Trying to get an accurate reading from thermocouple

[Eric-E]

You actually cannot do a differential measurement with this device. So this means that you should be just fine connecting the Omega Thermocouple Converter on COM and ai0. You should not be seeing this absolute value of the sine wave though. I think this is because you may be using a pseudo-differential task when this is not appropriate for this device.

With the thermocouple connected to the Omega Converter connected to your 9201, please provide a screenshot of what your data looks like if you view it in Test Panels in MAX:

1) Click on your 9201 in MAX and click on Test Panels at the top of the screen.

2) Make the Mode Finite and Rate something like 10 and samples to read 100 (this will acquire 10samples/sec for 100 samples which is 10sec of acquisition)

3) Channel Name should be set to your 9201 device name and channel (whatever the + cable of the Omega Converter is connected to)

4) Input configuration should be RSE (note all others are not an option).

5) leave the inputs limits as 10 and -10V.

With a simulated 9201 it looks like this because the default simulated signal is always a sign wave with a little noise added:

 

Please perform the above steps and post a shot of the test panel for us to really understand what's going on. You should see something reasonable, but keep the following in mind. To find the code width of this device take the range 20V and device by the 12-bit (2^12) resolution: 20V/(2^12) =5mV -> 5degC. Therefore we the best theoretical difference we can measure will be 5degC. Add noise to that and our accuracy is not that great. Just keep this in mind for the expectations of this device.

I hope this helps. Thanks!

[iinWill]

I attached the picture of the test. I had the NI9201->ai0->+Output Omega->/COM->-Output Omega. I have the Omega converter switched ON and connected to a thermocouple (K type). I'm pretty sure there's an problem with my results. 

[johnsold]

The image you posted is a fairly typical +/-2 LSB noise on about a -9 mV signal.  I did not look up the noise specs on your device but nothing there looks like a problem.

Note that with a 10 Hz sample rate, any power line frequency interference will be aliased.  This will make it much more dfiifcult to identify.

Lynn

[iinWill]

---

@johnsold wrote:

The image you posted is a fairly typical +/-2 LSB noise on about a -9 mV signal.  I did not look up the noise specs on your device but nothing there looks like a problem.

Note that with a 10 Hz sample rate, any power line frequency interference will be aliased.  This will make it much more dfiifcult to identify.

Lynn

---

What are some good VI i could use to get a temperature reading for the OMEGA?

[johnsold]

As has been pointed out earlier in this thread, the DAQ device you are using does not have enough resolution to give you good results with a 1 mV/degree signal. You cannot fix that in software.

You could use an external amplifier, but then you need to worry about offset voltages and gain calibration. Or you could get a DAQ device with better input voltage resolution.

Lynn

[iinWill]

---

@johnsold wrote:

As has been pointed out earlier in this thread, the DAQ device you are using does not have enough resolution to give you good results with a 1 mV/degree signal. You cannot fix that in software.

You could use an external amplifier, but then you need to worry about offset voltages and gain calibration. Or you could get a DAQ device with better input voltage resolution.

Lynn

---

Any suggestion for a DAQ that would work with the OMEGA? No specification, just to be used for OMEGA. Would an USB Thermocouple be better? 

[johnsold]

It may depend on other requirements.  If you want to stay in the CompactDAQ form factor, the 16-bit modules will give ~300mV resolution on a +/-10 V range.  This is equivalent to 0.3 degrees.  There is also a 24-bit voltage module which would have much finer resolution.

The 4-channel thremocouple module is preced lower thatn the device you are using now.  Of course you would also need to obtain thermocouples.

Lynn

[iinWill]

---

@johnsold wrote:

It may depend on other requirements.  If you want to stay in the CompactDAQ form factor, the 16-bit modules will give ~300mV resolution on a +/-10 V range.  This is equivalent to 0.3 degrees.  There is also a 24-bit voltage module which would have much finer resolution.

The 4-channel thremocouple module is preced lower thatn the device you are using now.  Of course you would also need to obtain thermocouples.

Lynn

Would the NI USB-TC01 with a Omega K-type Probe be fine? 

 

 

 

[johnsold]

I have not used them but the specs look OK.  The maximum sample rate is 4 samples/second. They recommend insulated or ungrounded thermocouples.

Lynn

[iinWill]

---

@johnsold wrote:

I have not used them but the specs look OK.  The maximum sample rate is 4 samples/second. They recommend insulated or ungrounded thermocouples.

Lynn

---

Let me recap, the NI 9201 DAQ will have a temperature error around 10C, including the error from the Omega Convertor, but not including signal interferances? So does that mean the 9201 is usable, but with a large error? Because when I use the VI on acquiring voltage, it would always fluctuate from a large negative number, like -200C to a large positive number, like 500C, which made no sense to me, would that be the VI issue or the DAQ issue? I can't seem to understand why the numbers I get from the VI are so big, when I'm trying to measure temperature from a cold metal plate. 

Also the Omega Thermocouple Probe is insulated.