ni-4351 erronous Thermocouple Readings

[Trebouchet]

Had been expierencing problems with our NI-4351 boards, we have several and the only one working is doing so without knowing why.

I took a board and setup a new system witha fresh install of windows XP SP3. I then installed TDAQ744.

It installed and then XP reconized the device and installed its drivers.

I then setup the properties of the 4351 for a TBX68T accessory.

Then, I created a virtual channel for measuring a j type thermocouple on channel 11, whcih is wored correctly.

I set the refrence to user at 25deg C. 

I then got 23.899 DegC, and it woudlnt change as the thermocuple was heated or cooled.

I tried changin the ref to built in, and then the reading went off to 2800 or so.

I repeated this with a different 4351, same problem.

I changed the TBX68T with a CB68T

my guess is a problem with TDAQ744, I may wipe the system and startover and install 7.2

Very disappointed that we were directed to buy these boards a few years ago only to have suppoort for them drop which left us in this bind.

[aNIta B]

Hi Trebouchet,

         I'm sorry to hear that you're having trouble with the 4351.  It sounds like you have things set up correctly, but I'm curious if there's anything other than the driver that's was changed when you moved from your old system to your new one.  Do you remember what your test setup was before?  How many 4351s do you have?  I'm also curious if the S1 and S2 jumper settings have changed in your configuration.  I would highly recommend taking a look at the Knowledge Base article, "PCI-4352 Measurements Are Incorrect When I Enable the Built-in CJC Sensor".  Have you been using CJC?  You may also find this knowledge base article helpful. 

To save yourself some time, I would simply uninstall TDAQ7.4.4 and install 7.2 (if that's what you had working before), rather than wiping the whole system.  You could also use a utility like MSI Blast if your afraid of not having a complete uninstall.  

If you continue to see this issue and would like to post screen shots of both your configuration and your results, that would be quite helpful. Thanks!

[Trebouchet]

One card in the machine with a TBX-68T, the card was in another machine with a CB-68T breakout.

Switches S1,S2 set to off.

Both thermistors (TBX and CBT) verfied as good resisitance, and showing correct voltage drop.

Windows xp fresh install (it takes me 5 minutes). Tring NIDAQ744, 722 now 693.

  Just created a J-Type Thermocouple on channel 11, TC wires not conencted at this time.

With no conenction, we see a temperature reading  0.6603 deg lower then Userset CJC of 25Deg

Connecting the TC makes no change

  Change the User CJC from 25 to 19 Deg C

  Now we read  0.6581 deg lower then user CJC of 19 Deg C

  Changed to Built-in CJC

  Now we measure TC as a high value (open TC)

  Now we change back to user CJC of 25 Deg C

We see a high TC measurement, but its dropping,,,,, slowly

  Dropping some more

  Droppign more

Settled down back to 24.3397.

I have tried this on two seperate system, and had the same results. I pulled out our thermocouple calibrator to generate some TC signals and the system doesnt respond., showing a constant measurement just below cjc if user setpoint is used, or very large value (open) if Built-in CJC is used

[aNIta B]

Hi Trebouchet,

         Thanks for the awesome screen shots and helpful info.  I haven't had a chance yet to set mine up the same, but a thought did cross my mind as something we should probably try.  I think we should look into how/if your thermocouples are grounded- there are two Knowledge Base articles regarding this:

Grounded Thermocouple Measurements

The Need to Grounding Thermocouples when they are already in Differential Mode

Looks like you've done some great troubleshooting- sorry about the inconvenience, but I think we can get this working!

Have a great weekend!

[Trebouchet]

No , grounding is not an issue, Yes its the one thing I ry to rule out first as it is the worst to faind ground faults.

I went furthure and added Nidaqmx8.80 as i have a rio board that will also run in this system.

Now my thercouples still lock in at a fraction of a degree from the user CJC setting, regardless if a thermocouple is even connected.

If i use the built in CJC , it now just pulls down to 0 deg C..

For kickers, i connected different TC's , grouneded and un grounded, no difference.

[Trebouchet]

I swapped the TBX68T with a CB68T , and still have the smae problems

on another note, its not reading voltages either.

I changed my TC calibrator to a voltage ref, and the DAS just reads 0.0000

[aNIta B]

I set this up here as well and was able to witness the 24.339C when reading an open channel with the CJC set to 25.  I believe that result is expected, because the open channel could float and then the reading that we're getting is going to be the temperature reading at our hot junction sensor minus the difference between the hot junction temperature (at the open channel) and the cold junction temperature (taken from the sensor on the terminal block or our manual temperature input).  As for the other odd behavior regarding the termperatures jumping exteremly high that you're seeing:  are you seeing the same behavior on all channels? In your first post, you stated, "Then, I created a virtual channel for measuring a j type thermocouple on channel 11, whcih is wored correctly".  If you set up a virtual channel again, as you had done before, can you get a good temperature reading? Or, are all readings bad all the time?  I'd be very curious as to what's changed between the working measurements and the ones that are giving erroneous readings. 

Hope you're doing well- happy halloween!

[Trebouchet · Accepted solution]

Finally got a chance to setup another system. Even found a spare cable, whihc as it turns out was the problem. It was damaged enough to cause problems intermittly.

[aNIta B]

Wow!  Way to go on figuring that out- sorry that it was such an inconvenience, but thanks for letting us know what the problem was.  Have a good one!