Signal Conditioning

cancel
Showing results for 
Search instead for 
Did you mean: 

Jumpy readings 4-20mA

I have some 9102 CRIO modules reading thermocouple data somewhat accurately except that the readings tend to spike randomly. Is there something in the FPGA/RT or CRIO hardware that could be attributing to this noise?  (+-10 degrees is acceptable but in some instances I'm seeing jump of over 60 degrees!)
 
Close up:
 
0 Kudos
Message 1 of 7
(5,730 Views)
craige,

Which cRIO modules are you using? the 9102 is the cRIO backplane.  Really the only suitable cRIO module for reading thermocouples is the 9211, no other modules have the resolution or input range necessary to take accurate measurements for normal thermocouples...

Message Edited by Doug M on 01-18-2007 05:04 PM

Doug M
Applications Engineer
National Instruments
For those unfamiliar with NBC's The Office, my icon is NOT a picture of me 🙂
0 Kudos
Message 2 of 7
(5,712 Views)
My mistake, I'm using the 9203 (8 channel, 4-20mA module, 16-bit resolution...etc)
0 Kudos
Message 3 of 7
(5,706 Views)
ok, are these normal thermocouples? Or are they some other sensor type that converts thermocouple readings to a 4-20mA reading?

Message Edited by Doug M on 01-18-2007 05:38 PM

Doug M
Applications Engineer
National Instruments
For those unfamiliar with NBC's The Office, my icon is NOT a picture of me 🙂
0 Kudos
Message 4 of 7
(5,703 Views)

Type K-thermo couples are fed into Honeywell Digital Recorder. The HoneyWell applies a programmable scaling

0 degrees = 4mA

2200 = 20mA

Since I am not onsite I am waiting to hear from my customer if the noise is seen at the output which runs to the CRIO. (50 degrees equates to +-.360 mA noise is this case). But before I sit back and twiddle my thumbs I wanted to make sure my set up wasn't the problem.

 

Thanks for your response,

Craig

0 Kudos
Message 5 of 7
(5,698 Views)
The operating instructions for the 9203 list the absolute accuracy as .052% of the range when the module is calibrated, which is much less than the .35 mA you describe.  I would take a look at the specs of the honeywell device to see what its resolution and accuracy for thermocouples are.  Another way to test the accuracy of your module is to use a power supply to output a known accurate current and see what noise you see from the 9203.
Doug M
Applications Engineer
National Instruments
For those unfamiliar with NBC's The Office, my icon is NOT a picture of me 🙂
0 Kudos
Message 6 of 7
(5,696 Views)

It looks like noise induced on the cabling, possibly due to insufficient grounding.. One easy way to resolve this is to add some filtering in the FPGA on the incoming signal. LV FPGA 8.2 includes an easy-to-use Butterworth filter that you can add right after your FPGA I/O node. Each Butterworth filter VI is configured in a dialog to set the sample rate and cutoff frequency.

 

Message Edited by Christian L on 01-19-2007 04:07 PM

authored by
Christian L, CLA
Systems Engineering Manager - Automotive and Transportation
NI - Austin, TX


  
0 Kudos
Message 7 of 7
(5,675 Views)