House DC bus stays energised by Victron chargers when main battery isolator is switched OFF

Ran into something similar with my shepherd's hut build that had me scratching my head for a good while.

I've got a Fogstar Drift 200Ah LiFePO4 behind a main isolator, with a Victron SmartSolar and a Victron IP22 shore charger both feeding in. What I noticed is that even with the isolator open, the DC bus downstream stays live — fridge, 12V lights, the lot. The chargers are effectively backfeeding through the load side and keeping everything energised.

It's not dangerous as such, but it does mean:

• The isolator isn't actually isolating anything whilst a charger is connected and active
• Any fault on the load side stays live even if you think you've killed the bank
• Slightly unnerving if you're doing any wiring work and assume the bus is dead

My current workaround is just disconnecting shore power and waiting for cloud cover before touching anything — not exactly elegant.

I think the proper solution is to either:

1. Put the isolator between the chargers and the bus, not just between the battery and the bus
2. Use a Victron BatteryProtect on the load side so loads drop off cleanly when the battery disconnects

But I'm curious whether anyone else has thought this through more carefully, particularly if you're running a BMS with a charge/discharge relay in the loop — does that change how you'd wire the isolator position?

Would be interested to hear how others have tackled it, especially on smaller off-grid setups where you don't necessarily want to over-engineer things.

@WattCharlie the Victron's doing its best impression of a zombie — technically "dead" but still haunting your DC bus through the PV input side 🧟

@WayneTaylor's zombie analogy is spot on 😄

Worth adding — this catches people out because the SmartSolar's PV input and the battery port are not isolated from each other internally. When you open the isolator, the charger itself becomes the path. Even with panels producing minimal current at dusk, you can measure a surprising 13–14V on the "dead" bus.

Practical fix I used on my Vivaro build: install the isolator between the SmartSolar load/battery terminals and the bus, not between the battery and the charger. That way opening the isolator truly breaks the downstream circuit regardless of what the charger is doing.

Alternatively, a Victron BatteryProtect on the DC distribution side gives you remote switching without back-feed headaches.

The Fogstar Drift's BMS won't protect against this scenario — it's purely managing cell-level events, not topology issues like yours.

@WayneTaylor's zombie analogy deserves a gold star, but I'd extend it slightly — your isolator isn't actually isolating anything, it's just offending the battery whilst the Victron cheerfully ignores the whole situation and carries on regardless.

Seen this exact scenario on my narrowboat. The load-side terminals of a Victron

@YorkshireBoater learnt this the hard way in my shepherd's hut when I nearly "isolated" myself into a false sense of security — your isolator needs to be on the load side of every charger input, otherwise you've just built a very expensive way to feel safe while your DC bus laughs at you.

@BatteryPaula raises the critical safety point here — a "dead" bus that isn't actually dead is a genuine shock risk.

One thing nobody's mentioned yet: the load output on the SmartSolar is also worth checking. If you've wired loads directly to that terminal rather than to your battery, they'll stay powered regardless of your isolator state too.

The cleanest solution I've implemented on my narrowboat is a Victron BatteryProtect on the DC bus downstream of everything, configured as a true system isolator. Unlike a manual switch, it can be driven by the BMS's pre-alarm signal to disconnect before the BMS hard-cuts — much more elegant.

Also worth verifying with a multimeter before you touch anything — never assume a bus is dead just because the isolator's open. Measure first, work second.