Noise between channels on 6210 DAQ

[jkusch28]

Hello,

I am using a NI-USB 6210 and Labview 2011 to acquire analog signals from 6 sources; 2 LVDT signals from LVDT signal conditioners, and 4 J-type thermocouple outputs amplified through AD594s. The issue I am having is large periodic noise and offsets on all of my thermocouple readings. I have contacted NI about this issue, and they suggested using all differential inputs, and to connect both input leads of each channel to ground using ~10k resistors. I have done this and tried many different configurations (including using RSE inputs, and switching DAQs with a 6211) with no better results. I am simply trying to read in the 6 voltage signals into Labview using DAQ Assistant. After much research in the issue, I have found that it could be due to the scales of the voltages that I am reading in, however this should only be an issue with fast sampling rates, and I am only sampling at 5Hz. The LVDT outputs can be as high as 10V, while the thermocouple outputs are usually on the order of .5V, never higher than 2V. I have tried using op amps to amplify the thermocouple outputs, but this gets very messy with the 4 temperature readings and I was not able to get that method to work properly. I have also read "Field wiring and noise considerations for analog signals", but was not able to get a good result.

I am wondering if anyone can help me out with this issue. I do not think the solution should be very difficult since I'm only trying to read in 6 consecutive voltages, however I have not been able to find any fixes yet.

One method I am considering is to use two separate DAQs to read in the samples. I could use the 6210 to read in the LVDT signals since I want 16bit resolution on those readings, and use a second cheaper, maybe 10-14bit DAQ to read in the thermocouple outputs. Then I could simply read in the data from each DAQ consecutively in Labview, after I figure out how to do this. 

I would be very thankful to anyone who could give me some insight into the issue and maybe offer any other options. This issue has been plaguing me for weeks now and I would love to be able to move on with my project with good data collection.

Thank you!

[johnsold]

What is the magnitude and frequency of the "large periodic noise and offsets" that you are measuring? Is it power line frequency, LVDT excitation frequency, or something else?  If you measure the AD594 outputs with an oscilloscope, do you see the same noise and interference? Are the thermocouples grounded?

Lynn

[jkusch28]

Hi Lynn,

The noise is about 50mV on the thermocouple output, and appears to happen consistently in about 45 minute blocks, spaced out by about an hour each. The LVDT excitation frequency is set at 10kHz, and when I do record thermocouple voltages without the LVDTs connected, the noise appears to go away (still in the process of confirming this). If I measure the AD594 outputs with a scope I do not see the same noise as seen when collecting with the 6210. And yes, the thermocouples are all connected to the same ground as the DAQ. Thanks!

[johnsold]

1. You clearly have interference from an external source.  Find out what is running 45 minutes on and 1 hour off. (Or, was that on 45 minutes of each hour?) That device, whatever it is, is the source of the problem.

2. Once you know the source you need to determine the means of coupling from the source to your measurement system.  Ground loops would be very high on my list of suspects.

3. What is the frequency of the interference?  This may help verify the source and may be useful in finding a fix.

4. What type of computer is being used for the data acquisition, desktop, laptop, or something else?

Lynn

[jkusch28]

Hi Lynn,

I have the thermocouple amplifiers powered by the 5V supply from an Arduino. I have the arduino so that I can control the temperatures of a few different systems using PID control and some solid state relays. I have connected all of the grounds together to avoid ground loops, but when I collected data this morning with the arduino not connected to the system, I got good results on all channels. I guess this says that the arduino could be the source of the noise, even though I can collect data on the arduino without the noise I saw in the DAQ. Also, I am collecting the data through Labview 2011 onto my laptop.

If the arduino is the source of the noise, how else could I control the SSRs? I just started looking into using Labview and the digitial output pins on the 6210 to run PID PWM, however this seems to be a little above my Labview experience. I would really like to use the arduino to control the temperatures, but still record the thermocouple outputs using the 6210. Thanks for any help!

[johnsold]

Connecting all the grounds together does not guarantee that you do not have a ground loop. If some device has a gounded power supply or chassis, you may have multiple ground paths.

What provides the Arduino 5 V? How is that source grounded?  The power line ground is intended to provide safety for the power distribution system, for people, and to a lesser extent, to the connected equipment.  It is not intended to be a good measurement system reference potential and usually is not.

The 6210 has software timed digital outputs so it will not be suited to PWM unless you are using frequencies of a few Hz or less.

You should be able to use the Arduino for controlling the SSRs.

You did not answer my question about the period or frequency of the interference.

Do you have an overall diagram of the wiring, especially the power and ground wiring of your system? That could be very helpful in identifying and eliminating the problem.

Lynn