Second Quattro QUA488024000 (fw 499) in parallel + mixed US5000/US3000C Pylontech bankdiy

Been down a similar rabbit hole lately, so figured I'd chip in.

Running a single Quattro 48/5000 in my static setup — emergency backup for the house, same sort of ESS arrangement. Started wondering whether throwing a second unit in parallel was worth the hassle versus just upgrading to a larger single inverter.

The Pylontech mixing question is the bit that really kept me up at night. I've got a US3000C bank and was eyeing up some US5000s to expand capacity. Everything I've dug up suggests Victron tolerates the mixed bank but doesn't exactly endorse it — the US5000s communicate differently enough that things can get a bit twitchy under the DVCC settings.

Few things I'd want confirmed before committing:

• Firmware parity — both Quattros need identical firmware, full stop. Seen too many threads where someone had a unit sitting in a box for six months and forgot to check.
• VE.Bus cabling — the interconnect cable length matters more than people admit in parallel setups
• BMS priority — which unit the Pylontech actually "talks" to when you've got two inverters in the mix

The van conversion side of my life has made me deeply suspicious of complexity for its own sake. Every additional component is another thing to diagnose at 11pm in the rain.

Anyone here actually running a mixed US5000/US3000C bank with parallel Quattros long-term? Genuinely curious whether it's settled down or remained a source of grief.

Hey @GafferTapeKing, looks like your post got cut off there, mate — "Started wondering w..." — did the forum eat the rest of it? 😄

Keen to hear where you were going with that though. I'm running a similar single Quattro 48/5000 setup myself up in Yorkshire, ESS with a small Pylontech bank (three US3000C's). The mixed US5000/US3000C combination you've mentioned in the thread title is something I've been curious about — specifically whether the BMS communication behaves itself when the cell capacities differ between units. Have you actually got both battery types running together already, or still planning stage?

Also worth clarifying — are you on DVCC with the Pylontech protocol enabled through VenusOS? Makes a fair difference to how gracefully everything talks to each other.

@YorkshireExplorer yeah the forum does that sometimes, mobile browser I'd guess.

@GafferTapeKing — interested to see where you were going with this. Mixed Pylontech banks are a bit of a minefield. US5000 and US3000C have different cell counts (15S vs 14S I believe) which means slightly different charge voltages. Victron's DVCC helps smooth things over but it's not a silver bullet.

If you're considering paralleling Quattros on top of that, make sure your firmware is identical across both units — fw 499 mismatch can cause sync headaches that aren't always obvious until load spikes. VE.Bus configuration via VEConfigure needs doing as a parallel pair too, not individually.

What's your intended use case — peak shaving, full backup, or something in between? Makes a difference to how aggressive your charge curves need to be.

@GafferTapeKing mixing US5000 and US3000C in the same bank is where Pylontech's own documentation quietly clears its throat and looks at the ceiling — different cell counts, different internal resistance, the BMS will technically allow it but you're essentially asking two slightly different batteries to pretend they're identical twins on a job application.

Victron's ESS assistant plays nicely with Pylontech over CAN-bus, but with parallel Quattros and a mixed bank you've got three potential sources of argument happening simultaneously — been there on the narrowboat with mismatched cells and it's a debugging nightmare at 11pm in a wet marina.

Worth checking whether fw 499 handles the dual DVCC arbitration cleanly; there were known quirks before ~500 on parallel Quattro setups.

@ExPostie86 that's what's been nagging at me — does Pylontech's BMS actually coordinate properly across the two models when they're in the same group, or does the US5000's higher capacity just mean it cycles harder whilst the US3000C sits relatively idle?

Wondering whether the VRM data would reveal that if you logged cell voltages per module over a full charge/discharge cycle. Has anyone actually pulled that granular data and compared state-of-charge across mixed units during peak draw?

Also curious whether the Quattro's charge current gets distributed sensibly or whether the BMS just broadcasts a single target and the modules do their own thing internally.