Thermocouple Pit Probe Issues on new HM build


 

J Halvo

TVWBB Member
First build, HM 4.2.4 w/ thermocouple pit. After installing in a case, I am getting a reading of 561f [plus/minus 3f] on the pit probe. Everything else seems to function fine including the thermistor probes. I just got the Adapt-a-Damper working tonight as well so almost ready to put it in service.

Things I've tried...
  • I have confirmed probe 0 is set as thermocouple in config.
  • I have tried with multiple K-type probes (none fancy but same results).
  • If I take it out of the case completely, things seems closer to reality but still flakey (jumps a bit up and down sometimes).

No obvious soldering issues but then again I am far from an expert.

Any tips for narrowing down the problem(s)? Wondering if a bad solder joint coupled with tight fitting case is partly to blame.

Thanks!

chCCqdZ.jpg
 
Last edited:
You can start by taking the thermocouple out of the equation by simply shorting the two pins of the thermocouple connector together using a short piece of wire or a bent paperclip. Then the thermocouple amp will output just the ambient temperature. If that reads a wonky value (like 561F) then I'd check the soldering on the thermocouple amp and components as well as the 100k/0.1uF filter closest to the pit probe (atmega pin 28). You can also install the noise firmware from AVR Firmware -> Online Repository -> hm-noise.hex. That will add a second graph to the home page that shows the raw ADC values and you can see how noisy the signal is. It should bounce around at most by 1-2 units.

If everything works fine with no thermocouple, then check to make sure your thermocouple's metal shield isn't connected to one of the two pins and picking up noise that way. You can also try a different power supply or outlet since some outlets (GFCI ones especially) have a lot of ground noise on them.
 
You can start by taking the thermocouple out of the equation by simply shorting the two pins of the thermocouple connector together using a short piece of wire or a bent paperclip. Then the thermocouple amp will output just the ambient temperature. If that reads a wonky value (like 561F) then I'd check the soldering on the thermocouple amp and components as well as the 100k/0.1uF filter closest to the pit probe (atmega pin 28). You can also install the noise firmware from AVR Firmware -> Online Repository -> hm-noise.hex. That will add a second graph to the home page that shows the raw ADC values and you can see how noisy the signal is. It should bounce around at most by 1-2 units.

If everything works fine with no thermocouple, then check to make sure your thermocouple's metal shield isn't connected to one of the two pins and picking up noise that way. You can also try a different power supply or outlet since some outlets (GFCI ones especially) have a lot of ground noise on them.

Shorted the pins, still ~561F.

Can I try the thermocouple amp testing even after things have been mounted? Yes I see the warning about the 3V limit, but didn't know if I can still use the same test.

Anything obvious on these pics?

2TQVKDO.jpg

qowC7bI.jpg
 
Last edited:
Tried the thermocouple amp test to the best of my ability. Using a AA battery (not sure this is enough to make the test work), which measured 1.48V, hooked up as instructed. Best measurement was 1.36V with most being down about 1V.
 
Last edited:
might just be the light on your pic, but check the bottom 2 pins on IC, rightside, looks like a solder bridge.
 
might just be the light on your pic, but check the bottom 2 pins on IC, rightside, looks like a solder bridge.
I saw that too but unfortunately those are actually connected in the schematic as well so that's expected.

The soldering looks great to me, like really good. I'm not sure where to tell you to go from here though. You can maybe check the thermocouple output with everything powered up by measuring the voltage on the big circle between the thermocouple amp and the 10k resistor on the TC+ side. That should read 5mv/C. 561F would be ~1.176V, so if you're seeing the same voltage there, then we know there's not a problem with the measurement. It is weird that the amp is putting out some value in the middle of the range rather than just nothing (0V) or full scale output (~3.2V). When you unplug the thermocouple does it switch to "No Probe" or does it stay 561F? I don't know what that would mean but I'm hoping we'll stumble across the answer here.
 
This was bought from the HM store with the TC surface mount stuff pre-soldered. So that wasn't my handy work :)

When you unplug the thermocouple does it switch to "No Probe" or does it stay 561F?
. It says No Probe.

So, some discouraging results....

Presuming I am measuring the proper spots, I measured about 1.53V. It seemed really finicky though.. at one point it was 1.0V and then 1.6V and also if I moved the probe around a bit then said a probe was connected and was measuring between 561F and 900F (moving target). There was even a brief time where it did seem to show the ambient temp of 63F and the voltage on the multimeter was down in range expected. Couldn't recreate that though.

Again, amateur here so appreciate the help and will give more info as I can. A couple shots and more in the gallery here: https://imgur.com/a/CY32CoW

Xij9LGW.jpg


bY4eIP3.jpg
 
This was bought from the HM store with the TC surface mount stuff pre-soldered. So that wasn't my handy work :)
I thought all the soldering looked perfect, as much as I'd like to take credit for the nice work you've done on the through hole components.

All the presoldered thermocouple boards go through a testing process with nothing plugged in and then at room temperature, so it seems strange that it is behaving so strangely. Would you be interested in sending your HeaterMeter board back and I can hook it up my test equipment and see if I can resolve the issue (I'd pay shipping both ways)? That, or I can send you out a replacement order. Email me at heatermeter@capnbry.net and we'll work it out.
 

 

Back
Top