Dodgy temperature controller.

It's pretty common. High voltage cables do fail over time and it's often at the terminations. Not so good if someone is down in the vault working.

Big Clive

Ooops I hit send before the question. Have you ever made sparks fly out of a manhole? <a href="https://youtu.be/eKzuV74BLVo" rel="nofollow noopener" target="_blank">https://youtu.be/eKzuV74BLVo</a>

Jim

How timely! I have a question about the safety of our towns infrastructure. A few days ago our home and about 60 others lost electrical power for a few hours. The cause of the outage was said to be electrical arcing under ground about a half mile away.

Jim

Do you have another thermocouple you can try?

Gadgetman

Yikes...looks like someone's supervisor was on vacation or something. Even for low-end Chinese "standards" this seems like a raw design that hasn't had the benefit of a more experienced hand overseeing things and adding the refinements it most certainly requires to be actually usable. I wonder also if the thermocouple itself might possibly have an issue? It's entirely possible given the circumstances.

Michael Thompson

A short on the sensor terminals should give the internal reference temperature reading. The thermocouple probe gives a + or - voltage depending on how hot or cold the end of the probe is compared to its terminal connections. The thermometer then calculates the temperature based on the sum of the its internal reference temperature sensor and the voltage measured on the probe (41 µV/°C). If the terminal connections are warmer or cooler than the device's internal temperature sensor (e.g. touching the terminals with warm fingers or right after taking the thermometer inside out of a cold or hot vehicle), it will give a false reading until the device's internal temperature sensor stabilises with the temperature of the probe terminals.

Seán Byrne

The only things that I thought you could have added, were testing the thermocouple itself and testing the temperature reading of the system at some more, relatively stable, points: For example, ice water is at a stable 0°C and boiling water is at a pretty constant 100°C.

Weird indeed.. Did you try a different thermocouple? The proximity effect is strange - I assume the cable is screened. Why would noise on the ground affect the signal - surely that's referenced to the board ground? What about modding in the "missing" decoupling cap from the reference design? Component choices look ok, though a 7-segment and matrix keypad driver combined seems like overkill - all that could be handled in software with 11 or 12 pins of the micro. Not crazy about those tiny pads on the relay switching pins - altogether a mixed bag, cumulating in a big fat thumbs down methinx...

Gordo

If you shorted the thermo while grounding it with your fingers as b4 would that have an effect for the weird low temp reading?

With the exception of the power / relay terminal separation and pad size I'd say all of your points were what most 'regular' consumers would expect. It seems to be a "My First Project" experiment that got sold.

Scissors, Clive...? Really? Go to your room. ;p (Also, what a piece of crap this "controller" is. I guess *technically* it *does* control things; just not in the way you might expect... What happens if you reverse the thermocouple leads?)

Daddy Bearcat