Heatermeter Problems - Thermocouple Seems To Be the Culprit

Hi All,

A couple of weeks ago I was doing a normal cook. Thermocouple and one probe in the first jack. About two hours in the thermocouple went berzerk.

The temperatures were jumping around like crazy, +/- 200 degrees.

To compensate, I put the meat probe that was in jack 1 into jack 2, grabbed another probe and stuck it into jack 1 (new pit probe), and disabled the thermocouple. This turned the first jack into the pit probe and all should be okay, right?

Well, the meat probe started going berzerk in this configuration.

In my desperation, I unplugged the thermocouple and things started behaving normally.

I didn't get around to testing for the problem until today.

Using the same thermocouple and two probes, I added another probe and went through all the permutations of them being plugged in that I could think of.

All three food probes are very new Thermoworks Pros. The thermocouple is relatively new from Auber, but has never given me trouble before that day. I also tried another brand new thermocouple with similar results.

The results are here:

https://www.dropbox.com/s/z4ohvg1nttaobln/Screenshot%202016-06-19%2013.20.50.jpg?dl=0



Any and all suggestions on what may be causing this would be appreciated.

You're quite organized with that chart of your testing, good work, but I'm not sure it tells us much of anything helpful. The only thing odd/interesting is it seems you only report the issues on food probes when the TC is plugged in. This could be coincidence and mean nothing, or it could mean the TC is causing some "cross talk" through flux to adjacent (isolated) probe circuits, or I guess it could mean the TC is effecting the 3.3v supply (which connects to the food probe circuits through the pullup resistor(s)), but I would think that would be unlikely and think it would probably happen whether the TC was plugged in or not if that were the case.

First of all, yes, if you disable Probe 0 Food Probe1 becomes the Pit Probe, if you disable Probe 0 & Food1 then Food2 becomes the pit probe, it flows down stream like that.

The standard probe circuits are pretty small and simple, there is only a pullup resistor and the resistor and cap of the RC filter between the probe jack and the ATMega, each probe input circuit is independent (none of them touch the other). The TC circuit is much more complicated with the TC amp and all, but it is also completely independent, meaning it doesn't touch any of the other probe input circuits. Each probe input circuit connects directly to it's own pin on the ATMega, counting backward from Pin28 on the corner, Pit, Food1,2,3 on pin 28-25 respectively. The probe input circuits are all contained on the edge of the board between the probe jacks and the ATMega, they don't cross over to the other side of the board at all.

The reason I said all that is to point out that one shouldn't effect the other, their paths never cross and each has it's own input leg on the ATMega. Since you have a similar problem randomly effecting all of the probes you should be looking at the things the probes DO have in common. What they all share is a connection to the same ground and also the 3.3v supply, food probes connect to 3.3v through the pullup resistor and the TC uses 3.3v to power the TC amp. So it is possibly a problem with your ground circuit on that side of the board or the 3.3v that goes over there.

The two common causes for this sort of mystery would be a bad solder joint somewhere (on the ground or 3.3v line) or a circuit board that has a lot of flux left on it from soldering, believe it or not extra flux on the board can make the probes flake out pretty good.

What I would suggest is a solder reflow and cleaning of the board. Personally I would first remove the ATMega from its socket, then reflow solder on its socket and also reflow solder on all the components between the socket and the probe jacks (minus the SMD parts). I would also reflow solder on the components right on the other side of the ATMega because the 3.3v passes through there and jumps from the bottom to the top of the board and then back to the bottom, so you may have a bad solder connection on the top side of the board that only manifests on occasion. I would also reflow solder on the 3.3v regulator for good measure. After the reflow you should take some Isopropyl Alcohol and wash the board, brush it while wet with a soft brush like a toothbrush and then rinse it off and let it dry. In the majority of cases with flaky probes this will remedy the problem. Most often it is flux shorting things out over there....

You could also be more technical by testing/monitoring the 3.3v supply on these parts while the problem is occurring and measuring continuity on the ground, but that can prove difficult and really isn't necessary. The reflow and wash process described above should fix those kind of problems without needing to actually track down the culprit.

Good luck with your repair....

I've seen similar issues and one bad probe was affecting the others. I'm not sure why but once I finally figured out it was a bad probe everything worked good with all other probes since. Here's my threads on the similar subject:

http://tvwbb.com/showthread.php?64608-Help-HM-broken&p=716394#post716394
http://tvwbb.com/showthread.php?64608-Help-HM-broken&p=716341#post716341

Bottom line is that if things were working good and "suddenly" start to misbehave I would recommend doing an independent probe test one at a time in cool to boiling water. This is how I clearly found my bad probe and once I took it out of the equation all the other probes stopped acting bad along with it. Sounds like you've narrowed it down the thermocouple. Buy a new TC and see if things become happy again. My issue was just with multiple probes in use at the same time but it sounds very much the same.