where and when to use/not to use R.G. electronic switch bypass

[Dimitree]

hi everyone,
I'm a fan of the electronic switch bypass presented by R.G.

but I'm wondering whenever it is good to use it and when not.
For example, can I use it in front as bypass for a fuzz pedal / high gain pedals?
And when I need to switch something inside the circuit, for example break a connection, or alternate 2 different caps, do I need some modification to the standard scheme?

thank you and sorry for my bad english

[R.G.]

Much depends on understanding the CMOS switch itself. CMOS switches look to the rest of a circuit as either an almost infinite resistance, or an ever-so-slightly non-linear low resistance. Much work on the insides of the switches has made the "slightly non-linear" be essentially a non-issue. But there is some on-resistance.

For the  CD4000 series, this is on the order of 100-300 ohms in most cases. So it's not a perfect "short circuit". But as long as you understand that the resistance when on is not zero, you can usually design around it.

There are power supply limitations. All of the pins on a CD4000 switch chip are protected by reverse-biased diodes to the most-negative power pin and to the most-positive power pin. These diodes conduct when signals bigger than the power supply voltages are applied to the circuit. Note that this can also happen on the "off" side of a switch. The solution is easy: ensure that you're not feeding the chip signals that go outside its power supply range.

There are slight offsets and control signal feed throughs inside the chip. The chips switch with nearly no switching noise when the signals being switched are exactly in the center of the chip's power supply, as the positive and negative going offsets tend to cancel there. This is not a big deal in actual practice, but it is something to think about. If you have a choice, make the CMOS switching pin have a DC level in the middle of its power supply.

And just like metal switches, CMOS switches will happily connect different DC levels into the signal path when switched. This makes for nasty pops. This is not a CMOS defect - it's doing exactly what you told it to. To avoid this issue, ensure that the signals being switched are both at the same DC level. This is most convenient to do if you use a capacitor to break the DC from the actual signal and bias the CMOS pin at half its power supply.

In using it to bypass a high gain pedal, whether by CMOS or metal switches, things will be quieter and less likely to oscillate if you AC-ground the effect input when it's bypassed.

[Dimitree]

many thanks
I hope I understood correctly,
one of the example I'm finding hard to understand how to implement an electronic switch solutions, is to get "trails bypass" in the deluxe memory man.

This should be a solution, but I'm not sure if I can fit an electronic switch without modifications (and without getting audible pop)

if I'm not wrong, both sides of the switch are already at half DC level, right?

(this is the complete schematic without the mod)

[PRR]

Does it pop with the mechanical switch?

Are all switch pins within 18V of each other (including peak signals)?

The it "should" be fine----

BUT! This plan is NEGative 15V and ground. That is not a real problem but you have to think differently to conventional positive-hot arrangement.

[Dimitree]

Does it pop with the mechanical switch?

didn't test it since I don't have the opportunity to try the pedal. I just tested it on simulations. Is there a way to predict it, with the help of some adequate knowledge? 

Are all switch pins within 18V of each other (including peak signals)?

One pin would be all the time at -7.5V, the other two depend on the audio signal level, but shouldn't be that high (even if the first opamp can give 10x gain)

BUT! This plan is NEGative 15V and ground. That is not a real problem but you have to think differently to conventional positive-hot arrangement.

I guess I could just wire the 4053 with -15V on the GND pin and 0V on the +V pin, just like the CD4047 used in the clock section, right?

[PRR]

> like the CD4047 used in the clock section, right?

Good guide.

[Dimitree]

and what about those pedals that don't like high impedance or buffers on its input, like fuzz pedals?

should I adjust the 1M resistors on the CD4053 when I want to bypass those pedals with such electronic switching?

[Fender3D]

Dimitree,

I know the topic asks about R.G. switch, but I think in your DMM will be easier use just a FET (and 3 caps and 3 resistors, but you'll need them with 4053 also).
Not a matter of money... I just feel sad for the other 2 switches (wasted) alone inside the chip....   

[Dimitree]

yeah I could do that, I wanted to use 4053 simply because I've made a lot of small "switch pcb" that I'm putting to my pedals to use with soft touch switches.

any advantage of the FET in this application?

[PRR]

> should I adjust the 1M resistors

Why? They hardly affect interface impedance.

When the switch is ON, it is about 300 Ohms plus whatever drives it. In most audio, this is "low".

When the switch is OFF, there is NO signal, so we generally do not care what the impedance is. (We are either silent or using another signal.)