HM 4.0 meets the braided probe cable of doom


 
P

Phil Lynch

New member
Hi all,

I built a HeaterMeter v4.0 a couple of years ago (great project, BTW - many thanks to Bryan and all involved). Everything worked nicely until last week when I had the unit powered up but out of its case and then managed to drag one of the braided probe cables over the assembly; it probably hit the Pi, but it may have been the HM itself. Needless to say, some of the "magic smoke" escaped and everything stopped working.

The Pi wouldn't boot at all and was generating a lot of heat on D17 and RG2, so it's scrap.

I temporarily installed the MCP1700 voltage regulator (it was optional on the 4.0) and the HeaterMeter itself works fine. I even did a couple of cooks using it like this. I've now bought an identical replacement Pi but I can't get it to talk to the HeaterMeter. I tried with the original software on the sd card and then with the latest 4.x software. I've also re-flashed the firmware, the process of which worked fine but made no difference to the communication between HM and Pi.

The error message on the LinkMeter->Configuration page says "HeaterMeter serial communication can not be established."

Getting the multimeter out reveals some strange voltages (i.e. not 3.3v or 5v). I've done my best to annotate these on the diagram below. It seems that the whacky voltages on the UART pins would explain the serial communication failure but I'm at a loss to trace this problem to source. If anyone can point me in the direction of what to check next I'd be very grateful.

27061075741_87a5de2185_b.jpg


Cheers,

Phil
 
Oof. Believe me, I have done this as well with a braided cable shorting out a HeaterMeter board and it does seem that the Pi is the most fragile when it comes to that. Those are definitely some strange voltages, especially on the first three pins. The fact that pin 1 is 2.78V would indicate that the reset pin is pulling current. It is only 52 microamps but that seems high for something that should be high impedance. Of bigger concern is the TX line, which should read 3.3V with nothing attached for sure (although it should vary once a second as data is transferred). Do you have a spare atmega lying around you could swap in and try?

You also could try pulling the chip out and measuring the resistance of pin 2 and 3 to RX and TX on the Pi connector, and also from them to 3.3V and GND to see if the PCB itself has shorted internally. If those all check out as being connected to RX/TX and not at all to 3.3V and GND then for sure you just need a new atmega328 chip.
 
Thanks Bryan, much appreciated. I'll pull the chip and run those tests as soon as I can get my hands on some de-soldering braid.
 
You're quite right! I've spent that much time looking at the underside of the board that I'd completely forgotten it was in a socket. Cheers!
 
Quick update: Pins 2 and 3 are connected to RX and TX on the Pi Connector and not to 3.3V or GND, so I'll replace the atmega and report back. Thanks again.
 
Final update - replacing the atmega did fix the problem and I'm back up and running :smilekettle:

Thanks very much, Bryan, for the help and advice. Thanks also, Dave, for saving me a facepalm moment when I got my soldering iron out!
 
You must log in or register to reply here.

 

Back
Top