Dual OpAmp: mutual impact of the different OpAmps which are in the same housing

[Bigman]

Hi,

I believe to remember that I was told not to use dual (or quad, etc.) OpAmps, because the two OpAmps within one housing could impact each other quite significantly.
I have checked some TI OpAmp specs, but  could not find a parameter, which would characterize how prone the OpAmps within the same housing are for a mutual influence.

Of course, at the end it will be a question about available space on the PCB, how much noise is acceptable, how I define "significant" etc.  ... however, I could image some Dual OpAmps are better "shielded" than others. I am wondering: isn't there any "standard" parameter in the specs, which would tell me how strong e.g for a Dual OpAmp the single OpAmps can influence each other?

Thank for all your comments!

[RandallMcRee]

Sure, its called crosstalk.
This is a signal from one opamp coupling into another.
I just looked in the LT1632 datasheet, they talk about channel separation:

Channel Separation VO = – 10V to 10V, RL = 2k         110 127 dB

High CMRR and PSRR are also necessary in order to prevent unwanted coupling. Also true for singles.

[RandallMcRee]

Oh, there is also thermal coupling. That is where one opamp heats up because its dissipating power and affects the operation of the other one in that package.

You can read between the lines, so to speak, as to this effect by looking at the matching of Vos and other parameters in the datasheet.

E.g.

Input Offset Voltage Match (Channel-to-Channel) VCM = V–, V+ (Note 5) 700 3500 µV

[SeanB]

As well most dual or quad opamps share internal bias circuitry, so if you drive the one opamp into saturation or into slew rate limiting then this will show up on all the others as either a spurious signal or as a DC offset or the output hitting the one power rail while the overload exists. Most common when using the one half as a comparator and the other half as an amplifier, you can get some odd effects as the comparator side switches and this can be only under certain conditions of input signals.

[danadak]

That spec of isolation is T and F and V dependent, and many datasheets
its not present. Usually however the product engineer would have done
some characterization testing, and sometimes manufacturer will share that
info, especially for critical applications. You can ask FAE or rep if there is
some specific OpAmp you want to use and is not speced.

You can always test for this yourself.......just a thought.


Regards, Dana.