The Most Important Solar Decision Nobody Tells You to Make

A solar proposal tells you the panel brand, the wattage, the price, and the monthly payment. It rarely names the part that matters most: the power electronics that turn what your panels make into power your house can use. Those components are the engine of the system. They decide how it performs, how it ages, what a repair costs, what your battery options are, and how many separate companies you will deal with for the next twenty-five years. It is the most consequential choice on the project, and it is almost never presented to you as a choice at all.

Your installer already picked the engine, often before you met, and often for reasons that had little to do with your roof: which system their crew installs fastest, which manufacturer they have a relationship with, which equipment their lender approves. None of that is necessarily wrong. But the engine should be specified for your home, the way a good mechanic spec’s a part for your car, not handed to you as a default you never knew you could question. This is the explanation that lets you question it.

The one idea everything rests on

Panels make DC power. Your home runs on AC. Something has to convert one to the other, and the only real questions are where that conversion happens and how many times. Everything below follows from that, including what happens when you add a battery. (If DC versus AC, or power versus energy, is still fuzzy, our piece on reading the numbers covers the groundwork.)

At the level of the panels and their power conversion there are really only two architectures. Everything a salesperson might name at you is one of these two, or a variation in how one is deployed.

Architecture one: the AC system (microinverters)

In an AC system, a small inverter sits under each panel and converts that panel’s DC to AC on the roof. Enphase is the name you will hear most; some panels, such as certain Q Cells models, arrive with a microinverter built in. Either way, the power leaves the roof already AC.

The advantages are real. Each panel works independently, so shade or a fault on one does not drag down the others; you get monitoring panel by panel; and with no high-voltage DC running down through the building, there is a genuine safety benefit.

There is also a ceiling worth understanding, because it causes a misunderstanding that trips up almost every homeowner. A microinverter can only convert so much; Enphase’s high-current IQ8HC, for example, tops out around 380 watts. So a 430-watt panel does not hand you a full 430 watts of AC beneath a microinverter rated below that. At the brightest moment of a clear day the microinverter passes only what it is built to pass, and anything above that line is left on the table. This is called clipping, and the lesson is simple: the nameplate on the panel is not the power you get; the pairing is.

Now the part that keeps it honest. At a modest mismatch like that one, the lost energy is tiny, often a fraction of a percent over a year. A panel rarely produces its lab-rated wattage on a real roof anyway, once you count heat, dust, and angle, and the only thing clipped is the sliver above the ceiling during the few brightest hours. Pairing a somewhat larger panel with a somewhat smaller microinverter is normal, sound engineering, and it is how most reputable systems are built. You may hear an installer say the panel’s extra dawn-and-dusk output “more than offsets” the clipping; within a modest range that is broadly true. Treat it warily as a blanket claim, though, because it is also a favorite reassurance line, and it stops being true as the mismatch grows: push the panel far enough past the microinverter and clipping climbs steeply, especially on sunny California roofs, until you are paying for wattage thrown away every clear afternoon. So the useful question is not whether your panels are larger than your microinverters, which is fine and expected, but whether the mismatch is modest or aggressive. Ask for the system’s DC-to-AC ratio and what the modeling shows for clipping on your roof. A good installer has those numbers ready; a careless one tells you not to worry about it.

Enphase is also a closed ecosystem: its microinverters, gateways, combiners, and batteries are built to work with each other and not with outside equipment. That closure cuts both ways, and the upside is worth saying plainly, because it answers a problem we will get to. When the whole system is one company’s, you have one company to call, one warranty, one app, and no one to point at but themselves. For a homeowner who wants simplicity, that is a real benefit, not a marketing line.

Architecture two: the DC system (string inverter)

In a DC system, the panels send DC down to a single string inverter that converts it once, usually on a wall near your panel. For a home this is a string inverter; “central inverters” are a commercial and utility-scale thing, and you should be mildly suspicious of any residential proposal that uses the grander term.

A plain string inverter gives you no per-panel control: wired in series, the weakest panel can hold back its whole string, so on a shaded or multi-direction roof a bare string system gives up production. On a clean, unshaded, single-orientation roof that hardly matters, and a quality string inverter is the simplest, usually least expensive, and longest-proven way to build. The established brands are mature technology, and maturity counts in a component you expect to run for a decade or more.

The DC architecture is also where you have the most freedom, because the pieces are not all forced to come from one company.

Safety, sized to your roof. Every rooftop system needs a way to shut the panels down at the module level so firefighters are not facing live DC. On a DC system this is a small device on each panel, and here is the unadvertised part: inverter-agnostic makers like Tigo and APsmart let you buy exactly as much capability as your roof needs. At the simplest tier the device only satisfies the safety rule, listening for a “keep-alive” signal and shutting the panel down if it stops; no app, no monitoring, nothing to manage, and you would never know it was there. If your roof has shade or mixed orientations that would benefit from monitoring or optimization, those same families offer stepped-up versions that add exactly that. You pay for what your roof needs, chosen independently of the inverter and the panels. This is the one place the brands genuinely cooperate, and only because a shared safety standard the code forced on them (SunSpec’s rapid-shutdown signal) makes that one handshake work across brands. It covers the shutdown signal and nothing else.

SolarEdge: a DC system in its own clothes. SolarEdge is still a DC system with a string inverter, deployed in a distinctive, more closed way. It puts a mandatory optimizer on every panel. The optimizer is not an inverter; it conditions each panel’s DC and handles per-panel tracking, monitoring, and shutdown, while the actual DC-to-AC conversion still happens once, at the string inverter. So you get the per-panel benefits an AC system is sold on, per-panel optimization, panel-level data, built-in rapid shutdown, while the power stays DC to the inverter, which, as the battery section shows, has real advantages.

Like Enphase, SolarEdge is a closed ecosystem: the optimizers work only with a SolarEdge inverter and the inverter only with SolarEdge optimizers, so the PV side is committed end to end, and its own battery is captive to SolarEdge equipment the way Enphase’s battery is captive to Enphase. That closure carries the same single-source upside, one company for warranty, service, and support, and a cost beyond the lock-in: an optimizer on every panel is one more piece of power electronics on every panel, and in the practice’s field experience, more electronics on the roof tends to mean more service calls over the years. That last point is our view from the field, not a published failure rate, and a fair-minded installer can argue it; the lock-in is simply in the spec sheets.

A word on reliability, since everyone asks and the honest answer resists a slogan. A string inverter is a single point of failure and a near-certain mid-life replacement; when it goes, the system stops until it is fixed. Microinverters spread that risk, so one failure costs one panel, not the array, and modern Enphase units have earned a strong record. But each microinverter or optimizer that fails is a trip onto the roof to replace. No architecture is simply “more reliable.” There are different failure modes, and the right one depends on your roof and how long you intend to own the home.

The battery decision is where the engine choice comes due

Everything above is the PV side. Storage is a separate system, but it is where the architecture you chose quietly dictates your options, and for most California homeowners shopping now a battery is part of the plan.

Start with efficiency. A DC-coupled battery, wired on the DC side, charges with fewer conversions, at a round-trip efficiency of roughly 95 to 98 percent. An AC-coupled battery, fed by a system that already converted everything to AC, has to turn it back to DC to store and to AC again to use; that round trip lands closer to 90 to 94 percent. Tesla’s own datasheets show the gap: the DC-coupled Powerwall 3 is rated 97.5 percent, the older AC-coupled Powerwall 2, 90. An AC system, because it makes AC on the roof, can only AC-couple a battery and pays that penalty every day it cycles. A DC system can go either way.

There is also a sizing ceiling no proposal explains. A battery’s built-in inverter has a power limit, often lower than you’d guess. The Powerwall 3’s inverter is rated at 11.5 kilowatts of AC. Wire solar straight into it as a DC system and Tesla supports an array up to 20 kilowatts against that inverter, close to a 1.74-to-1 ratio, because the excess DC can be steered to the battery. Pair that same Powerwall with a separate microinverter array instead, AC-coupling the two, and the solar you can put on it drops to roughly 7.68 kilowatts per Powerwall by Tesla’s own rule. Same hardware, a very different system you are allowed to build, decided entirely by how the power is converted.

One common misunderstanding is worth correcting: AC-coupling a battery is not a special trick of one ecosystem. Almost any system that outputs AC, microinverter or string inverter, can generally pair with a third-party AC-coupled battery, because AC is AC. FranklinWH, for one, markets itself as working with essentially any inverter brand, and the Tesla Powerwall 2, Q Cells’ AC storage, and Lunar Energy’s system are likewise built to AC-couple onto an existing array. The real lock is narrower: it is the proprietary, own-brand batteries that are captive. SolarEdge’s own battery takes power only from SolarEdge equipment, and Enphase’s requires Enphase, but a SolarEdge or Enphase array can itself AC-couple to an outside battery like most systems. Two caveats keep this honest. The genuine architecture-level advantage that remains is DC-coupling for that efficiency; an AC system cannot do it. And “can generally AC-couple” is not “any battery works with any inverter”: code and each manufacturer’s approved-pairing lists impose real, variable limits, and confirming a specific pairing is the installer’s due diligence. Ask them to show you they checked.

The pattern underneath all of it

Step back and the logic is plain. These manufacturers are not partners; they are competitors fighting over the same customers, with little reason to make their products work with one another’s, or to help you when a problem might belong to someone else’s box. Outside that one safety standard the code forced on them, there is no broad cooperation and no plug-and-play across brands. The lock-in is not an accident; it is the business model.

For you, that surfaces as fragmentation. A solar-plus-storage system can carry panels from one company, the power electronics from a second, and a battery from a third, each with its own warranty, its own support line, and its own rules for saying no. Put three manufacturers on one home and you have three companies to call and a predictable scene when something fails and the cause is not obvious: each points at the others. This is not a story about bad people; it is structural. Each warranty rationally covers only its own component, so when a fault sits in the seam between two brands, no one owns it, and you become the one mediating.

The person who normally absorbs all of this is your installer, the integrator who picks components that work together, commissions the system, holds the workmanship warranty, and chases the manufacturer claims for you. That holds right up until the installer is gone, which, after the wave of California solar failures of the last two years, is no longer a hypothetical. When the integrator disappears, the fragmentation you never noticed becomes your full-time job. We cover what to do in that situation, and how to vet an installer’s staying power before you sign, in our pieces on orphaned systems and installer bankruptcies. For now the point is narrower: the engine you choose determines how many separate companies you are exposed to when something breaks.

The honest other side

There is a real case for the closed ecosystems, and it is the mirror image of all this. The same closure that locks you in gives a single-brand system one app, one warranty, one company to call, and no finger-pointing, because there is only one party responsible. The integration is tested, the DC-coupled designs are often more efficient, and when it works the experience is genuinely cleaner. That is exactly how the manufacturers sell it, and the pitch is fair.

So the real decision is not “open good, closed bad.” It is a trade. A single-vendor ecosystem buys simplicity and accountability at the cost of flexibility and being tied to one company’s roadmap, pricing, and survival. A more open, mixed system buys flexibility and independent component choice at the cost of more relationships to manage and more seams where responsibility can slip. Neither is wrong for everyone. What is wrong is being steered into one without ever being told the other existed.

What to do with this

You do not need to become an engineer. You need to make the invisible choice visible, and ask the questions a good installer welcomes and a volume seller dodges. Did you consider an AC or a DC system for my roof, and why this one? If I want a battery now or later, is it DC- or AC-coupled, and what does that mean for efficiency and how much solar I can fit? What is my DC-to-AC ratio, and what does your modeling show for clipping? How many separate manufacturers will my system involve, and who holds the warranty on each part? Are you a certified partner of the brands you are quoting, and is the recommendation tied to a financing program?

In the practice’s view, for a clean and largely unshaded roof a quality string inverter with the right level of safety device is often the most sensible and cost-effective engine, and a DC-coupled design earns its keep the moment storage enters the picture. For a shaded or complex roof, per-panel electronics genuinely help, and the question becomes whether you get them inside a closed ecosystem, with its single-source simplicity and its lock-in, or from inverter-agnostic devices that leave you freer but add a relationship. Both are defensible. What is not defensible is letting the most important component in your system be chosen for you, by someone optimizing for their crew or their lender, without your ever being told a decision was made. The engine should be chosen for your home, with you. Insist on it.

---

The Installer’s View is an independent solar advisory practice for California homeowners. We do not sell, install, or service solar equipment, and we are not paid by any manufacturer named here. This article is general educational information, not legal, financial, or engineering advice, and equipment specifications and company circumstances change; verify current details before making a decision. Product and efficiency figures are drawn from manufacturers’ published documentation as of the date above.