RG Help! Andrew the tone god wicked switches

[Rayman]

RG,

I read your article on CMOS electric switching and I read the wicked switch article.  In your article you bias the ins and outs of the CD4053 with 510K resistors supplied with half voltage.  The schematics on the wicked switches use a CD4066, but don't have any biasing resistors.  Are they not required for the CD4066?

thanks,

Ray

[The Tone God]

The Wicked Switches(WS) article is written to show how to use multiplexers in a generic manner. WS's can be used to switch signals other then audio and in some cases the bias network suggested by R.G. could interfere with the intended function. I believe I quoted/linked to R.G.s article to help cover that territory.

So do you need to bias the signal for audio situations ? Sometimes you can get away without it if you are careful but if you want to be safe then bias the signals.

Andrew

[David]

RG,

I read your article on CMOS electric switching and I read the wicked switch article.  In your article you bias the ins and outs of the CD4053 with 510K resistors supplied with half voltage.  The schematics on the wicked switches use a CD4066, but don't have any biasing resistors.  Are they not required for the CD4066?

thanks,

Ray

Ray:

If it's any consolation, I breadboarded a Wicked Switch with the 4049 and SPST switch as diagrammed several years ago.  It worked just fine.  I didn't complete the project because I couldn't make PCBs at that time.  YMMV.

[Rayman]

David,

Thanks, did you bias the 4053 (or whichever IC you used)?

[R.G.]

CMOS analog switches can switch signals from the most positive power supply to the most negative one. They do this with or without biasing. The biasing is to prevent popping from feedthrough of the control signal. It's a peculiarity of the CMOS transmission gate that there is some feedthrough in both the N-channel and P-channel devices making up the actual pass gate. But the feedthrough is in different polarities for N and P devices, and it varies with the DC condition of the pass transistor. So near one side, the N channel may have more feedthrough than the P channel, the reverse for near the other power supply.

There is a sweet spot near the middle of the power supply where if a signal is right there, the N and P channel feedthrough cancels. This corresponds to the minimum pop condition, and is why I always bias CMOS gates of any kind if I'm switching low level audio. It makes the switches almost completely silent.

If it's not low level audio, the pops become a non-issue anyway, so you can relax the biasing. But there is no substitute for knowing the details of the application.

No generality is worth a darn. Including this one. 

[David]

David,

Thanks, did you bias the 4053 (or whichever IC you used)?

I used the 4049 and 4066 as specified by TTG.  I may have been lucky or am just spitting into the wind.  Mr. Keen has forgotten more than I will ever know about audio.  Heed him, not me.