Switching: CD4013 vs. CD4016

[modsquad]

Inquiring minds would like to know if you can use the CD4016 in place of the CD4013.  To me it looks like the main differences is that the 4016 is a quad switch versus the 4013 which is a dual switch.  I am wondering if there is a difference in the noise characteristics.

[gez]

The 4013 is a flip-flop and not a switch (no signal can go through it), although it can be used for switching purposes.

[modsquad]

So I take it that I can use the CD4016 to route signal between multiple effects.

[gez]

So I take it that I can use the CD4016 to route signal between multiple effects.

If you want one signal to be routed to multiple paths, you're probably better off with a multiplexer (can't remember codes of common ones as my CMOS cookbook is downstairs).

RG has a little article about switching using 4053s:

http://geofex.com/Article_Folders/cd4053/cd4053.htm

I think Andrew (Tone God) has done the same sort of thing with 4016s (possibly an article on his site), and Miguel (Bioroids) has done some comparisons between the 4016 and 4066.

[gez]

Here you go:

http://www.geocities.com/thetonegod/switches/switches.html

http://www.dedalofx.com/bioroids/experiments/e01_4016vs4066.php

[modsquad]

I have those articles and read and reread them.   I just bought about 20 CD4016s for about $1.00.  Couldn't you just input the signal to all 4 and then switch the outputs on and off to the effects?

[gez]

I have those articles and read and reread them.   I just bought about 20 CD4016s for about $1.00.  Couldn't you just input the signal to all 4 and then switch the outputs on and off to the effects?

You could do, yes.

[PerroGrande]

You *may* have to bias the signal at half-supply to avoid clipping when running these on a single supply.  The examples on the data sheet that used a sinusoidal input and output showed a split supply operation for the chip.  From this, it is not unreasonable to believe that biasing will be necessary for single supply operation.  If biasing is required, a coupling capacitor is probably also going to be necessary on both sides to remove the DC offset.  Unfortunately, I don't have any of these to test, so this is largely conjecture based on the datasheet.

They also present a 400 Ohm resistance when they are fully on, so make sure the incoming signal isn't negatively effected by this resistance. 

[gez]

You *may* have to bias the signal at half-supply to avoid clipping when running these on a single supply. 

It's almost certain that he will have do this, to avoid switch glitch etc; which is why I linked to RG's page on the 5043 (same principles apply to other switches).

[modsquad]

While doing some research I came up with some interesting uses for switching audio signals that I think will work for my application.

http://www.pic101.com/audiosw/index.html

[PerroGrande]

Yep -- that linked article looks to be doing stuff similar to what you want.  I also noticed that the 4016s in this application are biased to 1/2 the supply, which is as expected.

One other quick caveat -- you'll probably want to debounce the switch to avoid any chatter as a result of the switching operation.  Otherwise, you should be ready to go. 

[modsquad]

The only other issue that I have found while researching is that some claim the 4016 is not suitable due to bad "sonic performance".   Instead they used relays.

[CGDARK]

The only other issue that I have found while researching is that some claim the 4016 is not suitable due to bad "sonic performance".   Instead they used relays.

I agree with them, but if you don't want to use relays you can still use optofets like the H11Fx family. They are like LDRs, but much faster, lower on resistance and clickless operation (when used with the proper driver circuit).

CG

[modsquad]

Yeah thought of that but found this quote from Mr. Vex:

"you could also try using an H11F3 or H11F1 photofet optocoupler, but they usually add distortion to audio circuits... however that can sound interesting too."

[gez]

By default, you're going to end up buffering your signal (you're driving a few inputs so it's advisable), so on resistance of the 4016/4066 isn't really an issue.  They're low distortion too due to the way they're wired up (one MOSFET compensates for the other as the signal rises and falls).  My only issue with CMOS switches is that it's difficult to get totally glitch-free switching (though you can reduce it to an acceptable level), but that can also be true of relays.