Optimizing Thermocouple Measurement Accuracy With the Keysight 34980A

The 34921A/34921T is currently the only multiplexer module for the 34980A that provides an isothermal reference and temperature sensor for thermocouple measurements on the terminal block. However, other multiplexers can be used for thermocouples by using an external isothermal reference panel. This requires measuring the thermocouple voltage as well as the isothermal reference sensor and coverting to temperature using the standard voltage to temperature polynomials in your computer.

When the 34921T terminal block is plugged directly onto the 34921A multiplexer module, the isothermal gradient error is +/- 0.5C due to heat being conducted from mainframe through the Dsub connector pins into the 34921T terminal block. Locating the 34921T away from the mainframe using the Y1135A/36A cables where there is no airflow can reduce this error to +/- 0.1C. A third party external precision isothermal reference panel and sensor can reduce both the isothermal gradient error and the reference temperature measurement error. These panels can be used with the 34921A and the other multiplexer modules for the 34980A.

The 34921T uses a MAXIM DS1631U +/- 0.5 C accurate, surface mount IC sensor located on the under side of the PCA. It is read via a two wire serial link using the lines labeled TSIL that are on the two Dsub connectors. These lines are not accessible to the user on the terminal block.

The thermocouple voltage is measured using the 100 mV DC voltage range. To get the best accuracy, use the default 6 ½ digit, 1 PLC integration time or longer to reduce the effects of noise on the signal. The measurement error is essentially negligible in 5 1/2 or 6 1/2 digit mode with the 34980A. Faster measurements can be achieved using shorter integration times at the sacrifice of some accuracy. See the Additional Noise Error under DC Operating Characteristics in the Specifications section of the 34980A User’s Guide. To convert the voltage measurement accuracy to temperature, use the published Seebeck coefficient for the thermocouple type, the thermocouple voltage tables or thermocouple polynomials available from thermocouple reference handbooks such as those available at omega.com. Note that for absolute temperature measurements, the thermocouple manufacturer’s compliance error must be added. For relative thermocouple temperature measurements (temperature differences), most of this error cancels out for thermocouples made from the same batch of thermocouple wire for small temperature differences. www.omega.com

External reference sensors must be the RTD and thermistor types supported by the 34980A if you want to use the conversion built into the 34980A. These must be connected using one or two channels depending on measurement type used. These channels must be assigned as external reference channels. A calibration table for a precision PRT sensor can be used. This requires measuring the resistance of the PRT, converting to temperature using the calibration table in your PC, measuring voltage on the thermocouple channels, and finally converting the voltage to temperature in the PC.

Some of the third party provided external reference connector panels are:

• VXI Technologies, Inc., model VT1586A
• Kaye Instruments UTR panel
• Scanivalve Passive Uniform Temperature Reference
   

For the VXI Technologies VT1586A, the isothermal gradient error is +/- 0.2 C, but wrapping the VT1586A in insulation to protect it from air currents can improve this. That's not necessary with the Kaye Instruments UTR because it includes an insulated enclosure. With the optional calibrated PRT sensor the Kaye UTR is as good as you can get at +/- 0.1 C combined isothermal gradient and reference temperature measurement error.

Learn more about Keysight Multiplexer Modules for 34980A