I just spent the last few hours with my multimeter fixing the shorted power cable and resplicing the bad spots.
This must have been heat related because I was using it as a dashcam… I always used a micro usb 5V 1000mAh cord to utilize the protection circuits (instead of powering the bus directly) but it still managed to pop some transistors it looks like?
Now I know it’s not best practice but… for science could I just reflow this with solder and see if it at least powers on?
I got the GPIO pinout and verified clean, full voltage on the 5V and 3.3V GPIO board power pins after reflowing that bad resistor. Used electrical tape as a solder wall
The sd card slot had a lot of corrosion that I cleared out… sd card and slot are nice and shiny now.
I can hear the board power on too. Seems like the SD card is at fault now since the activity LED isn’t lighting up.
I don’t recommend doing this. You can bypass, but you may end up damaging the component it’s connected to.
You could try using an ultrasonic cleaner or just take some IPA with a toothbrush and gently scrub away the corrosion. Hopefully it’s not damaged anything major. If your lucky it’ll cleanup and most everything will work though it’s looking like that resistor is FUBAR, you can measure the resistance with a mutlimeter then attempt to order a replacement part.
The board should really be cleaned up first. The corrosion products are probably shorting things out. Might even consider using dilute acetic solution followed by a water and then IPA rinse.
Also the next time you deploy one of these to the same environment you should probably use some conformal coating on it so the corrosion doesn’t occur again.
I’ve cleaned the whole board with penetrating oil followed with air drying. Nice and shiny now. I’ll definitely coat it with at least silicone grease next time… foggy windshield protection?
sd card slot had a crap ton of corrosion + melted solder
I’ve tried so many times to reformat the sd card but I’m starting to realize the sd card slot is broken and the board is getting power. Sd card appears fine despite having a small burn mark on one of its pins from the slot on the pi zero w
I’m going to try and boot the pi over USB without the sd card in by utilizing usbboot gui. From my research it should work
I forgot to update my findings here, my apologies.
Pasted from lounge:
"And as I had it all setup I decided to give my pi zero w one last try and got (painfully) usbbootgui up and running in a Lubuntu VM. Had to “debuild” it from source, manually install it as an app with all the missing dependencies and got it working.
Only to find out either my cord is bad or the pi zero w is bad. Likely, the pi. If I can get a new one asap I can usbboot that just to prove it works. But I tried both hardware and software pass through of the usb cord in qemu. Nothing"
So Pi fails to usb boot without sd card despite using both hardware and software passthrough of the USB NIC on my laptop.
In the trash it goes
for anyone else trying to get usbboot GUI running, partly why it was so difficult for me is I realized only after successfully debuilding the project that it was last updated for Ubuntu 22.04. It’s possible to debuild on 24.04 but I had to mess around with repos including the raspberry pi repo, manually install by studying the broken debian installer script, and didn’t sleep for like 24 hours lol.
I would have just used Raspian X86 in a VM, but I couldn’t get it to work with QEMU.
Now that I think about it… could I have just setup raspian lite arm in QEMU to emulate another pi and get usbboot GUI from the raspberry pi repo that way? I think I remember reading you can usb boot from another pi
Ok wow. So just run raspian OS lite arm32 in QEMU, install rpiboot package. Dedicate usb port to VM. Then usbboot raspian os lite arm32 again over an OTG cable.