# Help on IR Proximity Sensor Inaccuracy

I want to measure the depth of a burning soil using IR proximity sensor. The sensor will be placed 20cm above the surface of the soil and measures it perpendicularly.

Problem #1: I'm quite concerned about the environment that the sensor is in, it would be hot (up to 60 degree Celsius) and dusty (as soil burning produces particulate (PM2.5). Will this two condition affect my measurement?

Problem #2 I was thinking that the temperature will, as any measurement device will have thermal drift, but the amount of thermal drift depends on the device itself, how sensitive it is to ambient temperature. Some source says that it changes the wavelength and thus makes the measurement inaccurate.

On the other hand, from some sources the smoke of the soil burning does not affect it too much because the IR wavelength is long enough to go past through the smoke..

Is this correct? If not please correct me and provide some reference

It will also be very useful for me if how much it affects the measurement is provided. Any ideas? I have no prior knowledge on this, so pardon if there's some mistakes.

• Start by looking at the data sheet for the device. Sep 4, 2021 at 7:55
• Problem 1. Proximity sensors are not temperature sensors. They detect changes in IR irradiance levels. Temperature sensors would need to measure IR wavelength and convert that to a voltage analog. Problem 2. An IR temperature sensor will measure the surface temperature only as that is where the IR irradiation is taking place. You can't measure the temperature at any depth using IR. Sep 4, 2021 at 9:38
• IR proximity sensors use near IR. Thermal IR is long IR. Sep 4, 2021 at 16:31
• Besides the previous comments, smoke and PMxx will certainly foul the sensor and change the intensity of any single frequency signal over time. You'd need some clever way of extracting your data without the benefit of predictable intensity (e.g. spectroscopic methods, flashing illumination), or else frequent calibration Sep 4, 2021 at 19:25
• @bintangf_m - probably much slower. But you should be able to simply put your sensor in the smoky environment you plan for, with a calibrated IR source (which is kept clean), and see the drift in the baseline Sep 5, 2021 at 22:41