Trying to use a grounded thermocouple with an instrumentation amplifier

[EricNava]

Hi all

I am trying to make my own JBC cartridge soldering station.
I would like to use an instrumentation amplifier to amplify the reading from the thermocouple in the JBC tip.

If possible I would like to use a single supply for the instrumentation amplifier, although it seems this can cause problems.
When looking at other people's designs, most use an op amp with dual supplies. To me it seems this shouldn't be necessary because the tip will never go below 0 degrees, and thus the thermocouple voltage will always be positive.
There is probably something I am missing here, but I can't figure it out.

Thermocouple outputs vith temperature:
https://en.wikipedia.org/wiki/Thermocouple#/media/File:Intermediate_temperature_thermocouples_reference_functions.svg

Another issue is that one lead of the thermocouple is grounded.
I believe that in most stations the outer shell of the tip is connected directly to earth, which makes sense for ESD protection.
However in the datasheet of the amplifier I want to use, (pg 18) it states:

"With single-supply operation, VIN+ and VIN– must both be 0.1 V more than ground for linear operation. For
instance, the inverting input cannot be connected to ground to measure a voltage connected to the noninverting
input."
Texas instruments INA333 instrumentation amplifier
https://www.ti.com/store/ti/en/p/product/?p=INA333AIDRGT

So it seems that using dual supplies would alleviate this issue, however I have seen some other amplifier datasheets show a circuit with one of the thermocouple leads connected through a 1M ohm resistor to ground.
Would that keep the thermocouple voltage more than 0.1V above ground?

Also if I use dual supplies, it seems it is ok for one of the thermocouple leads connected directly to ground, is that correct?

Sorry if this post doesn't make much sense. I have never used an instrumentation amplifier or op-amp before so I am trying to learn.
Thanks

[Miti]

If possible I would like to use a single supply for the instrumentation amplifier, although it seems this can cause problems.
When looking at other people's designs, most use an op amp with dual supplies. To me it seems this shouldn't be necessary because the tip will never go below 0 degrees, and thus the thermocouple voltage will always be positive.
There is probably something I am missing here, but I can't figure it out.

You don’t need to go below 0*C to have a negative output from the thermocouple. If the hot junction is colder than the cold junction, for whatever reason, the output is negative. When both junctions are at the same temperature, the output is zero mV.
I think what you mean is that you’re not interested in measuring temperatures close to the cold junction, so the output when the iron is hot should always be positive.

Take a look here, see how Sparky did it:
http://dangerousprototypes.com/forum/index.php?topic=7218.0

[EricNava]

You don’t need to go below 0*C to have a negative output from the thermocouple. If the hot junction is colder than the cold junction, for whatever reason, the output is negative. When both junctions are at the same temperature, the output is zero mV.
I think what you mean is that you’re not interested in measuring temperatures close to the cold junction, so the output when the iron is hot should always be positive.

Take a look here, see how Sparky did it:
http://dangerousprototypes.com/forum/index.php?topic=7218.0

Thank you for the reply.
I hadn't thought about how the relative temperatures between the thermocouple junctions could result in a negative voltage.

I have been quite intimidated by the Unisolder schematic, although I have tried my best to understand it now and it has helped. Looks like he is using 3 op-amps in an instrumentation amplifier configuration with 3.3V and -0.6V dual supply.
It is interesting how he used the LM2675M-3.3 to also generate the -0.6V, don't quite understand how he did that, however I will definitely use dual supplies for my circuit now do avoid the grounding issue and for when the cold junction is colder than the hot junction (even though that is a unlikely situation).

[magic]

Instrumentation amps amplify a differential voltage differentially in both directions before converting to single-ended. This can't work with one input at ground.

Any reasons why you can't use a normal single supply opamp like TS27L2, LM358 or LT1013?

[SiliconWizard]

I've used the AD8223 for the exact same kind of application. It doesn't have this limitation in single supply mode.
Note that you'll still need to offset the output using the Vref pin, otherwise it will be out of spec. In my case, I used a 1.2V voltage ref which I had available for other parts of my design.

[EricNava]

I've used the AD8223 for the exact same kind of application. It doesn't have this limitation in single supply mode.
Note that you'll still need to offset the output using the Vref pin, otherwise it will be out of spec. In my case, I used a 1.2V voltage ref which I had available for other parts of my design.

Thanks for bringing the AD8223 to my attention. It does indeed show a single supply circuit amplifying a grounded thermocouple. Seems pretty much perfect for what I need.
It seems like has much higher input bias current and a lower Common mode rejection ratio than the INA333, but I don't think that will matter for my purpose.

[Bgdn]

Hello,
i saw this discussion just moments ago.  Did you came up with a final design?

This caught my atention, because i also work on a design with inamp for reading a thermocouple. I am using AD8221 with +3.3 & -2V supply

[MathWizard]

IDK what opamp is best for your application. But if you don't have a dual supply, and the opamp doesn't use much power, what about using some little charge pump circuit, possibly with an IC, that creates -1, -2 or 5V or something ? They can work ok for some digital stuff, IDK how noisy they are for use with op-amps tho.