Capacitor bank switching using CD4051 ?

[tss]

I have some tone stack and I would like to switch various capacitors to a resistor and to ground. My supplies are +/- 9V and the audio signal is referenced around 0V. I'm intending to wire this up as shown below:


Will this work pretty much as I think it will?

[armdnrdy]

I would bring the audio into X....and connect each cap to it's own "output" with the other end of the cap grounded.

You might have an issue with all of the cap inputs connected together.

One thing that I've noticed with the CD405X series switching...sometimes things look good on paper but don't work as intended in reality.

Best to bread board! 

[amptramp]

The total of 18 volts across the 4051 is near the upper limit of the device power.  The TI spec sheet says 20 volts total but other manufacturers may specify a lower limit.

The CD4051 has an "on" resistance of 125 ohms which may be negligible in some circuits but fatal for others.  In this case, it will be the lower limit for the resistance connected to pin 3.  There is some non-linearity in this resistance which can go higher based on temperature and the voltage applied.  Check the spec sheet here:

http://www.ti.com/lit/ds/symlink/cd4051b.pdf

Since stompboxes are rarely linear devices, the linearity of the CD4051 may add to the expected effect.

[tss]

I would bring the audio into X....and connect each cap to it's own "output" with the other end of the cap grounded.

You might have an issue with all of the cap inputs connected together.

One thing that I've noticed with the CD405X series switching...sometimes things look good on paper but don't work as intended in reality.

Best to bread board! 

This is so the caps will discharge when they see high impedance from the 4051?

[armdnrdy]

I would bring the audio into X....and connect each cap to it's own "output" with the other end of the cap grounded.

You might have an issue with all of the cap inputs connected together.

One thing that I've noticed with the CD405X series switching...sometimes things look good on paper but don't work as intended in reality.

Best to bread board! 

This is so the caps will discharge when they see high impedance from the 4051?

No...

The CD405X series doesn't know if X0-X7 is the input or the output...
You have the audio signal coming into all of the "channels" through the capacitors.
The audio signal will pass through the capacitors to X0-X7 possibly causing each channel to be held high creating issues.
Used in this configuration, you might need pull down resistors on each channel to pull them low.
This is why I suggested switching the audio signal in to X and grounding the capacitors on each channel.
As I said before...best to breadboard before you commit to any type of a build. I know from my own mistakes...too often...I've assumed that something "looks good in theory" and proceeded with a build

[anotherjim]

As long as the input signal bias (no signal DC level) is ground, it might be ok. The caps won't discharge in any great hurry into the "off" switches, but they shouldn't have any significant DC across them to be discharged IF the input bias is ground.

If you do get switching "pop" noise, it may be necessary to fit high value (1M-10M) "anti-pop" resistors to ground from the switch side for each cap OR across each cap.

[PRR]

> You have the audio signal coming into all of the "channels" through the capacitors.

But the other way, you have the audio signal coming into all of the "channels" direct.

By symmetry, I don't see that it makes any difference.

Yes, there is a pop problem (either way). 10Meg from each cap-side to out-side bias is the fix. (Alternatively, 10Meg from X1 to X2, X2 to X3, etc.... saves one resistor. This assumes that one switch is ON, and that the caps can discharge through the up-to-70Megs of resistors.)

[tss]

Thanks for the info guys. I am waiting for some 4051 to arrive to test this. I am wondering how come we don't see this simple circuit used all over the place. I know the resistance is not linear and not insignificant, however many times you can scale networks so that the 100 - 125+ Ohms this introduces to the signal will be insignificant and the non linearity is not THAT bad and we are not doing any HI-FI stuff here so the output does not need to be any kind of reference. This gives a lot of flexibility in tone circuits.

[armdnrdy]

> You have the audio signal coming into all of the "channels" through the capacitors.

But the other way, you have the audio signal coming into all of the "channels" direct.

By symmetry, I don't see that it makes any difference.

Yes, there is a pop problem (either way). 10Meg from each cap-side to out-side bias is the fix. (Alternatively, 10Meg from X1 to X2, X2 to X3, etc.... saves one resistor. This assumes that one switch is ON, and that the caps can discharge through the up-to-70Megs of resistors.)

The audio signal isn't coming into any of the channels until A,B,C direct it to. X0-X7 will be sitting at ground.

There may not be a popping problem with the bipolar supply and the use of the VEE pin.
R.G. outlines the signal has to be biased at around 1/2 supply. If I recall...the VEE pin and a bipolar supply takes care of this issue. Read the data sheet to confirm.

[tss]

AFAIK all non selected channels are high impedance and not ground. Making unused channels grounded will limit the usability of the device a great deal... 

[armdnrdy]

I don't think you understood what I was referring to.

Each caps is connected to ground via X0-X7...the incoming audio signal enters X.

When a particular channel is selected via the control inputs, (A,B,C) the audio signal is connected to that channel's cap.

There is no grounding unused inputs.