geofex CD4053 switching problems

[MikeH]

I'm trying to do the 4053 bypass switching as detailed here on geofex: http://www.geofex.com/article_folders/cd4053/cd4053.htm

I have it on the breadboard right now and it thumps/pops like a mofo when I switch it on and off.  I've read these can be a bit finicky- anything I might be doing wrong.  I've triple checked and I have all of the right components in place.  As a side note; if I connect the fx in and fx out together (iow, no effect in the path) the switch is almost dead quiet.  I've tried it with multiple different effects and they all pop.  Thoughts?

edit: FYI, I'm using the SPDT control method.

[R.G.]

The secret to getting the CMOS switches to be quiet is to use capacitors to break the DC path from other places, and then to tie a biasing resistor to the middle of their power supply. When this is true, you're not switching between two DC levels, and the insides tend to cancel out any control signal feedthrough.

I promise you, if it's popping, you're either switching between two (or more!) DC levels by not breaking the DC paths from outside the chip, or you don't have the analog switch input/output pins pulled to half the ir power supply voltage.

Well, OK, the chip could be damaged. But the probabilities heavily favor a mistake in hooking the circuit up.

[trjones1]

One problem that I ran into when first playing with this circuit was that I kept the original pulldown resistors to ground on the circuit that I was switching, causing massive popping.  Clipping them out solved it. 

[MikeH]

One problem that I ran into when first playing with this circuit was that I kept the original pulldown resistors to ground on the circuit that I was switching, causing massive popping.  Clipping them out solved it. 

Hmmm- I'll try that. 

I promise you, if it's popping, you're either switching between two (or more!) DC levels by not breaking the DC paths from outside the chip, or you don't have the analog switch input/output pins pulled to half the ir power supply voltage.

I'm quite certain that I have the input and output pulled to vref (though I will double check voltages), but I'm not sure what you mean by "not breaking the dc paths from outside the chip".

[amptramp]

I promise you, if it's popping, you're either switching between two (or more!) DC levels by not breaking the DC paths from outside the chip, or you don't have the analog switch input/output pins pulled to half their power supply voltage.

I'm quite certain that I have the input and output pulled to vref (though I will double check voltages), but I'm not sure what you mean by "not breaking the dc paths from outside the chip".

This means your switch should be capacitively coupled at the input and output so there is no possibility of connecting to an input or output with a different DC voltage superimposed on it.  Another possibility would be to have the switch connect to an inverting op amp stage with the inverting node held at Vref.  It would pay to use a small resistor to isolate the amplifier input and a small feedback capacitor to stabilize it, but this would also keep the CD4053 biased near its preferred voltage midway between the positive and negative supply voltage (where the "on" resistance is lowest).