I am using an MRT511 thermopile temperature sensor and COS77DT for amplifying the temperature sensor output.

I used Recommended circuit in the MRT511 datasheet but output voltage of the op-amp not varied with thermopile sensor. Output voltage is always constant.

Can anyone share mistakes in that circuit?

  • \$\begingroup\$ There's something wrong with that link. I just get the start of an XML document. Edit your question to include a schematic of the device and show where you are measuring. \$\endgroup\$
    – Transistor
    Jul 23, 2020 at 10:28
  • \$\begingroup\$ Pins 2 and 4 measure the temperature of the part itself, not what the thermopile is "looking at." Have you tried varying the temperature of the device? \$\endgroup\$
    – JRE
    Jul 23, 2020 at 10:33
  • \$\begingroup\$ Pins 1 and 3 provide a voltage proportional to the temperature of what the thermopile "sees." \$\endgroup\$
    – JRE
    Jul 23, 2020 at 10:35
  • \$\begingroup\$ I tried to amplify the voltage across the 1st and 3rd pin of the thermopile sensor using OPAMP LM358.. I got amplified output and the output voltage varied with temperature..But had an issue when am using different IC's same LM358 package. Output voltage varied for same temperature, when used different IC's. \$\endgroup\$ Jul 23, 2020 at 10:37

1 Answer 1


Here is the datasheet.

The NTC sensor resistance between pins 2 and 4 should vary with the actual temperature of the device, about -5% per degree C (as shown on the chart), and 100K at 25°C. This is to be used to compensate the millivolts out of pins 1 and 3.

The latter (which reacts to the actual temperature being measured vs the sensor) is about +40uV or 50uV/°C so you would want a better op-amp than the LM358 for sure, in terms of Vos and TCVos. Also watch the bias current, since the source resistance is quite high (as much as 150K at 25°C). The offset of the LM358 can be as much as +/-9mV which corresponds to something like +/-150°C measured temperature.

In general you will calculate the device temperature from the NTC sensor resistance and then calculate from the curve in figure 1 the millivoltage corresponding to that temperature. You read the millivoltage off the thermopile, then add the millivoltage from the NTC calculation, then calculate the resulting temperature using the 25°C graph in figure 1.

Edit: Ignoring nonlinearity (which the slightly complex procedure above deals with properly), the thermopile measures (approximately) the temperature difference between the remote temperature being measured and the sensor ambient, and the NTC sensor measures the sensor ambient. So if the thermopile indicates a millivoltage that corresponds to 50°C (with the assumption of a 25°C sensor temperature) and the sensor temperature is actually 35°C, then the remote temperature is approximately 60°C.

  • \$\begingroup\$ Emphasize that it is measuring two temperatures... would it be fair to call these "local vs. remote"? "conducted vs. radiated"? \$\endgroup\$
    – mbedded
    Jul 23, 2020 at 15:15
  • \$\begingroup\$ @mbedded edited. \$\endgroup\$ Jul 23, 2020 at 15:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.