Upgrade Cerbo GX from v3.64 to v3.79 lost Renogy BMS connected via RV-C

Had a similar headache last month after pushing the Cerbo GX update on my static caravan setup.

Was running v3.64 without a single grumble — Renogy battery bank showing up perfectly over RV-C, everything talking nicely. Updated to v3.79 and the BMS just vanished from the device list. No errors, no warnings, just... gone.

Tried the obvious stuff first:

• Power cycled the Cerbo
• Checked all the CAN bus termination resistors
• Re-ran the RV-C network scan
• Swapped cables

Nothing. Spent a good few evenings on it before I found a thread buried in the Victron community forum suggesting the v3.79 image changed something in how RV-C devices are enumerated. Seems like certain Renogy BMS firmware versions are no longer being recognised properly after the update.

Rolled back to v3.64 using a fresh SD card and the Renogy came straight back up. Staying put for now until there's a proper fix or someone figures out a workaround.

Anyone else hit this? Particularly interested if folk are running the Renogy 200Ah lithiums — I've got four of those in parallel and wondering if it's specific to that model's BMS or a broader RV-C enumeration issue across the range.

Also curious whether anyone's managed to get it working on v3.79 by adjusting any of the RV-C instance settings manually. I know Victron's VRM portal has some diagnostic tools but haven't had much luck there yet.

Would be good to pool knowledge on this one rather than everyone reinventing the wheel separately. Drop your firmware versions and hardware configs below if you're affected.

@RetiredElectrician this is a known regression in the v3.70–v3.79 range affecting RV-C device enumeration — Victron acknowledged it on their community forum a few weeks back.

Worth trying two things before rolling back:

1. Disable then re-enable RV-C under Services in the Cerbo menu — sometimes forces a fresh device scan
2. Check /data/conf/rvcservice.conf via SSH; the DGID mapping for Renogy BMS occasionally gets corrupted during OTA updates

I hit something similar when I upgraded my Vivaro-based setup — Fogstar BMS disappeared entirely until I manually re-paired. The underlying issue seems to be how v3.75+ handles the BAT_STATUS DGN response timing differently.

If neither works, v3.71 is still available as a manual firmware file on the Victron Professional portal — much more stable with third-party RV-C devices in my experience.

@VivaroWanderer so Victron broke RV-C enumeration and the fix is... another update that's not out yet — classic "we'll patch the patch with a future patch" energy 🔧

Rolled back my Cerbo to v3.64 via the offline firmware file on Victron's site and the Renogy BMS came back like nothing happened — took about 8 minutes flat. If you're not sure how to pull the offline .swu file, it's buried under Professional Downloads on the Victron portal (needs a free account).

Worth noting: RV-C and VE.Can on the same bus seems to make this worse in my experience — isolating them helped stability before I gave up and rolled back entirely.

@RetiredElectrician ran into almost exactly this with a Fogstar-adjacent BMS on my shepherd's hut build — RV-C enumeration going silent after a firmware bump is deeply frustrating when you're relying on that state-of-charge data for load management.

Worth checking whether the Cerbo is still seeing the RV-C bus at all via the Venus OS diagnostic logs (/var/log/venus/). If the bus is present but device IDs aren't resolving, a temporary workaround some have had luck with is manually specifying the BMS instance number through the dbus interface — not elegant, but keeps you running.

@SaltyMechanic fair point, though Victron's beta channel (venus.local → Settings → Firmware → Online updates → set to Release Candidate) sometimes carries interim fixes ahead of stable. Worth checking whether anything post-3.79 is sitting there before you're left watching a blank battery widget.

@SaltyMechanic the Victron update cycle is just "break it, acknowledge it, fix it in six months, break something else" on an eternal loop — meanwhile my Cerbo's sitting on v3.64 like a smug tortoise and I'm not touching it until someone else has confirmed v3.80 doesn't eat their BMS for breakfast too 🐢

@RetiredElectrician worth checking whether the Cerbo is actually seeing the RV-C device at all post-update — on my narrowboat I had a similar enumeration dropout and the culprit turned out to be the device instance numbering shifting after the firmware change. Pull up the RV-C device list in the Cerbo's remote console (venus.local → Device List) and look for any "Unknown" entries sitting on instance 0.

Victron changed how RV-C device instances are prioritised somewhere around v3.70-ish — there's a note buried in the Venus OS changelog if you dig for it.

Temporary workaround: try manually assigning the Renogy BMS a non-zero device instance via the RV-C configuration panel. Sorted my setup within 20 minutes without rolling back firmware.