I'm looking for an appropriate IC to either amplify the voltage output of a thermopile or an IC to amplify and digitize its readout.
The thermopile is this one:
The IR radiometer should be able to read temperature via a thermopile and a thermistor. I know that other less expensive sensors exist that have built-in amplifiers and digitizers such as those you can get from adafruit but I'm expected to interface with this particular one.
I've done some reading already and identified a couple of candidates:
LMP91051 is one of them:
Unfortunately I have determined that this isn't for me. This amplifier with PGA is intended for NDIR gas sensors like for CO2 sensing so I think it can only amplify positive voltage, which is the case for gas sensors. I wonder if someone has experience with this IC and can comment on this. In my situation, the thermopile may have negative voltages when the target it's pointed at is cooler than its housing.
This has ADC built-in but I suspect its maximal voltage is too low. It's designed for sensing MEMS thermopile arrays with a 600uV max, although in its spec sheet 8.3 it mentioned both positive and negative range readings. My sensor has a 50uV/DegC output so the sensor and the target can only differ by 12 degC (600uV/50uV/DegC=12DegC) to max out this IC.
Any recommendations for IC (prefer with ADC built in) that can handle a large voltage range such as +- 1mV?
hello, I have been working on it for about three years.
I'm working on remote temperature measurement systems. It's good to know that there is someone interested in this issue. I share what I know and I want you to share your knowledge.
As you know, I started working for temperature sensing, with wavelengths between 0.8 and 14um. I am trying to measure the temperature of objects from afar. As an analog, I do not have difficulty in circuit design, but I will have to go a long way in software..
I have TS118-3 thermopile (8-14) I have uM different ones I have 0.8 nm ..1700nm sensors such as mrt511 AND I continue my studies I can share a few LTC6915 ADS1115 I measured in differential mode and 24 bit adcs
I will share more anolog ina188
I don't think we are working on similar things @nzmxp . I am reading a thermal pile+thermistor assembly with an analog front end that is designed for thermal piles plus and ADC and you are using instrumentation amplifiers with ADC. I'm not an expert in reading thermal piles but I thought the uV offsets on these in-amps will kill your signal. Have you been able to get anything that makes sense in temperature read out?
I am not an expert. We produced 20 samples of these devices to meet some needs at the factory. The design process is entirely mine. not for a commercial activity.
but if I develop it well in the future, I will consider producing it myself. There is a demand here in this sector, there are people who want from different factories.
I have a certificate for heat and precision measuring devices.
The first-class ambient temperature measurement certification office provides technical information. FRANCE https://www.bipm.org/en/home
Yes, I read an article on this topic. Real life simulator programs are prepared, they do not work like circuits. You are right
I set the output to zero volts. Or I usually reset with the opamp offset settings.. Of course, I am aware that the offset voltage specified in the opamp's data is determined.. Such a device will definitely make erroneous measurements and will make more erroneous measurements during calibration.
There are 1 uv offset opamps but they are not easy to find with adc. I have reviewed these series, but I think you will see the features you searched for.
ltc2400
ltc2421
AD8628
Looks pretty neat. As I said I am using LMP91051 and I'm reading its output with an ADS1115. This actually works. LMP91051 has a bias voltage generator to add to the thermal pile's reading before amplifying it so it will always be positive voltage despite thermal pile could have a small uV negative voltage if the target is cold.
The IC also has a 1:2 multiplexer for its input so I hooked one channel to the thermal pile and the other channel to the bias voltage. When doing so, in CH1 I get an amplified voltage from the thermal pile (plus a slight uV offset from the amp) and in CH 2 I get the uV offset from the amp (inverting and non-inverting inputs are shorted and still a non-zero voltage is read). I can subtract them to get the correct amplified voltage from my thermal pile. My readouts look pretty reasonable. My sensor isn't just a thermal pile. It's got a shield etc and comes from apogee. I have some bare thermal pile + thermistor chips like yours that I could try when I have time but I hope I can get the same accuracy as the apogee ones, which I think is just a bare thermal pile + thermistor with the fancy radiation shield and metal body for temperature stabilization and limiting the input angle. You should try the LMP91051. I'm on a consulting project so I can't share more than what I already have. But if you carefully read the spec sheet and do what it says you should do, you'll get it to work.
There is already a protector. protection from ambient temperature and protection from things that will damage the lens
I added some features.
thermal radiation I think it makes sense to read from a narrow place.
our heat source constantly emits radiation,
the sensor collects them and works in receiver mode....
you studied the sensor is very specific very narrow voltage range you are measuring cold and heat i am only reading hot radiation
LMP91051 has no thermistor reading. I think there is a separate entry on mcu.da.
In the software that I will try to read the thermopile, there will be no obstacles when I come to good places.
controls the need to increase or decrease the thermistor gain according to the calculation amount. Thermistor PTC (+) or NTC (-)
adding or subtracting ambient temperature from thermopile temperature
I understand that I read everything you wrote.
in a way yes, the ambient temperature is taken as a reference. We are currently trying to get accurate measurements,
Radiation heats objects in the environment, such as objects and air.
solid liquid gas they have reference temperature
There is no such thing as absolute temperature.
this thermal source is always there, if necessary calibrated at zero degrees or in an air conditioned environment then we are at the mercy of the radiation while in the field.
If you have a thermocouple it’s output depends on the difference in temperature between the “hot” and “ cold” junctions , both being the junctions of dissimilar metals . So if you know the temperature of one junction you can calculate the other . A thermo pile is stack of thermocouple with “hot” and “cold” junctions on either side . I would have thought if you turn the current off you could measure the generated voltage voltage and calculate the temperature of one side relative to the other . Worth an experiment.
If a thermopile is heated or cooled by the external environment , you can still measure it’s temperature difference .
If you strap a temperature sensor to one side ( eg DS18B20 on the cold side) you can then get the temperature of the other .
