1
\$\begingroup\$

This is abit a hypothetical question. Imagine a signal chain where a AC-DC power supply is powering a transducer and the transducer output is coupled to a data acquisition input by a BNC cable. Or I can illustrate the chain like:

AC mains->PSU->Transducer---------------->DAQ.

And the DAQ is single ended as well. I tried to simplify this by the drawing below: enter image description here

Now Vsig represents the transducer; Vcm represents the common-mode voltage. X and Y are the output terminals of the transducer. Rs is the output impedance, Rw is the BNC resistance. A and B are the terminals of the coax cable right at the DAQ side where signal is BNC coupled to the DAQ. Rin is the input impedance of the data acquisition board.

Now assuming inside the dotted line is a black box to us and we don't know initially what we measure as noise is CM interference originated or not. And our aim to make a quick test to verify if the noise is originated due to a CM interference. After that premise my two questions are as follows:

1-) If I short the terminal Y with a wire to the power supply ground, and if in that case the noise disappears does that reveal the existence of Vcm?

2-) If I only have access to A and B terminals, is there any way to reveal the existence of Vcm? Can this be done by a multimeter between the terminal B and DAQ ground(when the BNC to DAQ is decoupled)?

\$\endgroup\$
  • \$\begingroup\$ If you use a current probe on the BNC cable (the entire cable, not just the positive lead) this will show you only CM noise, as the DM will cancel itself out. You can then use your resistance measurements to calculate CM voltages. \$\endgroup\$ – Stiddily Dec 10 '19 at 17:28
1
\$\begingroup\$

In some cases, \$V_{cm}\$ is an AC voltage produced by current in a ground loop (in orange, below).
It cannot be shorted out.
If this ground loop can be broken, no current can flow around the loop, so no voltage is developed across Rw. One way out: disconnect (float) \$V_{sig}\$ so that Y doesn't connect to ground.

In some cases, the signal source cannot be floated, and is unavoidably grounded. And it is likely that the DAQ is grounded as well. So the loop cannot be easily broken.

Adding extra ground wires likely increases loop area or loop current. \$V_{cm}\$ gets worse, or at least doesn't decrease. A very annoying problem to solve, especially in complex measurement systems where lots of equipment is grounded.ground loop Vcm

|improve this answer|||||
\$\endgroup\$
  • \$\begingroup\$ But besides ground loops, how can we test if there is capacitive coupled CM voltage noise? \$\endgroup\$ – GNZ Dec 12 '19 at 10:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.