Engineering Article
The Solar Spec FAQ I Wish I'd Had 7 Years Ago (DEYE, Battery Leases, MPPT, Solar Generators)
-
What actually matters on a DEYE 12kW hybrid inverter datasheet?
-
What's the real AC-to-battery charging efficiency on a DEYE inverter?
-
Battery storage lease vs buying — is the lease ever actually cheaper?
-
Renogy Rover 20A MPPT controller — what do the specs not tell you?
-
What should I actually look for in a solar generator?
-
What's the question almost nobody asks, but should?
I've been handling solar system orders for integration projects for about seven years. In that time I've personally made (and documented) 23 significant mistakes — wrong inverter specs, missed MPPT ratings, misread lease terms — totaling roughly $11,400 in wasted budget. Most of them trace back to not knowing which questions to ask before I signed off on a spec.
So this is the FAQ I keep pinned to my desk. Six questions, in the order I wish someone had forced me to answer them. No "it depends" hand-waving — just specifics that cost me real money to learn.
What actually matters on a DEYE 12kW hybrid inverter datasheet?
If you're pulling a DEYE 12kW hybrid inverter datasheet, the easy part is the spec table. The trap is assuming every number on it is equally relevant. Three figures matter more than the rest, and they're usually the ones people skim.
The MPPT window, not the peak PV input. A unit might advertise a big PV input figure, but if the MPPT voltage range is narrower than your string design, daily yield drops. Check the full MPPT voltage window and the max input current per tracker, not the marketing headline.
AC-to-battery charging efficiency. See the next question — the datasheet figure isn't the field figure, and the gap matters.
Grounding and earthing requirements. I burned a $420 redo in 2019 by not reading the earthing conductor size closely enough. Neutral-ground bonding rules on hybrid systems vary by jurisdiction, and if you miss the difference at spec review, your install team eats the cost on site.
(Note to self: I now keep a one-page grounding checklist taped to the spec folder. Should've done that after the first incident, not the third.)
What's the real AC-to-battery charging efficiency on a DEYE inverter?
Honest answer: the number on the datasheet is a peak figure, and the field number is lower. Any hybrid inverter's AC-to-battery path — grid AC to DC bus to battery — loses somewhere in the 5–15% range at realistic load and temperature, and that range shifts with state of charge, ambient temp, and battery voltage.
Here's the thing about DEYE inverter AC-to-battery charging efficiency that most people miss: charging efficiency and discharging efficiency aren't the same figure, and the round-trip number is the one that matters for backup sizing or any kind of tariff arbitrage. Quoting either half alone flatters the system.
My rule now: never quote efficiency from the datasheet alone. Get the test condition — load ratio, ambient temperature, battery voltage — and ask whether that's a peak or a weighted average. If the vendor won't give you the test condition, that silence is itself an answer.
I still kick myself for not asking the test conditions on a 12kW unit back in 2020. We sized a backup system assuming datasheet efficiency and came up 8% short on runtime when the client actually needed it. Not catastrophic — nobody's system went dark — but embarrassing, and that client remembered.
Battery storage lease vs buying — is the lease ever actually cheaper?
Sometimes. But the headline monthly payment is basically meaningless until you total the whole thing up, and the items buried in the fine print are where the money actually goes.
The conventional wisdom when I started was "leases are just financing with extra steps — always buy." My experience across roughly 60 lease comparisons since then is more nuanced: for commercial clients with a 10-year horizon, buying usually wins. But for clients who genuinely value not owning the asset and not managing replacement, a well-structured battery storage lease can make sense. The catch is that "well-structured" is doing a lot of work in that sentence.
What the monthly number usually doesn't include:
- Installation and permitting (sometimes quoted separately, sometimes "included" at a markup)
- Inverter replacement mid-term (a hybrid inverter typically won't outlive a 20-year battery lease)
- Annual escalation clauses — 2–3% per year over 20 years adds up to a serious chunk of the total
- Early termination fees, which are usually brutal
- End-of-term terms — buyout price, removal cost, or "we'll take it back"
Here's the move I've learned: ask "what's NOT included" before "what's the price." The vendor who lists every fee upfront — even when the total looks higher — usually costs less in the end, and the one quoting a suspiciously low monthly almost always has an answer somewhere for where the rest of the money comes from.
It took me about three years and roughly 60 lease comparisons to understand that the monthly number is only useful as a screening tool. The real comparison is total cost of ownership across the term, with the boring line items filled in.
Renogy Rover 20A MPPT controller — what do the specs not tell you?
The Renogy Rover 20A MPPT charge controller is a solid unit for small standalone systems. Where it bites is when someone tries to stretch it past what the label implies.
The 20A rating is charge current, not panel wattage. A 20A controller on a 12V system tops out around 240W of usable PV, and slightly more on 24V. If someone's sizing a 600W array and assuming the 20A label covers it, they're going to clip hard most of the day and wonder why the battery isn't full by evening.
Open-circuit voltage margin is the one that actually burns components. Panel Voc rises in cold weather — a string rated at 100V Voc at 25°C can push past 110V at -10°C. The Rover's Voc ceiling is a hard limit, not a suggestion. It cost us two controllers in a single winter by not derating for temperature. Forty units on that order; we returned the two and rewrote our string voltage sheet. Lesson learned, and now written into the pre-order check.
Communications reliability is fine for monitoring, not for control. Bluetooth and the app are good enough to check state of charge from the truck. They are not something to build monitoring logic around. The controller runs the show; the app reports on it.
What should I actually look for in a solar generator?
Whatever the marketing says, the spec sheet is where the truth lives — and most of the numbers people compare aren't the ones that decide whether the thing works for them. Here's what I tell clients when they ask what to look for in a solar generator:
- Continuous output watts, not "peak." Peak is the number on the box; continuous is what actually runs the load. A unit advertising 3,000W peak with 1,500W continuous is a 1,500W unit.
- Battery chemistry and cycle life. LFP (lithium iron phosphate) is the current default for a reason — longer cycle life, better thermal behavior. If the spec doesn't say LFP, assume it isn't.
- Charge time from both AC and solar. "Solar ready" often means the MPPT input is small enough that a cloudy day barely tops it up.
- Can it take an external MPPT? Some units lock you to a proprietary panel connector, which limits you to their ecosystem. That's a pass if expansion is on the roadmap.
- Warranty terms on the battery. Look at what counts as "normal degradation." Anything below 80% nameplate capacity after the warranty window is usually your problem, not theirs.
If a client asks me which unit to buy, I give them the same line: pick the one whose continuous output covers their biggest single load plus a 20% buffer, then look at everything else. Everything else is a feature; that one is a requirement.
What's the question almost nobody asks, but should?
"What happens to my system when the manufacturer's cloud shuts down?"
Most hybrid inverters and battery systems report through a cloud portal. That portal is a product roadmap decision for the manufacturer, not a guarantee for you. If the monitoring service gets discontinued, or the app gets abandoned, or the local gateway depends on a server neither of us can control, the system still runs — but your visibility into it disappears, and any feature gated behind the cloud goes with it.
The second version of the same question: "does it run standalone, with local control, when the internet is down?" For backup-critical installs, the answer should be yes, and it should be documented before the PO goes out. Not after the panel is on the wall and the client is asking why the app shows nothing.
(I really should put this on the pre-order checklist. Adding it now — see, this is why I keep the document.)
Ask a technical follow-up