Incorrect Temperatures, how do I find root cause?

Hello so after a long time of wanting to do this project I finally jumped in. I'm probably more excited than I should be to get this up and running, but it's been a fun little project. And I'm an engineer/geek so I see so many uses for a device like this. I already have parts for a crock pot sous vide and my eye on a couple of things to make it a stand alone system.

Over the weekend I managed to get the board together and running. But the temptertures I'm getting are pretty far off from what they should be. The thermocouple is around 3-5F off, but that's more or less calibrate-able due to linear curve. However probes 1-3 are way off, 10-25F. I have two thermoworks probes that I'm using, and from switching positions at known temperatures (room/boiling water) I'm pretty sure the issue is with the heatermeter not the probes. And yes I did switch to the thermowork probe setting in the web UI.

It could be a fair assumption that I broke a component (or more) while soldering, soldering not a skill I use much (or at all). I actually had to buy a new soldering iron because the old one I had (borrowed) was just not working. With a new one I did much better (temperature control helped tremendously) with soldering but still have some issues especially with ground plain connections. So poor connection somewhere, or I over heated something would not surprise me.

How do I go about debugging the board, I started with getting the voltages off of the test pads for the probes, but I don't know where to go from there, or really what the readings should be. (3.3V +- some error right?) Without probes attached I get,
"3.3" - 3.27 V
"0"- 3.229 V
"1" - 3.25 V
"2" - 3.258 V
"3" - 3.244 V

So in the range of 1-3%error, is this too much or is that within spec?
I do get a voltage drop after attaching a probe.

FIY I'm using rev 4.3

Thanks for your help
-Dave

Sometimes, while testing the heatermeter, while its on, I will take the soldering iron to the probe jacks and every connection in between the them and the atmega chip. I do that to be able to have some feedback to help find the problem. But, you have to be very careful in doing that, and watch what you touch on the board.

Can you post some high res pictures of the front and back of the heatermeter board? Upload them to http://imgur.com/ and then post the links to the images.

-Steve
Yes I can, I'll take them after work when I get home and post them tonight

-John
Are you basically saying that you rework individual connections while the board is on?

D.Weiser said:

"3.3" - 3.27 V
"0"- 3.229 V
"1" - 3.25 V
"2" - 3.258 V
"3" - 3.244 V

Click to expand...

Your 3.3V is within spec, which is +/-3% so that number looks good (off by 0.9%). The probe points don't seem right though, they should read about 0.003V less than the 3.3V point, but largely depending on the impedance of your multimeter. Regardless of that, they should be identical between the 4 points. I just tested 3 of the 4.3 HeaterMeters I have on my workbench and they were all 3mV-4mV lower than the 3.3 point but all within 1mV of each other.

If you unplug the power on the HeaterMeter, you can also measure the resistance between the 3V3 point and each of the probe points. It should read within 1% of 10,000 ohms.

I finally got a chance to take some pictures, I uploaded them to a flickr account https://flic.kr/s/aHskFF4kVN

I measured the resistance at the probe points I'm getting 9.5-9.65K, which is out of spec. I notice that the resistance climbs a bit while holding the multimeter probes in position is this normal?

I'm assuming next step, remove 10k resistors measure them off the board?

I'd scrub all that flux away with some 99% isopropyl alcohol. It's not supposed to be conductive, but it's a common source of headaches.

You don't need to remove resistors to measure them accurately. You just need to ensure the resistor is the only thing between the meter probes.