How to amplify thermocouple readings

[sman]

I am writing a program in LabView to read the input of the temperature from the thermocouple, but the thermocouple is outputing a very small voltage and so how do i amplify that? Also, how do I correspond that with the actual temperature.  Thermocouple is type K.

[PauloM]

Hi,

Normally, that is done with signal conditioning outside the DAQ.

NI has some signal conditioners that can do what you need, but they are also available from other brands.

Do a search for thermocouple conditionning on NI site or on google, you find something.

To convert, use the Convert Thermocouple Reading.vi, on Data Acquisition -- Signal Conditionning functions pallete.

Hope this helps,

Paulo

[LesHammer]

> Normally, that is done with signal conditioning outside the DAQ.

To add to what Paulo said, without the signal conditioning module - specifically designed for thermocouples - you will need to handle the cold reference junction yourself.  Anyplace where you have wires (or screws) of different metals joined together you have a thermocouple.  Unless you account for all of these what are you amplifying?  Do yourself a favor and get the unit designed for this.

Les.Hammer@SmartSensorSystems.com

[sman]

Thank you.  It didn't help much, but thanks anyway. Later

[sman]

Thank you!

[G123]

hi,

check out this article it will guide you on how to condition thermocouple signal readings

"Automatic temperature controller: Designing a data logger for a slow cooker"

on  www.atmel.com

G123

[tbob]

To convert a thermocouple voltage into temperature, you would have to build an amplifier circuit using two op amps.  The first op amp would have a potentiometer from the positive input to ground, the second op amp would have a potentiometer from the output to the negative input (feedback resistor).  Then you would have to repeat the following:  Put the thermocouple in ice water and wait a few minutes.  Then set the first potentiometer to read 0 volts on the output of the ciruit (0 degrees C).  Then you have to move the thermocouple to a pot of boiling water and wait a few minutes.  Then set the second potentiometer to read 10 volts on the output (100 degrees C).  This will upset the 0 reading so you have to repeat the process several times.  After doing this, 0 volts will be 0 degrees C and 10 volts will be 100 degrees C.  To convert from volts to degrees, just multiply by 10.  Use a DMM or Analog Input board to measure the volts.  Labview can communicate with the DMM through GPIB.

So you see, this is a long drawn out process.  Instead of using a thermocouple, it would be much easier, and also cheaper considering the time involved, if you were to just buy a temperature device that already does this.

[LesHammer]

>  Then set the second potentiometer to read 10 volts on the output (100 degrees C). 

Except that the boiling point of water depends upon altitude and barometric pressure.  This can be 5 to 10 degrees C in some areas.

Also, this does not account for parasitic thermocouples.  The DMM lead to the first wire of the real thermocouple forms parasitic thermocouple where it touches that wire - as does the return lead from the DMM to the other thermocouple wire.  These two parasitic thermocouples are at the temperature of your terminal block (presumably "room" temperature) - which could vary 10 degrees C or more, depending upon whether you take this outside.  The only way to account for this is to have an isothermal block, measure its temperature with a thermistor, and subtract the parasitic thermocouple voltage characteristic.

On the other hand, if this is meant to measure hundreds of degrees C, rather than just 0-100, then room temperature may be down in the noise (and the calibration routine would need to be adjusted for the wider range).

-- Les Hammer

 This will upset the 0 reading so you have to repeat the process several times.  After doing this, 0 volts will be 0 degrees C and 10 volts will be 100 degrees C.  To convert from volts to degrees, just multiply by 10.  Use a DMM or Analog Input board to measure the volts.  Labview can communicate with the DMM through GPIB.

[Ryan_V[DE]]

Great suggestions here! everyone\, please let me know what hardware you currently have and I'd be more than happy to suggest the most cost effective solution for you.

Regards,

[Sheldon_Stokes]

For non-critical use, I use an Omega D5000 thermocouple logger.  It provides a virtual cold junction and digitizes the 4 thermocouples.  You talk to it through a serial port.  It's also a very easy serial device to work with.  I've done things as sleezy as use a RS232 equiped Protek DMM and a Fluke thermocouple conditioner to read temperatures.  I've used thermocouple data loggers from MadgeTech, but that's not a direct communication with LabView, so it really doesn't count. 

For critical temperature measurements, you don't want to use thermocouples, but rather switch to platinum resistance wire etc.  For those applications, I typically use a low noise amplifier and current source, and a DAQ board to read the output voltages. 

With thermocouples you need HARDWARE signal conditioning to measure temperatues with any resaonable accuracy.  There's a bunch of thermocouple signal conditioning options out there.  NI has a therocouple input on some of their DAQ products I think, but they use a temperature measurement at the terminal board to "fake" a cold junction (read: this is sleezy).

Sheldon