9205 differential measurement >10V

[Madottati]

See the attached block diagram of my setup, I have a 16V power supply supplies a DUT, in the +ve path I have a shunt. I measure the voltage drop over the shunt differential with an NI 9205. I have read in the operating instruction of the 9205:

Maximum working voltage for analog inputs (signal + common mode) => Each channel must remain within ±10.4V of COM. 

Well, in my case the working voltage is ~16V, still the DAQmx task returns plausible measurement values. It is nice, that it works, but I don't understand why?

I guess I should connect 16V to COM, but then I could not use toe other AI channels of the 9205 for GND referenced measurements. 

[santo_13]

It might work but the accuracy specifications are not guaranteed and prolonged usage might result in permanent damage to the analog front end.

If you want to protect the instrument and functionality within the accuracy specifications in the datasheet, then you've to abide by the limits of the instrument specified in the datasheet.

Alternatively, you can use a commercial high-side current-sense amplifier modules (like INA series from TI) and measure the voltage output of those. 

https://www.sparkfun.com/products/12040

https://www.adafruit.com/product/904