Help troubleshooting flow meter AI readings (9215 AI module)

[fraffan]

Hello all,

I have an instrument (wired by someone else) that  has 2 flow controllers and one flow meter. The controllers are controlled via a 9263 AO module (one channel for each controller), and the reading of the controller flow as well as the flow of the meter is read by 3 channels of a 9215 AI module. I recently encountered an issue where the signal of the meter exhibits a low frequency high amplitude oscillation (amplitude of about 0.1 V with no flow). I tried some troubleshooting and managed to isolate a possible cause. In the original wiring, the COM of the AO module is jumped to the AI- for both controllers. If this connection is broken, the reading from the meter becomes as expected. However, I feel that this would then render the controllers useless since I would not be able to control them (There is no wire going from the AO COM directly to the controllers). The strange thing is that this instrument has worked for several years with this wiring and it's only recently that this has been observed. Any insight is welcome. I am attaching a quick diagram as to how the modules are currently wired with noise and without noise as well as a plot of what the signal looks like for an expected reading of 0 volts (no flow) with and without the oscillation. 

[Henrik_Volkers]

Sounds like ground loop problems/currents.

A detailed schematic of your configuration, including the power supplies , sensors (type, link to datasheet?) (including all wires 😉 ) would be needed.

One cause of a low frequency drift (BTW the graph lacks of scale /plot information):

If you have line noise and minimize it by (coarsly) reading with line frequency (or a n/m fraction of it), the variance in line frequency can be the cause for low frequency drifts.

[cweeks12]

@fraffan

Looking at this, I agree that it could possibly be a ground loop issue. Here's an NI white paper that explains a bit about ground loops and how to avoid them: http://www.ni.com/white-paper/5362/en/ .

[fraffan]

Thanks for the insight. I have read this, and will still compare to the current wiring. To the best of my knowledge, the person who wired the setup originally took al of this into account and the setup has worked without issues for at least 5 years. It just started doing this recently when trying to calibrate the flow controllers/meter. I will upload some extra info in case you think you can help further. Much appreciated. 

[fraffan]

Thanks. I'll do my best to give more details in upcoming posts. For now, the instruments are 2 MKS G series controllers and an MKS G series meter. These use 15 pin cables, where the relevant pins are 2, 8 and 12. Pin 2 is the signal readout (which goes to the AI module to read the flow), pin 8 is the setpoint which goes from the AO module to set the flowrate (only to the controllers, not to the meter), and pin 12 is the reference ground for both the setpoint and the reading as per the manufacturer. This is why the "negative" or ground from the AI has to be bridged over to the AO ground so that both have access to pin 12. As a side note, we were troubleshooting today and when we removed all wires from the AI module (essentially leaving all channels open), Channel 1 was reading -10 V, whereas channel 2 and 3 were reading close to 0. I would expect all 3 channels to read the same, so I am now suspecting either a module or chassis issue.