Noise problem

[ABiermans]

Hi

I am using a PCI-6024E DAQ board to measure the voltages on a number of 100 ohm@0 C PRT probes and convert them to temperature, excitation current 1 mA.
When turning on some 3 phase pumps the voltages jump. I powered the measuring circuit with a battery and still see the same voltage jump so I know my groundlevel
is not stable. A different experiment in the room experiences the same AC induced on the signal so I know it is coming from the powerlines. How can I fix this short of
installing a transformer to get balanced AC power in the room? The AC now is 3 phase based unbalanced AC. Not an easy/cheap thing to change.

suggestions?

Thanks in advance.

Arne

[AnalogKid2DigitalMan]

Could it be induced voltage (EMI/RFI)? Is your wiring to your probes twisted or shielded in any way?

How long were the wires used and what type to connect to the battery in the test you mentioned?

If you have an oscilloscope, can you verify any noise on excitation and signal lines?

Are pumps properly grounded, do they have noise supression or snubbers to dampen transients they produce?

Is a 'soft start' on the pumps possible?

Do you have AC powerline monitoring equipment to verify noise, surges, spikes, etc?

It may not be a ground issue, just throwing out some ideas to verify what it is or not is.

[ABiermans]

Instead of probes I have also used just simple resistors on the board and I see the same thing so it is not EM pickup in the probewires.
I used a few 9V batteries in series, the wires are just hookup wire and they are short. The scope shows the noise on the signal lines.
The noise was also visible on the +- 15 V supply that previously powered the cicrcuit.

So, I have eliminated the probe wires as the noise source and also the simple power supply circuit on the board.
The pumps are grounded but do not have noise suppressors or snubbers. The pump controllers allow 'soft start' rather than just
100% off to 100% on but it makes no difference. The controllers are of the 1 phase in 3 phase out type, why? I dont know, especially since
we have 3 phase available I dont understand why these controllers were installed.

The jump I mentioned is not a spike it is a very reproduceable DC shift in the signal that would be explained by a DC shift on the DAQ ground.
I do not have AC powerline monitoring equipment but I do know that this sort of problem has been encountered before in this building.
The building AC is unbalanced and noisy.

[AnalogKid2DigitalMan]

Bummer

Would full differential mode on the DAQ improve the situation (thus no ground reference)?

This is where a laptop (running on battery power or floating supply) with the appropriate DAQ card could come in handy.

Maybe high quality AC power conditiong hardware?

Gotta run firewall timeout in 30 sec, be back

[densmore]

Try different outlets. Sometimes you can get lucky or unlucky that all the sockets don't go to the same  power panel.

[WM_John_Weeks]

You might try powering your DAQ device (which means the computer) from an isolation transformer.