NI-DAQmx Base: Thermocouples

[httcnj]

Hi. I'm using NI-DAQmx Base for the mac on LabVIEW 8.2 and I have trouble with getting the voltage differential for the k-type thermocouples of the DAQ I'm using. Getting the conversion is easy by using the thermocouple vi, but somehow the voltage readings don't seem right.

When using the acq. continuous samples example vi or the acquire single sample vi, the voltage reading I recieve for just a regular temperature thermocouple are 0 and -0.00048 (it mostly stays on 0 though); which to me, doesn't look right. Also the CJC voltage reading reads about 0.3 V (I don't really know how this works yet, but here it is if it helps).

A field sales engineer came by the other day and seemed to have it. When tinkering around with the example vi, he kept getting small positive values which look to work. Also, when linked with LabVIEW 6.0, the DAQ system works fine.

Please, any help would be appreciated.

[johnsold]

First, what kind of data acquisition device are you using?  Does it have amplification for thermocouples?  Is the device compatible with DAQmx Base?  Not to be critical of NI field personnel, but few of them are familiar with the Mac and many have not worked with DAQmx Base much.

Your low voltages suggest that you are measuring the millivolt thermocouple outputs on a 10 or 20 volt range. 

Lynn 

[httcnj]

I'm using the SCB-68 with a PCI-MIO-16E-4 which seems to be compatible with base. I find that its connected by looking at the lsdaq application. I don't really know about the amplification for thermocouples, but I think that would be unnecessary if the voltage outputs are correct since LabVIEW can probably do the amplification in the program.

I find that its neccessary to have low, positive voltages for the readings after reading a lot about thermocouples for the range of values I'm seeking, but the voltages I read from the VI are low but negative or 0. 

Is there a way of possibly troubleshooting this?

[johnsold]

The most sensitive range appears to be 100 mV with 12-bit resolution.  This leads to a least significant bit size of ~24 microvolts which should be adequate for some thermocouple measurements.  A type K thermocouple produces about 39 microvolts per degree at 0 degrees C.  The CJC output is 10 mV/degree C.  So the CJC output will need to be measured at a different sensitivity than the thermocouple channels.

In your original post you mentioned getting readings of zero and 480 microvolts.  That suggests that you are using a 2 volt range (assuming that you are seeing a 1 LSB shift).  That also implies that the smallest temperature change you could detect would be about 13 degrees.

Please tell us what voltage range(s) and which type of input connections (DIFF, NRSE, RSE) you are using. 

Lynn 

[httcnj]

Oh... now I understand the CJC reading better now. Thanks for that. Looks like the 300 mV reading is fine since its just around the temperature I'm working in. ~30 degree C.

I tinker with the voltage ranges somewhat. Going from -10 to 10 or to 0 to 5 (which I think is supposed to be the appropriate range for the temperatures I'm seeking) and I'm using DIFF input connections.

[johnsold]

What are you measuring?  The center of the sun?  What kind of thermocouple do you have?  The highest thermocouple output voltage listed in my Omega catalog is about 70 mV and most are around 20-50 mV at 1000 degrees.

Do you have the appropriate bias resistors connected to the differential inputs?

Lynn 

[httcnj]

I'm using a K-type thermocouple. So, what you're saying is to decrease the range of voltages measured to around 0 to .01 or something like that?

Also, I'm really don't know if there are the appropriate bias resistors connected to the inputs. Let's just assume so though.

Thanks, this is actually really helping a lot.

[johnsold · Accepted solution]

Type K thermocouples have a maximum temperature range of -200 to +1250 °C.  The output goes from about -6 mV to +51 mV for that temperature range.  So you should use the 0 to 100 mV or +/-50 mV range on your DAQ card, whichever matches best the actual temperatures you will be measuring.

If you did not explicitly connect bias resistors, they are not there.  Look at page 4-6 of the SCB-68 User Manual for Advanced Functions for information on connecting thermocouples.

Lynn 

[httcnj]

Thanks a lot. After applying the converter to the data I acquired (and making changes to what you specified), the temperature looks to be quite appropriate, although a little high. This is definitely a step in the right direction!!

I may come back if I need more help, really appreciated it.