DAQ Assistant: Clock Settings (Samples To Read, Rate) can affect signal readings?

[Cheng-Yu]

Dear all,

I'm totally new to Labview, and recently I get confused with the equipment I'm dealing with.

Technical details:

The Labview version is 7.1; computer operate system is Windows XP;

The equipment has a NI PCI-6220 and a 68 Pin Connector Block to read signals from the equipment;

There are 4 channels in DAQ Assistant (2 pressure reading, 2 temperature reading);

For the first pressure reading, Signal Input Range from 4m to 20m Amps;

Clock Settings are Samples To Read = 5, Rate (Hz) = 20.

Description of the problem:

I use Labvew to monitor and record pressure readings and temperature readings. The Labview configuration was set up by my advisor several years ago. Recently, I found the pressure reading vibrated a lot; for example, 5.01 to 5.05 bar within a second. In order to get a stable pressure reading, my advisor suggested me to change "Clock Settings" in DAQ Assistant from Samples To Read = 5, Rate (Hz) = 20, to Samples To Read = 250, Rate (Hz) = 1000. In this case, she believed that since we increase sample numbers and sampling rate, we could have more data, and thus have stable pressure readings.

At first I could have very stable pressure reading. The last digit (0.01) did not change within 20 seconds. However, somehow after a day the pressure reading became unstable and even worse than previous. (pressure reading vibrates from 5.01 to 5.30 within a second)

This is not the worst case. We found that when we set Clock Settings: Samples To Read = 5, Rate (Hz) = 20, the pressure reading is about 8 bar. However, when we set Clock Settings: Samples To Read = 250, Rate (Hz) = 1000, the pressure reading is about 5 bar. In this case, we even don't know which pressure reading is correct.

Labview records current, and transforms it into pressure reading. Thus my advisor tried to monitor the current reading by Labview, and she found the current reading changed when she changed the Clock Settings. (0.004 Amps (5 bar) when Samples To Read = 5, Rate (Hz) = 20; 0.005 Amps (8 bar) when Samples To Read = 250, Rate (Hz) = 1000)

Since we only change the sample numbers and sampling rate, the average readings should still be similar. However, the reading are not similar. That is what confuses me.

My questions are, if Clock Settings in DAQ Assistant could affect signal readings? If so, how it could affect the signal readings? What is the effect of "Samples To Read" and "Rate (Hz)"? How to determine these parameters to get the true pressure readings?

Thank you very much for your help. Hope to have some feedbacks from you.

Best regards,

Cheng-Yu

[TCPlomp]

A 6220 cannot read a current, it can only read a Voltage, so you'll probably have some (or should have) a resistor accros the voltage input. (normally 50 Ohm for a 0-20 mA signal).
My first step would be to measure this voltage with a multi-meter so you know what the actual voltage should be.

Then I would read that same voltage with MAX (measurement and automation explorer) to make sure you have the right value

Now about the changing of the voltage/current/pressure, how have you terminated the other signals? Have you provided a good earthing?

If you sample with a high frequency (1 kHz), perform an FFT on the acquired data, I can imagine a dominant 50 or 60 Hz (depends on where you live) in the signal that might cause your problem.

Ton

[RavensFan]

It sounds like you might be picking up some 60Hz noise.

With 5 samples at 20 Hz, or 250 samples at 1kHz, either way you are collecting only 1/4 second worth of data.  Do you need a pressure reading that frequently?

If you take more samples at a higher rate, than you should be less susceptible to things like 60Hz noise since you can average the data to minimize it.  Other alternatives are to sample at a higher frequency and run the data through a low pass filter to eliminate any higher frequency noise.

[Cheng-Yu]

Dear Ton,

You are right. 6220 can only read voltage, and I think that's why I found a resistor accross the pressure transducer input. (I'm sorry that I cannot tell what resistor it is because I'm not an electronic guy)

Somehow the problem was gone when I tested it again. Now we believe there is a noise which I believe is what you mention about 50 or 60 Hz frequency.

[Cheng-Yu]

Dear Ravens Fan,

You are right. We guess we have 60 Hz noise in our building.

The reason that we collect data every 0.25 sec is that we hope the equipment can react fast. In that case, if the operator accidentally over-pressurize the system, the equipment can react fast to vent the gas and protect itself.

I will try what you said to increase samples at the cost of fast reaction, or to add a low pass filter. Do you mean to add it in the block diagram right after the DAQ assitant? What configuration do you suggest in this low pass filter?

Thank you.

Cheng-Yu

[TCPlomp]

Yes I would add the low-pass filter right after the DAQ reading. However the best would be to remove the 60 Hz from the signal alltogether, have you removed every possibility of a ground loop?

How does your earthing schedule look like? (if you don't have one, I would start to build one).

Ton