LM358 Op Amp, Thermopile and Arduino

Projects General Guidance
1

I tried to follow this instructables project post about using an Arduino to connect a thermopile sensor with an LM358 dual op amp. (generic).

Link: https://www.instructables.com/Thermopile-Sensor/

Only issue is im using a different thermopile.

Link: TS318-1B0814

My question is can I use the 5v/3v3 rail from the arduino to power the LM358 and thermopile?

Also, does anyone have any experience with converting the voltage from a thermopile into temperature and display it on the serial port?

I created the circuit as in the instructables and I wasn't getting anything from my opamp (LM358). Any time I probed the resistors to check their voltages. My arduino micro turned off and then back on when I took the probes off the positive and negative side of the resistors.

Any help is greatly appreciated!

2

From the datasheet it says.

Voltage Response VTP 5.0 ± 1.3 mV

so it is a small difference to the voltage mentioned in the instructable

The maximum voltage from the thermopile is about 4.4 milli-volts which is way too small.  A 1000 times boost now makes it 4.4 volts.

You should check the datasheet of the original to compare for sure.

Yeah you can use the 5v rail to power the LM358, the thermopile is not powered i'd say.

It's instrucables, although there are quite a few good things out there, there is also a lot that isn't.

3

Not being sarcastic, what is your goal of your project? If you only want to see the output crudely change with temperature, then it should work as well as the instructables. However, if you want to make an actual measuring device then you need more capable circuit

  1. The LM358 is a "chevy" of opamps. The offset voltage (difference between input + and Input - when operating in the amplifying mode is almost as much as your sensor.

image
image857×220 16.2 KB

Voffset is between 2 and 7 mv, you full signal from the thermopile is about 5 mv. So, any changes in Voffset due to drift or temperature will look like your thermopile sensor is seeing a different temperature.

4

Hi,
That circuit will not be very stable with a gain of 1000, you need to lookup instrumentation amplifiers or use:

The amplification is all done for you and it has I2C connectivity.

Or lookup "instrumentation amplifier"

Tom... :smiley: :+1: :coffee: :australia:

5

The goal of my project is just to read the temperature of the object above the thermopile and print it on the Arduinos serial monitor.

How would you suggest I make a more stable circuit for this application?

6

Thanks for your help! Can you suggest an instrumentation amplifier to use and how I would hook it up to my thermopile?

7

Hi,
This may help, uses a differentr sensor but the basics are the same.
https://create.arduino.cc/projecthub/itmightbeworse/pressure-airflow-measure-device-with-analog-sensor-037d15?ref=tag&ref_id=amplifier&offset=2

Tom.. :smiley: :+1: :coffee: :australia:

8

There are likely others but the TC913A will more than do the job.

It would be connected the same as the LM358.

You don't need an instrumentation amplifier because the thermopile is not connected to any other circuit. An instrument amplifier is needed when you have to subtract a voltage from both input pins. For instance, in a bridge.

You also don't need a "rail to rail" amplifier as long as your supply is sufficiently above the expected max output voltage.

9

Hi,

You also don't need a "rail to rail" amplifier as long as your supply is sufficiently above the expected max output voltage.

Say the hottest object my thermopile will get is around 70*C which equates to 2.5mV. Since in the Instructables he uses a gain of 1000. It will bring it up to 2.5v. Will supplying the amplifier with 5v be sufficient for a 2.5v output?

I might try the LM358 until then unless its unusable for my application, im not looking for 100% accuracy

10

Every amplifier has a max output "swing" without looking it up again I think the LM358 is good for about 4 volts with a PS of 5V.

I might try the LM358 until then unless its unusable for my application, im not looking for 100% accuracy

Great idea, testing with a low cost "garden variety" amplifier is a smart approach. Get things working and go from there.

11

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.