Single-ended sensor with external power supply (NI 9201)

[Robin_Hartley]

Hi, I am hooking up a 0-5 V output accelerometer to a NI 9201 analog voltage input module.

The accelerometer requires between 7 - 36 VDC excitation which I am supplying from a 12V power supply. The power supply is floating - i.e. is not referenced to building ground. It's just a cheap, standard 12V power supply.

Unfortunately, the accelerometer is not behaving as expected - lots of noise and not very responsive, when compared to another accelerometer I have which I know works. I have spent a long time reading up about different ground potentials, single-ended wiring and floating vs ground referenced signals.

As far as I can tell, I am using the correct wiring, as recommended here for a single-ended, floating signal which is not referenced to ground: http://www.ni.com/tutorial/7113/en/

Is there something obvious I'm getting wrong here or is the problem not due to the wiring?

Really appreciate your help with this.

Best wishes,

Robin

Here is the configuration I am using:

[Henrik_Volkers · Accepted solution]

To check the influence of the powersupply, just replace it with a 9V battery.

To reduce power supply noise :

• Add a RC filter (R ~10Ohm, C 100nF in parallel with 10µF) in the powersupply, a choke (ferrite to force I+ and I- of the powersupply to be equal) also can help.
• Use a twisted pair cable from the 9201 and connect the AI-GND/COM at the Sensor (if the sensor has an extra GND connection seperated from power gnd use it!  EDIT: from the picture it doesn't seem so, but if the shield and the GND (black?) wire are connected inside the sensor (check with a DMM), try the shield for power GND and the black for singnal GND.

MEMS come in different tastes 🙂  so maybe the noise it inherent with that sensor.

Using a higher samplerate and applying a lowpass filter and decimation is a common software workaround. 

[Robin_Hartley]

Hi Henrik - thank you for such an informative and helpful answer.

Just by chance, I brought in a 9V battery today and when I tried the sensor, it worked perfectly! This indicates that the problem is due to noise from the power supply, not a problem with the wiring.

Therefore, the solution is to reduce the power supply noise - the suggestions you have made seem perfect for this so I have accepted your answer as the solution. I will now implement the noise reduction techniques you have suggested.

Thank you kindly for your help - it's very much appreciated 😄

Robin

Oh btw, for future readers, this is a piezoelectric sensor called: 8101-0040X-120 by Measurement Specialities - a good low-cost (~£105) sensor with ~6kHz bandwidth. Enjoy! 🙂