Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

How to read thermocouple output with NI USB-6009

So I have simple circuit setup using an AD 595 (link) IC chip with an output of 10mV/1deg C.  I have a type K thermocouple and I'm trying to read the output analog signal into a USB-DAQ.  The problem I have is that there is always an offset from what the appropriate voltage signal should be.  Using an oscilloscope or multimeter, the voltage is about 240mV which matches with the room temperature reading (23-24deg C).  When I use the test panel in MAX, I receive a signal output of 330-340mV.  I can't seem to figure out what causes this offset and when I try to cool the thermocouple probe, it does not decrease the voltage reading, but if I heat it up the output signal responds accordingly.  So there appears to be a minimum voltage level of about 340mV.  Does anyone know what may cause this issue?  Thanks.
0 Kudos
Message 1 of 21
(8,387 Views)

There is resistive biasing on the inputs that can raise the measured voltage of a signal.

 

This is especially noticeable with higher output resistance signals.

 

This is explained in the manual, but the summary is that the design uses resistors to bias the open-circuit voltage on the 6009 up to around the middle of the conversion range so that differential signals can be measured.

 

Try using differential inputs, with the + input connected to your AD595 output and the - input connected to the AD595 ground.

Message Edited by ned_konz on 09-24-2009 09:17 AM
0 Kudos
Message 2 of 21
(8,371 Views)

"Try using differential inputs, with the + input connected to your AD595 output and the - input connected to the AD595 ground. "

 

I have attempted this, but it did not result in any change.  I do not understand why there is a lower cutoff limit of 330mV.  I have used a lighter to heat the thermocouple probe and it responded just fine.  Is it a grounding issue?  I have searched these forums and saw mentioning of this.  Any other suggestions would be most helpful, thank you.

0 Kudos
Message 3 of 21
(8,347 Views)

If you could show a schematic of how you've got this hooked up, it would help.

 

Actually, a schematic of the whole thing would help.

 

If you have this on a breadboard, you could hook the power supply inputs to your circuit to the 5V and ground terminals on the 6009 (I believe that the supply should be able to source a couple of mA easily).

 

Grounding shouldn't be an issue if you have no external power supplies hooked up, etc.; however, when you have multiple systems, it can be an issue. Which is one of the nice things about a differential connection -- it can help if you have small amounts of ground difference.

 

Have you done a "sanity check" and just tied the inverting and non-inverting inputs of your 6009 together and then read the input (remember that you have to re-configure the 6009 to do differential input)?

0 Kudos
Message 4 of 21
(8,337 Views)
NI USB-6009 does not have thermocouple cold junction compensation. What you need is to measure the cold junction temperature using another device and process the voltage with the hot junction voltage.
0 Kudos
Message 5 of 21
(7,981 Views)
The AD595 does cold junction compensation (and amplification).
0 Kudos
Message 6 of 21
(7,962 Views)

As I suggested earlier:

 

Have you done a "sanity check" and just tied the inverting and non-inverting inputs of your 6009 together and then read the input (remember that you have to re-configure the 6009 to do differential input)?

 

0 Kudos
Message 7 of 21
(7,961 Views)

Sorry for the late reply, but this issue has now surfaced again with me doing the same testing.

 

But yes I have tied the two inputs together and get about a 1 mV reading, so that does not seem to be an issue.  I'm almost running out of ideas and I swear my lab partner and I was using this same setup just the other day and got the correct results.  Trying to figure out what could be wrong.

Message Edited by mdzz on 03-23-2010 11:39 PM
0 Kudos
Message 8 of 21
(7,833 Views)

Could you post your schematic and maybe a picture of your setup?

 

I'm out of ideas, frankly.

 

Have you put an oscilloscope on your USB-6009 inputs? Maybe you have oscillation. Have you bypassed the power supply pin 11 to ground with a capacitor? Though the example circuits in the data sheet don't show capacitors, I'd certainly add one.

 

I'd recommend something like the AD595 data sheet's basic connection circuit (Figure 1) (shown here), with the addition of a 0.1μF capacitor between pin 11 (the power supply input) and pin 7 and 4 (ground).

 

Your USB-6009 should be connected differentially:

  • the USB-6009 + input to pins 8 and 9 of the AD595
  • the USB-6009 - input to pins 7, 4, and 13 of the AD595
  • ground of the USB-6009 also to pins 7,4, and 13 of the AD595 (run a separate wire over to the AD595 circuit; don't just connect the USB-6009 ground to the USB-6009 - input)
  • the USB-6009 5V output to pin 11 of the AD595 (with a 0.1μF capacitor added near the AD595 between its pins 11 and 7)

 

AD595BasicCircuit.png

 

If you duplicate the schematic given in the AD595 data sheet as Figure 13 (shown below) (connect pin 7 to ground) you will have a simple thermometer (this circuit replaces the thermocouple with a short circuit). Does this work? If you heat the chip, do you see the output increase?

 

AD595AsThermometer.png

Message Edited by ned_konz on 03-24-2010 07:00 AM
Message Edited by ned_konz on 03-24-2010 07:01 AM
0 Kudos
Message 9 of 21
(7,823 Views)
photo1.jpg
 photo.jpg
 
 
Hopefully you can see the wiring from these two picturesThis is using the basic connection along with the suggestion of adding a capacitor between the supply voltage and pin 7I still haven't resolved the issue though.  Using a multimeter or oscilloscope provides the proper mV reading, but I cannot figure out why the DAQ is susceptible to this lower cutoff.
0 Kudos
Message 10 of 21
(7,802 Views)