Hello.
Disclaimer, I wouldn't recommend doing this if your vendor is available and willing to fix your device for you. Mine has since gone with the wind so I am on my own. And of course always exercise utmost caution when working with anything electrical, especially connecting your devices to high-current batteries.
So after doing a system maintenance, the partly colorblind and overheated me managed to wire the Cerbo GX the wrong way to the bus bar of a ~1000 Ah 48V LFP battery. Red caps on both terminals certainly didn't help. Immediately after turning on the first BMS I heard an intense hissing from inside the Cerbo then it released a huge cloud of magic smoke. I tore off the wire as soon as I was able to react but the device was done for.
After removing the cover and extracting the PCBs I evaluated the extent of damage. Slow-blow inline fuse was intact. The big electrolytic capacitor, added there to protect Cerbo from dying at 48V, blew up, soaked the lower PCB in electrolyte and shorted itself, causing the input filter inductor L101 to blow up tearing off a chunk of a trace with itself. The small ceramic capacitor C101 accompanying L101, looked like it seen better days but turned out alright.
The rest of the board, as well as the CPU board looked suspiciously fine so I decided to give it a go. Removed the smoldering remains of the electrolytic capacitor, washed it with a large amount of organic ethanol, scraped up the remains of the trace blown off by L101 and replaced it with an ultra-fast 4A 1206 fuse, semiconductor grade. NOTE: this is a bad thing to do, but L101/C101 looks like an input filter so it'll have to do for a bit.
Then I took a multimeter and checked the input resistance. To my delight, I saw it gradually getting lower then jumping up again, meaning the buck converter has likely survived and was trying to start.
After connecting a zink-carbon 9V battery (which is a nice self-limiting device) I heard the converter singing song of their people but device didn't show any signs of life.
Then, when I replaced the battery with a 12V/5A PSU Cerbo lit up and started.
Finally, after narrowly avoiding this disaster, I cut off the terminals and soldered new ones, with red and black caps. Then I took a 5A|200V diode, cut up the positive wire and soldered the diode inline. And did the same to the negative wire just because I have a whole box of soviet diodes lol. Now, should I ever reverse the polarity again, the Cerbo just won't power up. I don't know why Victron designers have decided against it and can't think of a good reason to not do it - except maybe saving 100-200 mW in heat dissipation. Overall the board design and choice of components look very decent. But leaving a large electrolytic capacitor at mercy of the user just doesn't seem wise. I mean, this thing can easily start a fire before the slow blow fuse decides it's time to say goodbye.
Any comments and suggestions are welcome. I'd also love to learn the value of the L101 so I could eventually order it and put it back instead of a fuse.
