how to reduce analog signal noise

[hjc]

Hi,

Currently facing a lot noise in my system during an analog measurement. My analog input signal is connected in the NRSE mode, that is the measurement device voltage out is connected to one of the analog input and the 0V to the AI sense. Beside that, I've an analog output signal connected to a pressure regulator in the RSE mode.

Is it necessary to connect the AI sense to the AI ground with a biasing resistor?

How about the AI input, does it also require the same biasing resistor?

Is 10K ohm (0.5W) sufficient?

Or is there any better way to connect a single analog measurement signal? 

Any advise is greatly appreciated! Thanks in advance!

Jason

[J_Thomas]

Hi hjc,

Thanks for posting to the NI discussion forums.  A great place to start would be to read this document:

Field Wiring and Noise Considerations for Analog Signals

In particular, pay attention to the nice summary table (Table 1 in the document) that shows all of the basic configurations methods that you can use with your device.  In general, a differential measurement system is preferable because it rejects not only ground loop-induced errors, but also the noise picked up in the environment to a certain degree. 

If you do use NRSE, and you are measuring a floating signal source it is recommended that you have 10k – 100 k Ohm resistor between AI and AI GND and between AI SENSE and AI GND.

If the input signal is DC-coupled, only one resistor connected from the (–) input to the measurement system ground is required to satisfy the bias current path requirement, but this leads to an unbalanced system if the source impedance of the signal source is relatively high. Balanced systems are desirable from a noise immunity point of view. Consequently, two resistors of equal value—one for signal high (+) input and the other for signal low (–) input to ground—should be used if the source impedance of the signal source is high. A single bias resistor is sufficient for low-impedance DC-coupled sources such as thermocouples. If the input signal is AC-coupled, two bias resistors are required to satisfy the bias current path requirement of the instrumentation amplifier.

[hjc]

Hi Thomas,

Sorry for the late reply, as I was busy trying and understanding the various mode of measurement.

I've tried adding the bias resistors (10K) in the NRSE mode but still see a lot of noise in the reading and the value shown doesn't reflect correctly even I increase the pressure in the measuring device (This is not true previously before adding the resistors which the reading still quite close to the applied pressure but fluctuating with noise). Even after I remove the resistor, the reading does not reflect correctly also which I can achieve previously in the NRSE mode.

I've tried to connect using RSE mode instead, and the reading go back close to the applied pressure but still fluctuating with noise.

I've do up a simple sketch of my current circuitry in the RSE mode with x01 analog input and output connection.

Please advise me further on how can I eliminate the noise in my system as I now confuse with the problem I had with NRSE previously which can work.

Do let me know if you need more info.

Thanks again for your help.

Regards,

Jason

[J_Thomas]

Hi hjc,

In general, a differential measurement system is preferable because it rejects not only ground loop-induced errors, but also the noise picked up in the environment to a certain degree. Have you tried setting the measurement up in differential mode? (again, you can refer to the field wiring document linked above for wiring assistance)

If a differential reading still gives erroneous readings, try reviewing the Solving Noise Problems in Measurement Setups section in the field wiring document or the suggestions in the following links:

(I have included several, though some of the information is repeated):

Data Acquisition: Troubleshooting Offset, Incorrect, and Noisy Readings

Troubleshooting Unexpected Voltages, Floating, or Crosstalk on Analog Input Channels

If none of the suggestions in these documents work, and you have double checked your connections and setup, it would be helpful if you could provide some more information about your setup:

1. What kind of sensor do you have connected to your DAQ card? Do you have data sheet for it? What is its impedance? What is the voltage level output of the sensor?

2. Can you describe the noise fluctuations in more detail?  What is the expected level of the signal you are measuring? How much does the signal fluctuate?

3. How are you measuring the signal? In Measurement & Automation Explorer (MAX)? In LabVIEW?

4. Do you still the get the same kind of noise if you measure a different signal? For example, do your readings still fluctuate if you measurement the voltage on a battery or another source with a known voltage?  If you have an external DMM, does it show the same level of noise if you measure the signal from your sensor?

Message Edited by J_Thomas on 04-29-2008 03:08 PM