When 4 inputs aren't 4 MPPTs A Sunsynk 8kW diagnosed by graph

Patrick pinged me this week about his Sunsynk 8kW. The system was less than twelve months old — installed by All Seasons Energy, who are now in administration — and the numbers weren't adding up. Three strings of panels (9, 8 and 5) on the roof, and the total generation just wasn't where it should be.

It was a question of where the daylight was actually going. Let's start with the inverter.

The first thing that wasn't right: four inputs isn't four MPPTs

Sunsynk's 8kW hybrid has four physical PV input connectors. So at a glance, you might think you have four independent MPPT inputs and you can hang anything off them.

You don't. Inside the inverter those four connectors are wired as two pairs in parallel. There are only two real MPPT trackers — they just give you a second connector each so you don't have to use a Y-splitter at the roof. Anything you plug into a pair has to behave electrically as if it were one combined string.

That distinction matters. An MPPT is an algorithm hunting for the voltage that gives the panels their best power output at any given instant. Each tracker can only sit on one voltage at a time. If you parallel two strings of different lengths into the same MPPT, you've handed it two different optimum voltages and asked it to find one — which doesn't exist.

What Patrick actually had on the roof

Three strings — 9 panels, 8 panels, and 5 panels. Two MPPTs to land them on. So the installer had to make a choice. There were exactly three options:

• 9+8 parallel, 5 alone — closest voltage match, only a one-panel gap.
• 9+5 parallel, 8 alone — four-panel gap on the pair. The bad one.
• 8+5 parallel, 9 alone — three-panel gap. Better than option two, worse than option one.

All Seasons Energy chose option two. The four-panel-gap option. The 9-panel string and the 5-panel string were paralleled into the same MPPT, with the 8-panel string on its own.

That single wiring choice was the entire problem.

The signature, every sunny morning

This was the moment the diagnosis became free. Five sunny days in a row, the same cliff at the same time of morning, the same plateau for the rest of the day. If the weather had been unsettled, this would have taken hours of detective work; the perfect-sun week handed it over.

What you're looking at is the MPPT making a survival decision. Up to about 08:13 the voltages on the parallel-paired strings are still close enough that the tracker can hold a compromise that works for both. As the sun rises higher and irradiance ramps the panels into their normal operating range, the voltage gap between a 9-panel string and a 5-panel string widens — and at some point the algorithm gives up and clamps to the lower of the two power peaks. From that moment until sunset, the longer string is being run below its operating point and the system is leaving generation on the table.

The bodge — and what a clean fix would look like

The fastest possible fix is a re-pairing. Swap which connectors go where so the inverter sees 9+8 paralleled on one MPPT and the 5 on its own. The voltage gap shrinks from four panels to one, the MPPT can hold both peaks, and the cliff disappears. That's the bodge.

With the re-pair done, the system performs properly again. Not perfectly — the 9+8 pair still isn't an exact match, and the 5-panel string then has a whole MPPT input to itself — but it's the best you can do with the inverter you've already paid for and the panels already on the roof.

A cleaner fix exists, but cost-considerations kick in. Options on the table:

• Add a second small inverter so the 5-panel string lives on its own properly matched MPPT.
• Swap the Sunsynk for one with three or more independent MPPTs.

For Patrick, the re-pair gets him 90% of the way back. The cleaner fixes are on the shelf for if and when he wants them.

What Patrick said

Prompt and useful support regarding my Sunsynk system lack of performance. Ronald simplified the technical issues to make them understandable. Many thanks, looking forward to follow up.

Bad engineering design strikes again

Patrick paid for a 22-panel solar system. The hardware was there. The installer's design choice — paralleling two strings four panels apart on a single MPPT pair when a one-panel-apart option was sitting right there — was costing him generation every sunny day for twelve months before anyone looked.

This is the part of the industry I keep writing up. A homeowner pays a five-figure sum for solar, gets a system that quietly underperforms by 10-15% because somebody skipped the first-year design rule on string matching, and then has no recourse when the installer disappears into administration. The hardware is fine. The panels are fine. The inverter is fine. The engineering decision behind the wiring is what was broken.

If your system has more strings than the inverter has independent MPPTs, and the strings aren't all the same length, get someone to look at the wiring. It's a ten-minute job to confirm or rule out from monitoring data.

If you suspect this might be your story too — drop me a line. The graph tells you within a single sunny day.

Written up · STS field note · May 2026 · Installer gone bust? · All customer reviews