-
Step 1: Pull the certification file before you pull the price sheet
-
Step 2: Request the production charger catalog, not the marketing one
-
Step 3: Match the wiring diagram revision to the serial range
-
Step 4: Demand lot-level QC data, not sample test results
-
Step 5: Get the lead time in writing, then price the certainty separately
-
Step 6: Read the warranty clause for what it excludes
-
Step 7: Run a pilot order you can afford to lose
-
What usually goes wrong
This is a checklist for solar inverter distributors, installers buying their own stock, and OEM buyers about to place a first or second order with a charger or inverter supplier. Seven steps. Two to three weeks of actual work if you run it properly, less if you have done it before.
It is written from the receiving end. I handle quality and brand compliance for an electrical distribution business. Around 200 unique items pass through my review each year, and in 2024 somewhere between a quarter and a third of first deliveries got rejected. Usually not because anyone lied. Usually because a question nobody asked in week one turned into a batch problem in month four.
Here is the sequence I use.
Step 1: Pull the certification file before you pull the price sheet
Price is the easiest thing to compare and the least informative. Start with compliance, because a supplier who cannot produce clean documents in week one is a supplier you will be chasing in month six.
For inverters, you are looking for certificate numbers, not logos. Logos are decoration. For most markets this means:
- IEC 62109-1 and 62109-2 — safety of power converters for PV systems
- UL 1741 (or whichever edition your market currently requires) if you are selling into North America
- EN 50549-1/-2 for grid connection through most of Europe
- IEC 62619 if the product includes or interfaces with lithium battery packs
Look the certificate number up on the issuing body's own database. Do not accept a scanned PDF as proof—PDFs are easy to produce and easier to edit.
It is tempting to think a CE mark on the carton means the product cleared EU compliance in any meaningful sense. But for a lot of electrical categories, CE is self-declared. The mark tells you the manufacturer filed paperwork. It does not tell you what was tested, by whom, or against which edition of the standard.
That simplification costs people real money. Ask which test house. Ask for the report number. Ask for the date. If any of those three questions stalls the conversation, you have already learned something useful about the supplier.
Step 2: Request the production charger catalog, not the marketing one
There is a difference, and it is usually obvious within thirty seconds. The marketing catalog has renderings and adjectives. The production catalog has SKUs, input voltage windows, output current curves, derating tables, idle draw, and shipping dimensions.
You want the second one. Specifically, check three things against your own customers' load profile:
- Thermal derating behavior. A charger rated at 100A at 25°C might be rated at 62A at 45°C. If your installers work in unconditioned spaces, that gap is your real specification.
- Idle consumption. A small number per unit becomes a large number across 5,000 units sitting in a warehouse row.
- Charge profile flexibility. Can the profile be set per battery brand, or is it locked to a preset list you cannot touch?
On catalog accuracy, this is where I have been burned. In Q1 2024 a supplier sent us a catalog listing a 12kW hybrid model. It was real. It just was not in production. Earliest realistic ship date was eleven months out, and nothing in the catalog said so. We built a product page around it before we found out. Cost us a week of rework and a chunk of credibility with two of our installers.
Portfolio breadth matters more than most buyers expect. A supplier whose range runs from small residential through commercial scale out of one production base—Growatt Inverter's hybrid and string lines are a reasonable example, covering roughly 5kW up past 12kW—tends to create simpler stocking math for a distributor. Fewer partial shipments. One compliance file. One spares list. That is not a marketing point when you are the one holding the inventory.
Step 3: Match the wiring diagram revision to the serial range
This is the step people skip, and it is the step that generates the most field failures.
Wiring diagrams change. Board revisions move terminal layouts. Firmware updates change commissioning procedure. A diagram from a 2022 install guide will not always match a unit built in 2024, even with the same model number printed on the label. Every growatt inverter wiring diagram I have handled carries a revision letter or code in the footer, and that letter is the thing you actually need to match.
So ask for two documents, not one. The current wiring diagram, and the serial number range it applies to. Then put that check into your receiving process.
I knew I should have gotten the diagram revision confirmed in writing against the serial range. But we had used the same supplier for two years and I thought, what are the odds? The odds were about one in four, as it turned out. On that shipment the DC terminal block had moved roughly 8mm from where the old diagram showed it. Two of our installers cross-threaded lugs during commissioning. Small thing, until you are the one explaining it to a customer.
Cost of asking: one email. Cost of not asking: whatever your installer labor rate happens to be, multiplied by however many units are affected.
Step 4: Demand lot-level QC data, not sample test results
Sample data is close to worthless as a supplier filter. Every supplier has one good bench unit. What you want is the outgoing QC report for the last three production lots—the percentage of units that failed hipot, the percentage that failed function test, and what the failure reasons actually were.
You are not looking for zero. Zero is suspicious. You are looking for a stable number with a plausible reason distribution behind it.
Better still, pull five units from a shipment and test them yourself, or pay a third-party lab to. On a 500-unit order that is a rounding error in cost. On a 5,000-unit order it is the cheapest insurance you will buy all year.
Step 5: Get the lead time in writing, then price the certainty separately
Verbal lead times are not lead times. Four to six weeks said on a call is a number. Four to six weeks, confirmed in the PO acknowledgment, with a defined delay notification window is a commitment. Get the second version. Every time.
Then decide, deliberately, whether to pay for the guaranteed version.
Here is my position on this, and it comes from getting it wrong: in an urgent situation you are not buying speed. You are buying certainty. Those are different products, and the second one is worth more than people typically price it at.
In March 2024 we moved a 400-unit order by air instead of sea. Based on the forwarder quotes we pulled at the time, the air option added somewhere in the 30–55% range to the landed cost per unit, depending on weight and consolidation. That is a wide band, and it should be—freight rates move weekly. Get a live quote before you commit.
We paid the premium. The alternative was missing a $60,000 installation contract with a deadline we could not move. Units landed nine days early. Nobody wrote a thank-you note, but nobody had to explain a missed deadline to a customer either.
The inverse is what actually hurts. A supplier who quotes probably three weeks and delivers in six has cost you more than a supplier who quotes five weeks and delivers in five—even if the second one charges more per unit. The cheap option that arrives late is not the cheap option.
Step 6: Read the warranty clause for what it excludes
Warranty length is the number everyone compares. The exclusions are where the money actually is.
Ask specifically about four things:
- Who pays return freight on a failed unit—you or the supplier?
- Is labor for replacement covered, or parts only?
- Is the replacement pro-rated after year one?
- Does installation by a non-certified electrician void it, and how would that be verified in practice?
A five-year warranty with parts-only replacement and buyer-paid freight is worth less than a three-year warranty that covers return shipping. Run the math on your own failure assumptions before the clause ever matters.
Step 7: Run a pilot order you can afford to lose
Twenty to fifty units. Yes, the unit price will be worse than container pricing. That is the point—you are buying information, not inventory.
If the pilot clears, open one unit and look inside. Conformal coating coverage, solder joint quality, whether the heatsink is actually mounted to anything structural. You do not need a lab to spot a unit that was assembled with care versus one that was not.
Then, and only then, talk about annual volume.
What usually goes wrong
A few patterns I see repeatedly, roughly in the order they cause damage.
Accepting industry standard as a specification. It is not a spec. If the supplier cannot give you a number with a tolerance attached, you do not have a specification. You have a hope.
Letting the catalog be the spec sheet. Catalogs are marketing documents with technical decoration. The spec sheet and the compliance file are the real documents.
Skipping the diagram check because you are in a hurry. This is the one people skip most often and the one with the most immediate field cost. Ten minutes at the PO stage versus a week of installer time and a cross-threaded terminal.
Choosing on unit price across a rush timeline. If you are up against a deadline, the cheapest quote is fairly likely to be the worst one. Not because cheap suppliers are dishonest—because the cheapest way to quote a tight deadline is to under-commit on the delivery date and hope. In my experience, that is the failure mode that actually burns you.
Run the first four steps and you will filter out most of the suppliers who would have caused a problem. Run all seven and you will know which ones you can put on a container order without losing sleep.
