So I actually recommended you use the TMUX7308F in your design instead of the CD4051B, It power off protection and overvoltage protection  which both mean there is no esd path from the io pins to supply/ground. This device should operate in a way that eliminates the high leakage current/measurement error readings you are seeing with the cd4051b.  With the TMUX7308F no external protection should be needed.

Lights Australia

However there are devices that have an integrated feature called power off protection that do not have the esd path from input from gnd/supply.

What happens with other channels if 1 channel is current limited overdriven that diode to supply/gnd is conducting a few mA?

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CheapLighting

In theory, as long as that power rail is able to source/sink the current without changing its voltage, nothing happens. But when going outside the recommended operating conditions, correct operation is no longer guaranteed.

yes this would solve issue, same as other instrumentation multiplexer. Also other spces are better but prices are high (about 20 times low cost MUX). We have low cost boards and price would be too high there. I think I´ll make tests to see if it is better than CD4051. At least we can remove the extern level shifter.

So I agree with Clemens in the fact that once the  internal esd diode is  forward biased( aka when a overvoltage event occurs input signal voltage is higher than supply) and the current exceeds what the diode can handle, then that can cause permanent damage to the device in which TI cannot guarantee correct operation after. I believe this FAQ can explain this in further detail.

ALAutomotiveLighting

Please let me know what you think about this device and check out this faq and video for more understanding of why this device should simplify your design.

thanks. Overvoltage is cause by wrong wiring or short between sensor signals and 24VDC. We have all inputs protected with resistors to limit current and nothing get damaged. But currently we use CD4051 and if 1 channel is overdriven (internal protecting diode conducting) high leakage currents on other channels/output appear.  Because we multiplex high impedance sources this cause totally crazy results. As soon as overdrive is removed it works correct again. To avoid this we have to add external overvoltage protection which limit voltages within supply raiils (need many components and cause additional leakage currents of diodes.