I wanted to confirm my thoughts to spare a pi from the bin just in case I'm over looking something.
It is a Pi 4B that is looking after a pair of timber kilns by way of 8 solid state relays (opto22 - they sink 1-2mA each when on). It runs an instance of node-red which provides the interface and logic control.
The other day a vent motor failed (start capacitor..?) which to keep running I was going to resolve by swapping the ducting to the one of the kiln not in use and reprogramming the pin, and somewhere in this exercise I have stopped the pi being able to connect over ethernet.
I've tried writing an OS to a new SD card and swapping out, with no success. Tried removing all of the static IP lines in the dchpcd.conf file with no success (it is assigned 169.254.111.225 - interestingly when I do set a static IP (a new one I've never used before) then plug it in, the router does see a device using this IP in its client list, but no data is transmitted or received, nor is it able to be pinged. The two LEDs on the pis ethernet port light up and flicker also.
Lastly I've tried the new SD card in a new pi, on the end of the same ethernet cable, and it connects no problem.
To me, confirming that the pis ethernet is mostly working but, not quite?
I'll blab on forever if I go about the whole story but during the vent fix I noticed GPIO 19-PCM_FS (configured as an output - could've been high or low at any time) was accidentally heat shrinked together with GPIO13-PWM1 next to it (configured floating - unused). I doubt they touched but, potentially they could have. This also resulted in the next pin in use along being counted wrong hah, I counted the pair in the heat shrink as one pin and so had another relay connected to GPIO 0-ID_SD which was left floating (this only partially drove the SSR with ~.8V and eventually caused the failure of it). Other than this, soldering and the usual poking and prodding that goes along with a fix like this I can't think of anything else that could've caused it. For context - the second kiln has been added in the past couple of weeks where these installation errors came about - the first one has been running flawlessly for a few years.
Eventually I enabled wifi with the wpa supplicant file and was able to shell into it and see it is all running fine, i change the vent pins, and can see it is doing the job correctly, but it has to be able to connect to ethernet. I'm wondering if there is anything else to try before I tackle swapping them out. Can shorting GPIO pins cause ethernet to work but not work, like this?
Thanks in advance!
Josh
It is a Pi 4B that is looking after a pair of timber kilns by way of 8 solid state relays (opto22 - they sink 1-2mA each when on). It runs an instance of node-red which provides the interface and logic control.
The other day a vent motor failed (start capacitor..?) which to keep running I was going to resolve by swapping the ducting to the one of the kiln not in use and reprogramming the pin, and somewhere in this exercise I have stopped the pi being able to connect over ethernet.
I've tried writing an OS to a new SD card and swapping out, with no success. Tried removing all of the static IP lines in the dchpcd.conf file with no success (it is assigned 169.254.111.225 - interestingly when I do set a static IP (a new one I've never used before) then plug it in, the router does see a device using this IP in its client list, but no data is transmitted or received, nor is it able to be pinged. The two LEDs on the pis ethernet port light up and flicker also.
Lastly I've tried the new SD card in a new pi, on the end of the same ethernet cable, and it connects no problem.
To me, confirming that the pis ethernet is mostly working but, not quite?
I'll blab on forever if I go about the whole story but during the vent fix I noticed GPIO 19-PCM_FS (configured as an output - could've been high or low at any time) was accidentally heat shrinked together with GPIO13-PWM1 next to it (configured floating - unused). I doubt they touched but, potentially they could have. This also resulted in the next pin in use along being counted wrong hah, I counted the pair in the heat shrink as one pin and so had another relay connected to GPIO 0-ID_SD which was left floating (this only partially drove the SSR with ~.8V and eventually caused the failure of it). Other than this, soldering and the usual poking and prodding that goes along with a fix like this I can't think of anything else that could've caused it. For context - the second kiln has been added in the past couple of weeks where these installation errors came about - the first one has been running flawlessly for a few years.
Eventually I enabled wifi with the wpa supplicant file and was able to shell into it and see it is all running fine, i change the vent pins, and can see it is doing the job correctly, but it has to be able to connect to ethernet. I'm wondering if there is anything else to try before I tackle swapping them out. Can shorting GPIO pins cause ethernet to work but not work, like this?
Thanks in advance!
Josh
Statistics: Posted by JoshNZ — Mon Jul 28, 2025 8:51 am