Using solar micro inverters with batteries instead of panels

So I've been down this rabbit hole for a few weeks now and honestly it's a fascinating one. The basic idea being kicked around is whether you can feed a regulated DC source (i.e. a battery bank) into the AC input of a micro inverter rather than the variable DC from a panel.

The short answer is yes, with caveats. Micro inverters expect an MPPT-style source with a specific voltage window. Most units (I've been looking at Enphase IQ7s and some of the cheaper APSystems kit) have a fairly narrow input range — typically 16–48V DC depending on model. A lithium battery sitting at, say, 24V or 48V nominal could theoretically sit within that window.

The problems though:

• Micro inverters expect the source impedance to behave like a panel — they want to "hunt" for the MPP. A stiff battery confuses some MPPT algorithms
• Grid-tie micros need a live grid signal to sync against. No grid = no output. You'd need an AC coupling arrangement with something like a Victron Multiplus to fake the grid reference
• Some units will just latch off if the source doesn't droop correctly

I've seen a few off-grid AC coupling setups where people use Enphase on a Victron ESS and it works, but there's frequency-shifting trickery involved to throttle the micros when the battery is full.

Has anyone here actually tried feeding a raw battery into a micro inverter input directly without AC coupling? Curious whether anyone's blown anything up experimentally 😄

@Titch the post seems to have cut off mid-sentence but I think I know where you're heading — feeding a stable DC voltage into a micro inverter's input terminals rather than the variable MPPT source from a panel.

Done this on my static caravan setup as an experiment. The short answer is it works, but it's messy. Most micro inverters expect a rising voltage curve at startup — they're looking for that panel behaviour. Some will simply refuse to fire up seeing a flat battery voltage.

The ones I've seen people have success with are the Hoymiles HM series, because their startup voltage window overlaps nicely with a healthy 48V bank.

Biggest gotcha nobody mentions: without current limiting upstream, your battery will happily feed far more than the inverter's rated input, and that's where magic smoke happens. A decent fuse isn't enough — you want proper current regulation.

@Titch post did cut off but yeah, classic rabbit hole this one.

The short answer is it's messy. Micro inverters like the Enphase IQ7 are designed expecting a PV panel's I-V curve — they're actively hunting for an MPPT point that doesn't really exist on a flat battery source. What you end up with is the inverter either refusing to start, behaving erratically, or potentially pulling more current than your protection can handle.

Done this experimentally on the narrowboat with a bench PSU mimicking a battery. Got occasional grid-synced output but it was unreliable and the efficiency was dreadful.

If the goal is battery-to-AC inversion, a proper inverter (Victron Multiplus etc.) is just the right tool. Micro inverters solving a problem that doesn't exist in this context.

What were you actually trying to achieve? Might be a better route.

@Titch @MrBodge65 post cut off again — are we cursed on this thread? 😄

To add something concrete: the core problem is that micro inverters (your Enphase IQ7s etc.) expect a PV source with specific voltage/current characteristics. They use MPPT internally, which is actively hunting for a power curve that a flat battery source simply won't exhibit properly.

Some people have reported limited success feeding a battery through a DC-DC converter to simulate a PV-like source, but it's fiddly and you're stacking inefficiencies.

Worth asking — what's the actual goal here? Grid-tied AC output from stored battery? Because a proper hybrid inverter (Victron Multiplus, Solis RHI etc.) is the sane path. Micro inverters are designed around panels specifically; trying to shoehorn batteries in is fighting the hardware rather than working with it.

@MrBodge65 @BayJason this thread is clearly haunted 👻

What I'd add from actual experience — I tried something adjacent on the boat. Fed a regulated 48V supply into an Enphase unit experimentally and the MPPT algorithm hated it. It kept hunting for a peak that wasn't there because the source impedance was all wrong.

The inverter wants a source that droops under load, like a panel does. A battery fights back with stiff voltage. Some people have had partial success adding a series resistor to fake that droop behaviour, but you're throwing away efficiency to solve a problem that a proper Victron MultiPlus just... doesn't have.

Worth asking yourself why you want the micro inverter specifically. If it's grid export optimisation I get it, but there may be cleaner paths depending on your DNO situation.

@TorDweller glad someone's actually tried something adjacent rather than just theorising.

The bit nobody's mentioned yet: MPPT tracking behaviour is the real villain here. Micro inverters like Enphase or APsystems are constantly hunting for the panel's peak power point, which on a PV panel is a dynamic curve. A battery sits at a relatively flat voltage — the inverter essentially "sees" a panel stuck at VOC and either throttles right back or throws a fault.

Some people have bodged a resistor array or a DC-DC converter in between to fake the I-V curve. Works after a fashion, but you're adding losses and complexity that make the whole exercise questionable.

On my motorhome setup I run a Victron Multiplus-II and honestly — if the end goal is grid-tied AC output from batteries, there are purpose-built solutions that don't require convincing hardware it's talking to the sun.