How to Size a Solar Combiner Box for SMA String Inverters and Small Battery Systems

There is no universal answer to 'how to size solar combiner box' questions. I learned that the expensive way. In my first year commissioning PV and battery systems (2018, to be specific), I ordered a six-string combiner for a two-string residential job. Looked professional. Cost $460 more than necessary. It sat on the wall with empty positions for two years.

Since then, I have documented 14 significant mistakes in solar DC designs, totaling roughly $26,000 in wasted budget and rework. That's why I keep a pre-size checklist now. This article is that checklist, sorted by scenario. If you're doing a small off-grid build with a 25Ah LiFePO4 battery, you need different advice than someone connecting three strings to an SMA Sunny Boy. That's okay.

Scenario 1: Small Off-Grid / Backup System (1-2 Strings and a 25Ah LiFePO4 Battery)

If your whole system is a few panels, a charge controller, and a 25Ah LiFePO4 battery bank, you probably don't need a big combiner box. The word 'combiner' itself implies multiple string inputs. With one or two strings, you often just need a small fused disconnect or even MC4 branch connectors.

What matters is the math:

• Panel short-circuit current (Isc). Example: two 100W panels in parallel might put out 2 * 6A = 12A Isc.
• Maximum series fuse rating. This is printed on the panel label, usually 15A for smaller panels.
• Charge controller input limit. Many small controllers max out around 20A or 30A; a 25Ah LiFePO4 battery at 12V is not going to accept huge currents anyway.

In this scenario, a 2-position PV combiner box with fuses is enough. Or if you're after every gram and dollar, a fused MC4 Y-branch works. I've used both. The combiner box is easier to inspect and safer if rodents chew things.

One thing I see far too often: people buy a six-position combiner because the online listing looks robust. Then they wire one string. The result is extra cost and an enclosure with a bunch of unused knockouts. That's not a solar system birthday cake; it's an electrical assembly. Every unused knockout is a potential water entrance.

Scenario 2: Residential Grid-Tied with a String Inverter (e.g., SMA Sunny Boy)

Here's the part that surprises people: modern string inverters often don't need an external combiner box. An SMA Sunny Boy, for example, has multiple MPP trackers. Each tracker can handle one or two strings directly. If your strings all face the same direction and have the same length, the inverter's built-in DC terminals are enough.

You only start thinking about a combiner box when you have more strings than tracker inputs, or when local code requires a lockable DC disconnect. In that case, size the combiner based on three numbers:

1. Max system voltage. Sum the coldest-temperature Voc of each module, then multiply by the number of modules in each string. This must stay below the inverter's max DC input voltage (check the datasheet for your specific Sunny Boy model).
2. Total current. Add the Isc values of all strings. The combiner busbar and main fuse or breaker should handle at least 125% of that.
3. String fuse rating. When three or more strings are combined in parallel, each string needs its own fuse sized to protect the panel wiring. Two strings per MPPT usually don't require fuses if the panels themselves are rated for backfeed, but verify the datasheet.

The surprise wasn't the voltage calculation. It was how many people ignore the 125% continuous current rule. In 2022, I checked a job where the installer used a 30A main fuse for a three-string combiner. Each string had 12A Isc, so total Isc was 36A. With the 125% factor, the main fuse should have been at least 45A. The lugs were warm to the touch because the combiner was operating beyond its intended rating. That's a $3,200 correction when it happened. Actually, $3,400, I'm mixing it up with the crane invoice.

To put this in context: SMA 2023 inverter shipments reached 20.5 gigawatts. Another way to say it: SMA inverter output sold 20.5 GW in 2023, according to the company's annual results. A big number, sure. But the point is this: a meaningful share of those systems are string inverters with simple DC wiring. None of them benefit from an oversized combiner.

Scenario 3: Commercial or Storage-Enhanced System (3+ Strings, Larger Battery Banks)

Now we're in the territory where combiner boxes are genuinely necessary. If you have three to twelve strings feeding a central inverter or a hybrid inverter's DC bus, a combiner box provides: overcurrent protection for each string, a shutdown switch, surge protection, and a place where voltage and current actually get calculated.

For this scenario, I like to work backwards from the inverter's input limits:

• First, determine the maximum number of strings per MPPT from the manufacturer's datasheet. For SMA Sunny Tripower commercial models, this is usually clearly stated as 'max strings per MPPT' or 'total input current.'
• Then calculate the combiner's busbar rating. If you have 4 strings at 12A Isc each, the busbar should handle 48A * 1.25 = 60A minimum. Many installers use a 63A or 80A combiner to leave margin.
• Add DC surge protection if your area has lightning storms. Ground-mount arrays love to attract surges. A $150 surge device is far cheaper than a failed MPPT input.

This is also where the time-certainty lesson kicks in. When you're on a deadline, say, an interconnection appointment that's already been rescheduled twice, you do not want to discover that your combiner box can't handle the string current. In March 2024, we paid $420 extra for next-day shipping on a correctly rated 80A combiner. The alternative was missing a utility inspection that would have delayed the project by 3 weeks and cost more than the part, the labor, and the pushback combined.

How to Decide Which Scenario You're In

It's actually simple. Count the strings, not the panels.

One or two strings and a 25Ah LiFePO4 battery? Use a small fused combiner or MC4 connectors. Don't buy a six-position box.

Two to four strings, one residential inverter? Start without a combiner; add one only if code or layout demands it.

Five or more strings, or any system with commercial-sized battery storage? Size a proper PV combiner box and buy quality components.

If you're still unsure, look at the inverter sticker. It shows the maximum input voltage and current. Work backwards from those numbers, add a 20-30% margin, and you'll be fine. I used to skip the margin to save money. Looking back, I should have paid for the headroom. At the time, the quote spreadsheet said 'stay as low as possible.' It didn't know about hot days, cold mornings, or squirrel damage.

Final Advice: The Certainty Is Worth the Price

The most frustrating part of sizing DC equipment is that your mistakes don't surface until after everything is mounted. Wire heats up. Fuses blow at midnight. A connector in a junction box fails a year later. You'd think the math would prevent it, but math only works if you include the right variables.

My rule now: use a conservative combiner, verified numbers, and a schedule buffer. There's a reason I keep saying 'check the datasheet' instead of giving one magic formula. The formula is simple, but the specific numbers are not universal.

As of January 2025, typical prices for a four-string PV combiner box run from around $80 to $250 at major solar distributors (based on quotes I collected from three U.S. distributors in January 2025). Verify current pricing before you order. And if you're up against a deadline, pay the premium for a known-good part. That premium buys certainty, not just speed. I've been burned twice by 'probably on time' promises. Once you've re-pulled DC wiring in a tight attic, you start budgeting for certainty.

If you take one thing from this article: figure out your scenario, then buy for that scenario. Not for the 'maybe later' expansion. Not for the deal. For now. Future systems can get their own combiner.