제목: Re: DAQ noise in AIGND

[labmaster]

One high impedance (microphone ~1.5kohm) and other one low impedance(pressure) transducers are used in DAQ.

(Anyway, forget the low impedance transducer. and I have to use very low ripple (~2mV of 15V) power supply for precise measurement.) 

Because of high impedance, I used two op-amp followers in each line in differential mode in channel 0/8.

NI recommend me to use 10k~100kohm bias resistor to AIGND for +/- lines.

In this case, the noise (about 60Hz) of higher than 20 mV level could be observed in my oscilloscope and DAQ result.

(I didn't measure the signal by oscilloscope and DAQ simultaneously.) 

In oscilloscope, by disconnecting AIGND terminal, I can reduce the 60 Hz noise  level less than 5 mV.

How to isolate the noise to AIGND in DAQ?

메시지가 10-23-2008 12:18 PM

에 labmaster에 의해 편집되었음

[Hillman]

Hi Labmaster,

Thanks for the post. I notice you hadn't had any support.

So when you disconnect the AIGND terminal are you leaving it to float? 

This is the standard advice: 

 And are you asking how to disconnect the AIGND in DAQ? i.e. how to elimate noise in software?

Kind Regards, 

[labmaster]

Thank you, Hillman:

When I connected to AIGND through the bias resistor, I found the electrical noise in oscilloscope.

In DAQ, like NI recommendation of differential and floating singal you attached,

I have to connect to AIGND because of removing the crosstalk and drift up to saturation voltage in multichannel scanning.

I don't want to elminate the noise (comparable to singal amplitude) in software for singal modification.

How to resolve my problem?

Lee

[Hillman]

Hi Lee,

Im still struggling to under your issue - can you provide screenshots?

Here's a great guide to reducing noise:

Field Wiring and Noise Considerations for Analog Signals

http://zone.ni.com/devzone/cda/tut/p/id/3344

Let me know what you think,

Kind Regards

[johnsold]

Lee,

One possibility is that you have a ground loop.  DIfferent parts of your system get power from different sources and current flows through a grounding conductor somewhere.  Due to the impedance of the grounding conductor voltage drops occur along that conductor and the voltage or potential is not identically equal to zero at all points which are supposed connected to "ground."

You mentioned pressure transducers and a microphone.  They must be connected to or mounted on some apparatus.  You also have a low ripple power supply, signal conditioning amplifiers, and your computer.  This is enough equipment to produce ground loop problems.  Single point, "star" grounding is the first step in eliminating ground loops.

Please describe your system so that someone may be able to offer suggestions.

Lynn