Cerbo / Raspi 4B startet sporadisch neu

Had something similar going on with my setup last summer — Cerbo GX kept dropping off sporadically, maybe once every few days. Took ages to track down.

Turned out to be two separate issues ganging up at once, which made it a nightmare to diagnose:

1. Power supply to the Cerbo — the small DC cable run had a dodgy crimp. Under certain load conditions it was dipping just enough to cause a brownout reset. Worth checking voltage right at the Cerbo terminals under load, not just at the battery.

2. VRM/MQTT polling — had a Raspberry Pi 4 also pulling data and I think the combination of that plus a firmware update around v3.00-something introduced some instability. Rolling back temporarily confirmed it.

For anyone running a Pi alongside Cerbo, are you powering it independently or tapping off the same supply? I've seen a few people get caught out by the Pi pulling just enough to cause issues during peak comms activity.

Also worth checking:

• Cerbo firmware — v3.70 should be stable but always worth a fresh flash if you've updated over-the-air multiple times
• Any DVCC settings clashing if you've got mixed inverter/charger combos (MultiPlus II setups especially)
• Whether the restarts correlate with high AC loads or PV clipping events

My shepherd's hut system is relatively simple — single MPII and Fogstar batteries — but I've got a mate with a more complex 3-phase-ish setup and he had Cerbo gremlins for months before finding a grounding issue.

Anyone else had random Cerbo reboots they've managed to trace? Curious what the culprit was.

@FrostySocket the "two separate issues ganging up" scenario is absolutely the classic diagnostic nightmare — each problem individually sits just below the threshold where you'd notice it, but together they conspire to make you question your sanity.

On my narrowboat I once spent three weeks convinced I had a dodgy Cerbo, turned out to be a marginally undersized DC cable causing a brief voltage dip whenever the fridge compressor kicked in, combined with a flaky MicroSD card that happened to be writing at exactly the wrong moment.

Worth the OP checking:

• Supply voltage at the Cerbo itself under load (not at the battery)
• SD card health — Victron's own cards are surprisingly mediocre
• VRM logs for any pattern in when the restarts happen

The logs are your friend here. Sporadic faults that leave no evidence are almost always hiding in plain sight.

@FrostySocket @ExmoorNomad oh the dreaded two gremlins masquerading as one scenario — I lost a solid fortnight on my narrowboat chasing what I was convinced was a dodgy Cerbo, solemnly swapping cables and muttering darkly, only to discover the real villain was a JBD BMS throwing occasional fits and taking everything downstream with it when it sulked.

The Cerbo itself was completely innocent. Practically owed it an apology.

Worth checking your BMS comms cable first if you haven't — that UART/RS485 connection can get flaky and cause the Cerbo to have what I can only describe as an existential crisis rather than a clean reboot. Victron's own docs are a bit coy about this edge case but the community wiki has a decent troubleshooting flowchart.

What BMS are you running?

@FrostySocket genuinely curious what the two culprits turned out to be in your case? I had a Cerbo rebooting sporadically last autumn and eventually pinned it down to a loose 12V supply connection on the VE.Direct port — just enough resistance to cause a brown-out under load. Combined with a Multiplus comms fault that was throwing its own errors, the logs were an absolute mess to interpret. Took me ages to separate the two threads.

One thing that helped massively was enabling the Cerbo's remote console logging and just leaving it running for a week — the timestamps on the reboot events eventually showed a clear correlation with high inverter draw rather than anything BMS-related, which pointed me toward the power supply side first.

@FrostySocket yeah really interested to hear what the two issues were as well. Currently dealing with something similar on my narrowboat — Cerbo GX dropping off roughly every 3-4 days, no obvious pattern.

Have checked the 12V supply to it (running off a Victron Orion DC-DC) and that looks stable, but I've not ruled out brief voltage dips when the inverter kicks in.

Is it worth logging the voltage on the Cerbo's input rail separately rather than just relying on what the Cerbo itself reports? Bit of a chicken-and-egg situation if it's rebooting before it can log the event properly.

@FrostySocket yeah go on then, don't leave us hanging — what were the two issues?

I ask because I've got a Cerbo on the boat and one in the static caravan, and the "sporadic reboot" problem is genuinely one of the harder ones to chase down. Could be power supply, could be a dodgy USB device hammering the CPU, could be VRM comms going weird, could be a BMS doing something spicy on the VE.Can bus.

The fact it was two separate things is what makes it so frustrating — fix one and it still misbehaves, so you assume you've fixed nothing and start second-guessing yourself.

Did either issue turn out to be something daft like a cheap USB cable or a flaky 12V supply to the Cerbo itself? That catches people out more than they'd admit.