I want smooth thermocouple signal

[Coq_Rouge]

Please post a snapshot of your noisy measurement. Without some idea of how your noise looks like we can not help you. You can attach your picture to a message that is not a problem.

[sahil]

hi.. i think you have come wid the answer ..

now please solve my problem??? 

 i want to acquire data in a text file through DAQ-NI 9234..i m able to get this will using other sources (other than DAQ-NI 9263).

but when i give a signal to NI 9263 then acquire the same signal i am not getting reliable results...

[Coq_Rouge]

This thread was started by instail, so are you and sahil working with the same project, which is OK. Or has sahil just jumped in and hijacked this thread???. Hijacking threads is this not acceptable behavior.

[sahil]

ok allright..

there is nothing like hijacking..i was just trying to get my problem solved..

can u solve my problem..

[Coq_Rouge]

You have asked that question here http://forums.ni.com/ni/board/message?board.id=170&thread.id=432876. Be patient and it will be solved in that thread 😉 

[TSDAN]

Thank you all for reply

this is the temperature wave I got 

thank you

[Coq_Rouge]

Can you please send code also. Are connecting the thermocouple directly to your daq unit?

[Joseph_Loo]

Unless I am wrong, yu are using a USB DAQ 6008 A/D converter. I would summarize that you have it set at 2.5 volt range with a direct coupling. This is inadequate to use directly with a thermocouple. The resolution, would probably give you +/10 deg (even at the 1 V scale) resolution depending on the thermocouple type (which you have not mentioned). This will give you that noisy spikes you are seeing. In addition, if you are using the single ended mode, you loose 1 bit.

[Coq_Rouge]

Index Temp degC       EMF mVolts
0     20.000              0.790
1     30.000              1.196
2     40.000              1.612
3     50.000              2.036
4     60.000             2.468
5     70.000             2.909
6     80.000             3.358
7     90.000             3.814
8     100.000          4.279
Here is some voltages for a T type thermocouple. '

So you have to introduce some gain in your application. The gain stage shall be tuned for the temperature range you want. Say you want to have 100 degC as your max temp. Then the gain should be (1 volt/4.279e-3) given that you are using the +/- 1 volt range

[Alain_S]

Allow me to do some math's from one of the documents I suggested you at the start 🙂

Code width (Resolution) =  This represents the smallest change in a signal that the DAQ board can detect. = 1.53mV at +/-1V range (from the datasheet)

Sensitivity of the thermocouple  = This represents the approximate change in voltage for a 1 unit change in temperature. = 3.489mV/80°C = 0.0436 mV/°C (for your type T TC in a 20 to 100°C range, data provided by Coq Rouge)

Minimum Detectable Change = This approximates the smallest temperature change that the DAQ board can detect. = 1.53mV/0.0436mV/°C = 35.0817°C 

From your graph one can see that you have about 13 bits for a temperature range of 75°C, meaning 1 bit is about 5°C.

That's seven times better than theory :).

Do you already apply some amplification?

Either way, a TC directly connected to a USB 6008 is something you better never do

You could of course amplify your TC signal by let say 200 so that for 100°C you get about 855.8mV but amplifying a TC signal 200 times isn't a sinecure, especially if you don't have experience in electronic design.

That's why I suggested a ready made amplifier like the Omega one.

Message Edited by Alain S on 08-17-2009 07:43 PM