ATMega328P failure? Diagnostic advice please?

[Web Dove]

Hiho,

Yesterday when I started my 5 year old HMv4.2 it reported a pit temp of 670 when the fire was cold (32ish in New Hampshire). Today I pulled the board to experiment.

3.3v power is good. Food probe 3 is at 2.5v when connected to a minijack probe at 69 degrees F. The HM reports that probe at 105.7 degrees F. That reading rises to 120F when I hold the probe tip. A Thermoworks J-Type probe which is known to work on a thermoworks sensor is causing .1V output from the AD849x at 20 degrees C. The AD849x data sheet says that is correct. The HM reading is 658 degrees F. If I hold the probe tip the HM reading falls to 620F (wierd). The computer, webpage and USB wifi all appear to be working normally.

It sounds like the A/D conversion is broken. Do you have any suggestions on how I could confirm that? Is it possible that the ATMega328P failed? I tried removing it from the socket and reinserting it, but the problem persist. Thanks in advance for any advice you can provide to me.

Web

 

[Bryan Mayland]

I've never seen an ADC go bad before. Did you check the configuration in the webui to see if the probe configs just got scrambled somehow? You could also try doing a "Reset Config" from the LCD menu on the device itself to reset everything to default. This also wipes the Pi's config but if you disconnect the Pi from the HeaterMeter when you do it then only the HeaterMeter will be reset.

If you've done that, and you've checked AVCC (labeled AVC) for 3.3V and verified that the voltages you're reading are measured at the ATmega pins, then I guess it could be possible that it has become damaged in some way. I'm not clear on what the failure mode of these microcontrollers is, but it seems more likely that the configuration is just messed up somehow.

 

[Web Dove]

You were exactly correct. The linkmeter had forgotten most of its settings. Fortunately I had some old emails and onenotes with the settings and was able to recover. Thanks very much Bryan!