Significant signal oscillation with USB-6002

[AtlasCH]

I currently have a working setup to measure and adjust the pH of a solution using a pH electrode and peristaltic pump connected to the USB-6002. I am also using a pH sensor amplifier to amplify and treat the high impedance millivolt signal coming from the pH electrode. When measuring and controlling the pH in a usual benchtop environment (glass beaker & magnet stirrer), the results look quite promising where stable and accurate measurements are achieved. However, when I want to run measurements on a solution in a reservoir of a TFF system, i get intense oscillations spanning from pH 2 to pH 8, making the whole setup unusable in this configuration. I was able to get stable and accurate readings only when disconnecting the power supply of both the laptop and pump connected to the DAQ device which is of course suboptimal. I also took measurements of the solution outside of the reservoir in the usual configuration and the measurements were once again stable and accurate.

Some more information about my current setup: I have a Hamilton EasyFerm PLUS PHI S8 120 pH electrode connected to a amplifier module (https://www.aliexpress.com/item/1005006366896348.html?src=google&pdp_npi=4%40dis%21CHF%215.79%215.79...) via a S7/BNC cable. The amplifier is powered using the 5V power supply of the USB-6002 and the analog output signal of the amplifier is sent to the USB-6002 via differential input. The peristaltic pump used is a LabQ pump (https://www.drifton.eu/shop/9-flow-rate-peristaltic-pumps/2525-labq-peristaltic-pump/) and is connected to the USB-6002 through the AO 0 and AO GND channels. The USB-6002 itself is connected to a laptop running LabVIEW.

Is this a grounding issue or due to some other kind of interference? What would be the best way to fix this issue? Thanks in advance.

[Leonard_Gartner]

Hi AtlasCH,

The pump is only part of the TFF system, while the simpler benchtop system does not include the pump, right?
Do you have to remove both the laptop and pump power supply to get better measurements? Or is it enough to remove the pump power supply? Does the power supply of your pump and laptop include ferrite beads to reduce electromagnetic noise? If not, this might help.

Another idea:
If your noise is high frequency but your signal is DC/low frequency you can eliminate the noise using a low pass filter on your laptop. You can use python with numpy for that.

Let me know if this helped or if there are further questions!

Best regards

Leonard

[AtlasCH]

Hi Leonard,

Thank you for your quick response. I'm sorry for my late reply, I was away the last few days. If you are talking about the dosing pump then no, that pump is always part of the main setup (DAQ device, pH electrode, pH signal amplifier, pump, laptop) and was also used in the benchtop runs: 

The TFF system itself of course also has built in pumps but those are not connected with my dosing setup. Concerning signal quality, this were the observations made in regards to which power supply was connected:

Concerning the power supplies, for the laptop I'm using a 65W standard AC adapter (USB type C) from Lenovo which does not come with an external ferrite bead but should still offer internal electromagnetic noise reduction from what I know. For the pump, I'm using a Mean Well NGE30E12-P1J Plug-in power supply that also doesn't come with a ferrite bead attached.

The oscillations themselves are quite low frequency however as you can see from the graph above so I'm not sure a low pass filter would help here.

I'm assuming that maybe the grounding or wiring could be an issue as when i fully unplug the TFF system, the measurements become once again stable and precise. I assume the main difference that is causing the issues is that the glass beaker in the benchtop setup is electrically isolated while the reservoir of the TFF system is connected to the system and thus there is an electrical potential causing these major oscillations. The amplifier I'm using is also not electrically isolated so that could also be a point of improvement.

Thanks again for your inputs, let me know if you have any more ideas.

Best regards

AtlasCH

[Leonard_Gartner]

Hi AtlasCH,

thanks for the graphs, that gives me a clearer picture!

A lowpass filter could still work, depending on implementation. For example moving average or butterworth filter. This would be done digitally on your PC, for example in Python. You would need to capture sufficiently many samples for the filter to have the desired effect.

I am not sure wrt the ground loop, since I dont know where you have multiple different ground connections. 

How is the reservoir electrically connected to the rest of the system? In my imagination, the reservoir is just

some glass vessel containing the liquid to measure. Would it be possible to provide a quick sketch of the setup?

After seeing the schematic of the amplifier: Did you connect both the power negative and signal ground to the USB-6002 GND?

Best regards

Leonard

[AtlasCH]

Hi Leonard,

Once again thank you for the quick response.

I will keep the lowpass filter in mind but I don't have access to LabVIEW at the moment so I will have to try and implement this at a later point in time.

Concerning the reservoir, it is indeed a glass vessel containing the liquid to measure but it also has a pressure-based level sensor built in its base that is connected to the rest of the system via a pressure transmitter. The inlet and outlet are also both in the base:

I also tested measuring the same solution in a glass beaker in close proximity of the TFF system and in the reservoir with the power supply of the system disconnected and in both cases the measurements were stable and precise.

Concerning the amplifier, yes, both the power negative and signal ground are currently connected to the USB-6002 GND. The amplifier is currently powered by the +5V power supply of the USB-6002 so the power positive of the amplifier is connected to the +5V channel and the power negative to the DGND channel according to the USB-6001/6002/6003 User Guide. For the signal, the 0-5V AO of the amplifier is connected to the AI0 channel and the signal ground to the AI4 channel with the input being defined as differential in the application.

I hope this makes things clearer.

Best regards,

AtlasCH

[Leonard_Gartner]

Hi AtlasCH,

Wait, so is the signal ground connected to your differential AI (ai4) or wired to the DAQ GND. Or both at the same time, shorted? Since you stated both in your reply I was not sure 🙂

Could you try to connect it to ai4 only?

Best regards

Leonard

[AtlasCH]

Hi Leonard,

Sorry for the confusion, I was thinking of two different setups where I tried both RSE and differential input. Currently the signal wires are connected via differential input so the 0-5V AO signal from the amplifier is connected to the AI 0 channel and the signal ground from the amplifier to the differential AI (AI 4) channel. So concerning the amplifier, only the power negative is connected to the DAQ GND (D GND). Besides the amplifier, the pump is also connected to the DAQ GND where the signal wire is connected to the AO 0 channel and the ground wire is connected to the AO GND channel.

Once again sorry for the confusion, I hope this clears it up.

Best regards

AtlasCH

[Leonard_Gartner]

Hi AtlasCH,

Regarding the reservior: It doesnt seem to be grounded since all parts are glass / plastic, hance electrically floating.

Regarding the pump: Does it reduce the oscillation if the pump is not connected to the DAQ AOs, and controlled manually ?

Best regards

Leonard

[AtlasCH]

Hi Leonard,

Thanks as always for your response. That's what I assumed as well regarding the reservoir. Concerning the pump, I haven't tested controlling it manually but I would assume that the signal would behave similarly like in the previous graph I posted here:

I'm not sure how I could wire the setup differently without sacrificing the automated ph regulation aspect of whole setup (e.g. manually controlling the pump). I'm considering either using a battery to power the amplifier or swapping the amplifier alltogether with one that is galvanically isolated but I'm too inexperienced with electronics to know if that would actually help with my issue.

Best regards

AtlasCH