Thermocouples versus RTD

  • Thread starter Fred Lherminier
  • Start date

Thread Starter

Fred Lherminier

From a general perspective, what are the advantages / drawbacks of RTDs versus Thermocouples for controls application (I keep it wide open purposedly)?

Accuracy, response time, industry specific tradition, cost of implementation (with or without transmitter, and cost of the PLC modules), temperature range, linearity, life time, noise, maintenance, availability....

(I guess I could name a few market specifics, like food industry, power generation, aerospace, HVAC...). (I guess I could also mention thermistors in the mix of the discussion).

Thanks all.
> From a general perspective, what are the advantages / drawbacks of RTDs versus Thermocouples for controls application (I keep it wide open purposedly)? <

Hmmm... Basically I'd recommend to use RTDs (e.g. Pt100) for most applications below around 300 °C (about 600 °F). Also they don't require special wiring hardware. For high temperatures thermocouples are better, RTDs should not be used too close to their range limits.
According to precision requirements 3- or 4-wire RTDs should be used. Also take in account self-heating effect of RTDs.
Thermistors (PTCs) are widely used for winding temperature limit control of motors. Larger motors rather use sets of Pt100 (at least 1 per stator winding). Large power plant turbogenerators can have several hundred thermal measurement points.
Concerning PLC inputs one must consider the input isolation. Usually, but not always, inputs are isolated toward data bus and power supply; check isolation among individual analog input channels.
Good but expensive practice is to use single a 3-channel (=input//output//auxiliary powersupply galvanic isolation) isolation converter or each analog input, so you can use basic 4-20 mA AI PLC modules. In the real industry (excluding HVAC which is a special domain I don't know much about), Pt100 and TC are the most used temperature sensor technologies.

We use both, but typically use thermocouples for engine exhaust temperatures, and RTDs for generator winding temperatures and bearing
temperatures. I like the idea of a 3 wire RTD over a thermocouple because the third wire is used to calibrate the loop resistance, which improves accuracy.

Paul Baker
I think thermocouples tend to be better for higher temperature applications, and are good for a wider range. RTD's tend to be more accurate, and
thermocouples tend to be cheaper to use. They both need to be linearized. Thermocouples are trickier because they require cold junction compensation and they're susceptible to errors in measurement technique. Thermocouples are easier to multiplex and don't require excitation.

Rick Daniel
Intelligent Instrumentation

B M Renukaprasad


I am currently involved in a similar decision making process to choose between T/c or a RTD.

The application is for a precision 300+ channel data logging approach for a temperature range between -40 to +250 Deg C. Body to be measured is an Aluminum rectangular tube surface. Accuracy required is better than 0.4 deg C.

I welcome suggestions.

Renukaprasad BM
Avasarala Automation Limited
According to my experience in handlingT/C & RTD ,we should use 3 Wire RTD for low temperature(-190 to 250 degree C) which is more accurate and linear but you have to pay more price.
For more sesitivity & fast response we must use E type grounded thermo-couples upto (0 to 300 degree C ) .For higher temperature we may use K type upto 1000 Degree C.How ever linearisation should be done in the DCS system or the transducer(mV to current).For further higher temperature we can go for S type.