I am having trouble understanding what would happen if I used a an analog switch IC that is powered by a VCC and GND-1, and controlled by (select line) a high/low signal on GND-1, but switches the i/o of the device between two signals that are referenced to GND-2 and GND-3. What I want to know is if this is possible or if every signal in the switch needs to be referenced to the same ground. Below is the switch logic.There is also a VCC and GND pin for the device. 1S/2S are on the same GND as VCC. nY0 signals are GND-2, nY1 signals are GND-3. Does this work or do the internal FETS need all signals referenced to the same GND?

 

Depending on the magnitude of the ground difference and the current through the switch, massive destruction is possible.

 

Do the grounds ultimately relate to each other? If not (and even if they do) you will probably run into common mode issues. With some notable (and usually expensive) exceptions all the signals must be within the supply rails (or within a volt or so). There are devices that can handle differential signals with a high common mode but they are rare...

What's the application? Maybe there's an alternative approach.

 

Specifically, how much Current and Voltage are you
wanting to Switch in your "not -the-same-Ground" devices ?

Is it AC, or DC, Power,
or is it a varying low-Current, AC-Signal,
or is it a low-Current-DC-Switch ????
.
.
.

 

It seems there is a flaw in your design if you have to switch separate grounds thru the same analog switch.
Most analog switches are CMOS type devices internally referenced to the same ground. I'm pretty sure the switched signals would have the same ground reference. What is the ultimate destination of the signals after switching? Which ground will the device that processes these signals use?
Otherwise, why not just use separate switches?

Can you post a schematic?
Maybe you can optical switches in some way...

 

A friend of mine built a volume control with an IC. There were two L/R channels. One was controlled by a TTL signal and the other a CMOS signal. He didn't catch the design error, I did.

 

I'd start with a differential amplifier, to get all the signals referenced to the same ground.

 

The CD4066 has been around for a long time. You do have Vss and Vdd. You can switch +-7.5V signals, but you will have wacky logic voltages to deal with.

(1)All controlinputsare protected by the CMOSprotection network.(
2)All p substrates are connected to VDD.
(3)Normal operation control-line biasing:switchon (logic1), VC= VDD; switchoff (logic0), VC= VSS.
(4)Signal-levelrange:VSS≤Vis≤VDD.