U3 with PT100 RTD sensor
Posted 26 April 2006 - 06:56 PM
Posted 27 April 2006 - 08:23 PM
The problem with an RTD is that the sensors themselves might be very accurate (resistance versus temperature), but they simply provide a resistance that varies (just slightly) with temperature, and it is often difficult and expensive to determine the resistance of the RTD with similar accuracy. It is much easier to use an accurate probe like the EI-1034 (available at labjack.com) with a high-level voltage output.
The typical way to measure resistance is to put a constant current through the resistor (RTD in this case) and measure the voltage that results. The post above points out some of the easy sources of error than can cause problems. I see 3 major sources of error:
1. Accuracy of the resistance versus temperature for the RTD.
2. Accuracy and stability of the constant current source.
3. Accuracy of the LabJack's analog input.
Assume you want an accuracy of 0.03 degrees C from -200 to +850 degrees C as mentioned for one class of RTD in the above post. That means 1 part in 35000 or an accuracy of 0.003% full scale. All those errors above have to combine to less than 0.003%. It seems that it would be tough to find a constant current source that is sufficienty accurate and stable, and that accuracy is beyond the U3.
Another way to handle RTDs is with a bridge circuit, but this has just as many problems and complications as the constant current source.
An RTD can certainly be used with a LabJack, but just because you are using an RTD that is accurate to 0.03 degrees C, do not expect your overall system to easily achieve that same accuracy.
Beyond creating your own current source or bridge circuit, following are some commercial signal conditioning units that can be used to interface an RTD to a LabJack analog input.
The TX94A from Omega is an option with 4-20 mA output. See Section 184.108.40.206 of the U3 User's Guide to handle the 4-20 mA signal:
5B34 or 5B35 from Analog Devices have simple voltage outputs and good accuracy:
Posted 01 May 2006 - 01:20 AM
Posted 17 May 2006 - 09:14 AM
Posted 12 January 2009 - 07:57 AM
Posted 12 January 2009 - 10:04 AM
This equation does have a lot of simplifications, but is a good place to start. Later if you need you can back up to the full unsimplified bridge equations.
Since you only have one active element (say dR1), the other 3 delta values are 0 and the equation is just:
dVg = (dR1 * Vin) / 4R
That lets you solve for the resistance of your RTD, and then you can use the resistance-to-temperature relationship provided for your RTD to get temperature.
Posted 17 February 2009 - 05:22 AM
Posted 17 February 2009 - 08:35 AM
A raw RTD connected directly to a U3-HV will not do anything useful. You need some signal conditioning for the RTD, as discussed above.
If the RTD probe already has signal conditioning built-in, and is providing 4-20 mA output, then yes the LJTick-CurrentShunt is good for monitoring a 4-20 mA signal.
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users