Thermocouple calibration - way off


 

Steve_M

TVWBB Guru
Decided to calibrate my thermocouple using the method described on the wiki

I've currently got it in a pot of boiling water and it's not reading any higher than 176F

My trust thermapen bounces between 211-212 in the boiling water.

The room temperature readings seem relatively normal.
 
Yep, it's at 5.

Based on the calculation in the calibration doc, 5 * 176 / 211 = 4.17

If I use 4.17 I end up with 204 as the new readout, not 211
 
Just curious, could you have the resistor that is not needed when using the thermocouple. I had that and it would not read below 50. Just thought I ask.
 
Do you have a good multimeter than can read mV fairly accurately? In boiling water the thermocouple output should be exactly 0.500V. At 176F it would be 0.400V. If it reads 0.5V then the problem is somewhere in the 3.3V reference to the ATmega, if it reads 0.4V then the problem is somewhere on the thermocouple circuit side.
 
So I finally got around to looking into this some more and I think my thermocouples are faulty.

In boiling water, measuring across the pins of the male thermocouple connector, I'm seeing a max of 3.05 millivolts using my Fluke 8060A.
 
The PCB for my v4.2.4 do-over arrived tonight and it's up and running with no display as of yet.

I performed the 2 voltage tests with a 9v battery on the TC circuit before continuing on with the build, and both checked out fine.

I've got my k type thermocouple plugged in and the temps are still out of whack, showing about 116F, while sitting at room temperature.

You can see it live here, I'll leave it on overnight to see if it settles down by morning.
 
Last edited:
I am at a loss. I checked your configuration and it looks good, the ADC isn't noisy. Are you sure it is a type-k thermocouple? I just don't understand how it can not just work.

I don't have an open HeaterMeter to test with, so I am not sure what the differential measurement would be at the thermocouple connector. It should b pretty small though I thought.
 
Yep, it's a K type. I'm going to source another one and see what happens. I'm pretty confident that everything on the HM side of things is fine.
 
Hey I got one of those too, to compare my thermocouple readings from the HeaterMeter! They work alright and read withing a degree (C) of HeaterMeter.
 
Is there a way to test the TC amp on a fully assembled HeaterMeter?

Edit 1: I see that the docs for the AD8495 say that you can ground both input connections and it will report its own ambient temperature.

Edit 2: With the inputs grounded, I'm seeing the AD8495 report an ambient temp of 92F, while the Vishay 10K is reporting 68F. I've removed it from the case, so the vishay reading is pretty close to what the temps are in my basement.

Edit 3: Some measurements I've taken with my meter. Based on the TC testing doc on the wiki:

Nothing plugged into the TC jack, I'm seeing 3.25v between GND and "4"

Between "5" and "2" shows 3.3v

Between "5" and "4" shows 0.05v

TC inputs shorted to ground, I'm seeing a reported temp of 102F and 0.2v between GND and "4"

Edit 4:

I left it overnight with the inputs grounded. It seems to have lowered in temp after time, but the temps are still quite high. For the most part it's following the same trend as the Vishay 10k.

 
Last edited:
I'll give that a try, though I did tests across the 1K and 10K SMD resistors on the board and was getting proper values.
 
So I'm pretty much at a loss to what's going on with my TC circuit. I used solder wick and scrubbed it all with 99% isopropyl alcohol, I even wound up re-soldering the TC amp and I'm still getting readings that are sky high. It's currently at 168F and very slowly coming down, but I doubt it's going to reach the 70F that the Vishay 10k is reporting.

My $25 Amazon unit arrived today and so far, just at room temperature, my eBay K Type probes are spot on within 0.1 degrees of each other.

I don't consider myself a novice, but this sure is making me feel like one!
 
Is it possible you have either SMD resistors or capacitors that are the wrong value (or in the wrong place)? Sometime parts suppliers can make a mistake, if you have any left over SMD components perhaps try measuring them to see if they are the right value?
 
Is it possible you have either SMD resistors or capacitors that are the wrong value (or in the wrong place)? Sometime parts suppliers can make a mistake, if you have any left over SMD components perhaps try measuring them to see if they are the right value?

The resistors I've confirmed are the correct values. No way to test the caps, but I do have spares. The temp reading is currently at 141F.
 

 

Back
Top