here the man explains his whole method : https://www.coda-effects.com/2016/08/relay-bypass-with-anti-pop-system.html

Other circuits are available for sale :
https://www.muzique.com/schem/bypass.htm
http://www.griffineffects.com/home/69-silent-step-true-bypass.html
https://www.modestmikesmods.com/silent-step-true-bypass
https://pettyjohnelectronics.com/shop/quite-bypass-switch/
http://labo-k-effects.com/en/index_mod_en.html

Sometimes they use a PIC sometimes an AT TINY chip, most of them seem to only use a micro-controller and a relay, whereas the first man uses an additional solid state relay.

Does anyone knows more about the principles used to make a silent switch ?

saint gillis said:

  Hello, I'm interested in understanding how to make a popless bypass switch.
Apparently a micro-controller is often used.

Here the man explains his whole method : https://www.coda-effects.com/2016/08/relay-bypass-with-anti-pop-system.html

Other circuits are available for sale :
https://www.muzique.com/schem/bypass.htm
http://www.griffineffects.com/home/69-silent-step-true-bypass.html
https://www.modestmikesmods.com/silent-step-true-bypass
https://pettyjohnelectronics.com/shop/quite-bypass-switch/
http://labo-k-effects.com/en/index_mod_en.html

Sometimes they use a PIC sometimes an AT TINY chip, most of them seem to only use a micro-controller and a relay, whereas the first man uses an additional solid state relay.

Does anyone knows more about the principles used to make a silent switch ?

Click to expand...

A noiseless switch is necessarily based on a time-variable element*, which a relay is not. A relay breabs or establishes contact in a very abrupt way, going from a fraction of ohm to several Megohms in a fraction of second. Silent switching requires a timed transition of a few dozen milliseconds.
The usual implementations are:

• optocouplers: probably the most effective solution in terms of performance, though attention is needed to timing. Its only serious drawback is current consumption, about 10mA. Also, they are going to be difficult to source, since they are not RoHS
• JFET's: another effective solution, where timing is more difficult to manage, and puts restrictions to the possible levels and impedances. Managing feed-through of the control voltage into the audio is difficukt. Very low current consumption.
• optoFEt's: Very similar in performance to bare FET's, they require a less complex drive circuitry, because there is no leak-through of the control voltage into the audio.
• VCA's: certainly the ultimate in terms of performance, at the cost of increased circuitry, they are more costly. I would include OTA's in this category.

The solution proposed in the link is designed to solve one problem that should not exist: clicks due to DC offset. It shuts the output for a time. That ceratinly creates a puncture in the signal.
It will not solve the problem of switching on or off a signal.
The DC offset problem should be solved the usual way: coupling capacitors and/or servo-control.

* I didn't mentioned zero-crossing switching, which requires a much more complex implementation than timed switching.

In fact I'm looking for info because I'm switching a balanced microphone signal, and I have a bit of pops during switching, here's the circuit (I've only showed the input and output stages for clarification) :

https://i.ibb.co/dmqYJ8J/switch.jpg

What would be the cause of the pops according to you?

I mean, obviously it's not a DC thing.
But this technique of grounding the effected output few milliseconds before switching, would it work in this case?

saint gillis said:

I mean, obviously it's not a DC thing.
But this technique of grounding the effected output few milliseconds before switching, would it work in this case?

Click to expand...

Sorry I did not read all of your links and bailed on the one I opened.

Grounding a source just before selecting, could discharge any inadvertent residual voltage in a output blocking capacitor. 

JR

saint gillis said:

In fact I'm looking for info because I'm switching a balanced microphone signal, and I have a bit of pops during switching, here's the circuit (I've only showed the input and output stages for clarification) :

https://i.ibb.co/dmqYJ8J/switch.jpg

What would be the cause of the pops according to you?

Click to expand...

There could be some DC from the capacitors leakage and/or from the load, but the most likely cause is that the preamp sees a very high impedance all of a sudden, which results in a short burst of Johnson noise.
If you can live with a puncture in the signal (actually it seems you are switching two different signals so it would not be an issue), you should short the output first, than divert the source and only then open the output. Relays would be adequate.

saint gillis said:

https://i.ibb.co/dmqYJ8J/switch.jpg

Click to expand...

R51 470k will NOT keep C47 100uFd totally drained, especially at mike level.

The TL072 will drive 2K fairly well, R51 R52 can just as well be 10k and get 1/50th the leakage voltage.

Have you checked these points with a milli-volt-meter?? 1mV at mike level is BIG. It could be wise to get a 4.5-digit DMM which may read sub-mV.

Yes that's right, in reality I have two 22µ bipolar in parallel equivalent to 44µ at the output. I thought something like 10K to drain could parallel with a potential 10K input stage of another device and could start to make an audible low-pass.

In fact I realize that if this output is connected into say a 1KOhm impedance mic input preamp, the value of the output capacitors will have a much more bigger influence on the low-pass than the drain resistors. So your right I will check this out, thank you.

So the DC offset is very low, before AND after the ouput capacitors.
I guess the solution is to short the outputs (hot and cold) few ms before and after switching.