ESS immediately shifts to 63 Hz and disconnects Huawei inverter when AC Input is present but rejecte

Ran into something uncannily similar last winter with my setup — Victron Quattro 48/5000 paired with a Sofar Solar inverter AC-coupled on the output side, Fogstar Drift lithium pack with DVCC enabled.

The moment grid input was present but getting rejected (voltage out of tolerance, which happens constantly round here in rural Wales), the Quattro would shunt straight to that ~53 Hz frequency shift to curtail the AC-coupled inverter. Normal behaviour in isolation — but with a fussy inverter on the output that interprets anything outside 50 Hz ±2 as a fault condition, it just... drops off entirely and takes ages to reconnect.

The core tension seems to be ESS frequency shifting fighting against inverters that weren't really designed with Victron's AC coupling frequency-watt curve in mind. Huawei's SUN2000 range is notoriously conservative about frequency deviations from what I've read on the Victron Community forum.

A few things worth checking:

• Grid code settings on the Huawei — some firmware versions let you widen the frequency tolerance window
• ESS minimum SOC and feed-in excess settings, which can trigger more aggressive frequency shifts than you'd expect
• Whether your DVCC shared voltage sense is playing nicely — I had ghost voltage readings causing erratic behaviour

Curious whether others are seeing this with specific Huawei firmware versions, or if it's more of a fundamental incompatibility with ESS frequency shifting logic.

Anyone successfully stabilised an AC-coupled Huawei on a Quattro-based ESS without it throwing a fit every time grid drops in and out? Would love to know what actually worked.

Hey @48VWizard, interesting one! Just wanted to flag something that might be relevant — have you checked your ESS assistant version? There was a period where older ESS firmware had a quirk where the frequency shift would trigger almost immediately if the AC input voltage was even slightly outside tolerance, rather than waiting through the proper rejection window.

Worth pulling the VEConfigure logs to see exactly what the Quattro "saw" on the AC input at the moment of disconnect. Sometimes the grid is momentarily dipping or spiking in ways you wouldn't notice otherwise. A decent clamp meter on the input live during the event can be revealing.

Also, what's your AC input current limit set to? If it's too low for the combined loads, the Quattro can get a bit confused about whether to accept the input or not.

@48VWizard this is a classic frequency-shift rejection loop and I've tumbled down exactly this rabbit hole myself. When the Quattro decides it doesn't trust the AC input — usually because something's slightly off with the voltage or frequency tolerance settings — it boots the AC-coupled inverter off via that 50→53Hz nudge as designed, but then it can get stuck in a rejection cycle if your ESS grid setpoint is misconfigured.

Check your AC Input 1 voltage range in VEConfigure — mine was set absurdly tight (207-253V) which caused phantom rejections constantly. Widening it slightly sorted it immediately.

Also worth confirming your ESS Mode isn't set to "Keep batteries charged" when you actually want self-consumption — that combination with DVCC on Fogstar packs produces some genuinely baffling behaviour that looks exactly like what you're describing.

@48VWizard one thing worth double-checking that nobody's mentioned yet — your AC Input current limit within the ESS assistant itself. If the Quattro's input current sense is briefly reading above that threshold (even transiently during switching), it'll reject the AC input and the ESS logic simultaneously tries to offload via frequency shift to the AC-coupled inverter. You end up in a feedback loop where the Huawei sees 53Hz+ and disconnects, which then dumps load back onto the Quattro, which briefly stabilises, rinse and repeat.

In my garden office setup I had similar behaviour with DVCC active — the fix was increasing the minimum input current limit slightly and also checking the sustain voltage wasn't fighting the BMS's own lower CVL. Worth pulling the VRM logs and correlating the frequency graph against the input current readings millisecond by millisecond.

@48VWizard had almost the exact same headache with my Quattro 15 months ago. Worth checking your grid code settings — if the Quattro is configured for a stricter grid code (say, G99 rather than G98), it can reject AC input that doesn't meet tighter voltage/frequency tolerances, which then triggers the frequency shift almost immediately as it islands. The Huawei sees that 63Hz signal and dutifully shuts down, which is correct behaviour but masks the root cause.

Have a look in VEConfigure under the inverter tab → grid and confirm the code matches what your DNO actually requires. Also worth cross-referencing with @JubileeClipHero's point — the two issues can compound each other quite nastily.