Unstable AC out on L2 of a split phase system with Quattros

Been mulling this over after a similar headache on my static caravan setup last winter, so thought I'd throw it open here.

I'm running a pair of Victron Quattro 48V/5kVA units in a stacked configuration — not split phase obviously, since we're on 230V single phase here in the UK — but the parallel master/slave relationship throws up its own gremlins. I had chronic instability on the output of my slave unit whenever heavy loads kicked in (immersion heater, mostly). Voltage was wandering by nearly 8V, which was making my Cerbo GX throw fits.

Turned out the CANbus daisy chain between the two units had a dodgy termination resistor at the slave end. Half an hour with a multimeter and a fresh RJ45 sorted it, but it took me weeks to track down.

Now I know some of you are running larger systems — I've seen mentions of 10kVA Quattros on this forum powering smallholdings and off-grid workshops. If you've got multiple LFP battery banks in closed-loop comms (DVCC enabled), I'd be especially curious whether you've seen the slave unit's AC output drift independently of the master.

A few questions worth chewing over:

• Is your VE.Bus firmware identical across both units?
• Have you checked the ESS assistant isn't fighting the battery BMS for control authority?
• What does your Cerbo GX show on the individual inverter tabs — are both units reporting the same AC out voltage simultaneously?

Would love to hear from anyone who's wrestled with this, particularly on bigger three-phase or parallel setups. The Victron community forums are useful but sometimes the UK-specific nu

Hey @MasterSolar, interesting one! With stacked Quattros in split-phase you'll want to double-check your phase master/slave configuration in VEConfigure - it's surprisingly easy to end up with both units trying to assert control independently, which will cause exactly the kind of instability you're describing on L2.

Also worth checking the RJ45 interconnect cable between the two units. I've seen a dodgy connection there cause all sorts of grief with synchronisation. Swap it out for a known-good cable before diving too deep into config changes.

What does your VRM portal show for L2 voltage and frequency over time? A graph would help narrow down whether it's a synchronisation dropout or something upstream on the AC input side.

@QG_Marine has the master/slave side covered, so I'll add something slightly different from my own experience running stacked Quattros.

One thing that caught me out was RS485 communication cable quality between units. I'd used a generic screened cable rather than a proper twisted-pair, and L2 would drift noticeably under asymmetric loading. Swapping to a decent Belden-equivalent twisted-pair with the screen earthed at one end only sorted it immediately.

Also worth checking your VE.Bus system configurator settings — specifically whether both units share identical firmware versions. Mixed firmware on stacked units causes subtle sync instability that doesn't always throw obvious alarms in VRM.

Finally, are both Quattros on the same DC busbar with equal cable lengths to your battery bank? Impedance imbalance between units causes real headaches under high load transitions.

Right, so @QG_Marine and @VivaroWanderer have got the sensible stuff covered, which leaves me free to contribute the bit nobody wants to hear.

Nine times out of ten when L2's gone wobbly on a split-phase setup, someone's plugged something into the wrong inverter and blamed the universe. I speak from experience — spent

Something worth adding that nobody's mentioned yet — check your AC output cable lengths between the two Quattros and your load distribution panel. If L1 and L2 have unequal run lengths or differing cable cross-sections, you'll get impedance mismatches that manifest as exactly the kind of instability you're describing, particularly under asymmetric loads.

I ran into this on my van conversion (different scenario, 24V single-phase, but the principle holds) where a dodgy termination was causing voltage sag on one leg only under load.

Also worth pulling the VictronConnect logs and overlaying L1/L2 output voltage graphs. If the instability correlates with load spikes rather than being constant, that points toward wiring rather than configuration. Constant drift suggests the master/slave sync that @QG_Marine mentioned.

Really practical thread — similar config to what I've been planning for my static caravan so I'm watching closely.

One thing I'd add: have you checked the VE.Bus system frequency settings in VEConfigure? I had a related issue where both units weren't locked to exactly 50Hz in island mode, which caused subtle phase drift under dynamic loads (EV charging is particularly brutal for this — the charger keeps renegotiating current draw).

Also worth asking — are you monitoring L1 and L2 independently on VRM? I found setting up separate grid meter inputs revealed the instability was actually load-related rather than inverter firmware. The graphs made it obvious once I could see both phases overlaid.

What firmware version are the Quattros running? There were some known stacked-phase sync fixes in releases around v500+.