0
\$\begingroup\$

I am trying to use an LM358 to amplify a thermocouple and read in the resulting voltage into an Arduino.

I bought an Amico thermocouple from Amazon: http://www.amazon.com/Probe-Thermocouple-Sensor-Temperature-Controller/dp/B00899A4LY/ref=sr_1_7?ie=UTF8&qid=1448423872&sr=8-7&keywords=thermocouple

It is type K, and if connected to my multimeter with the temperature setting, it shows 22C for room temperature. We tested it in a toaster oven, and it works at a range of temperatures.

Testing with Vcc = 6.3V (4 AA cells in series), I created the following circuit. It didn't work. The input voltage is correct.Admittedly, my knowledge of op amps is spotty.

enter image description here

Next, I looked for a schematic, even though I don't understand it. It doesn't have two resistors in the voltage divider that is familiar to me. I tried to duplicate it but didn't have a 100k potentiometer or a 270k resistor, so I tried 120k ohm resistor that was handy instead. I didn't bother with the source follower on the output since it's going to my DMM which is high impedance.

enter image description here The voltage differential is 40mV for example, dunked into hot water, and the output I am getting is 1.0mV (although it's noisy as well). Note that the author of this circuit says to put the positive side of the thermocouple to the negative terminal of the opamp (TP2). I tried both ways just in case.

Here is a picture of my circuit. I apologize that the terminals block some of the view but at least you can see some of it.

enter image description here

Any clues on what is wrong?

\$\endgroup\$
5
  • \$\begingroup\$ If you were trying to do an inverting amp if your first (Fritzing) schematic, you connected the ground wrong. I can't really say what you've connected in the jumble of wires in your photo. \$\endgroup\$
    – Fizz
    Nov 25 '15 at 6:52
  • \$\begingroup\$ I was trying to do a non-inverting amp (as shown in the schematic, the voltage divider is connected to the negative input of the op amp. \$\endgroup\$
    – Dov
    Nov 25 '15 at 6:55
  • \$\begingroup\$ No, what "no"?? \$\endgroup\$
    – Fizz
    Nov 25 '15 at 6:55
  • \$\begingroup\$ Just to add something a little more to his no, please review that the - terminal makes it inverting. \$\endgroup\$
    – mcmiln
    Nov 25 '15 at 7:08
  • \$\begingroup\$ Also that 2nd circuit is a rather unwise way to use two opamps in this application. If you're going to "splurge" on two [cheap, one-package] opamps, making a two-opamp inamp looks like a better idea. \$\endgroup\$
    – Fizz
    Nov 26 '15 at 8:38
2
\$\begingroup\$

Be aware that using a thermocouple like this will basically only give you a signal that depends on the temperature difference between the thermocouple tip and the connection to your opamp's input.

The signal is very small (41 uV/deg C), so generally opamp offset voltages will provide more signal than your thermocouple will. For this reason, to get good results you need to use an offset compensated (auto-zeroed) opamp; even then it's hard to get accurate results.

\$\endgroup\$
7
  • \$\begingroup\$ any suggestion for a part? I take it LM358/324 isn't good. \$\endgroup\$
    – Dov
    Nov 25 '15 at 7:46
  • \$\begingroup\$ I second on that. LM358 is bad op-amp for thermocouple. I would suggest you go with Instrumentation amps or dedicated IC's for thermocouple \$\endgroup\$
    – ammar.cma
    Nov 25 '15 at 8:32
  • \$\begingroup\$ @Dov: while LM358 won't give amazing performance, it can work in this application youtube.com/watch?v=gPyD8Ygq-qA If I were you I'd rather focus on what went wrong before moving on to [busting] more expensive opamps. People managed it even with 741 basically: youtu.be/MtccB9K09ck?t=244 \$\endgroup\$
    – Fizz
    Nov 25 '15 at 8:50
  • \$\begingroup\$ Sorry (and apologies to jp314 for spamming his post), but I realized pasted the wrong link for the last vid. One of the fabled Polish electricians did the latter youtube.com/watch?v=iTvZYBbPgVY \$\endgroup\$
    – Fizz
    Nov 25 '15 at 9:26
  • \$\begingroup\$ @respawnedfluff: your videos don't show anything. Not a single schematic, not a closeup of a board, not even dialog that describes what is happening. The first video shows a person plugging a thermocouple into an already-built board. The last video shows lights blinking on and off. How about an actual design using the LM358 that works? \$\endgroup\$
    – Dov
    Nov 26 '15 at 4:54
1
\$\begingroup\$

Also zooming in on your Fritzing schematic

enter image description here

you've swapped the polarity of the + and - supply pins of your opamp (pins 4 and 8), so it's probably damaged / busted by now if you've actually connected it that way in your first circuit.

enter image description here

\$\endgroup\$
1
  • \$\begingroup\$ No, apologies for the schematic, the circuit is correctly powered. \$\endgroup\$
    – Dov
    Nov 26 '15 at 4:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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