How I Vet Commercial LED Lighting Shipments: 7 Checklist Steps Before Sign-Off

How I Vet Commercial LED Lighting Shipments: 7 Checklist Steps Before Sign-Off

If you take delivery of commercial lighting and you're the one who has to sign off before payment, this checklist is for you. Not the operations-manual version with all the rough edges sanded off—the one I actually use, which has caught faulty batches before they reached a customer or a grow room.

I'm the quality and brand compliance manager at a lighting company. Since I rebuilt our verification protocol in 2021, I've personally reviewed more than 200 supplier shipments and rejected roughly 8% of first deliveries. That rejection rate isn't because suppliers are careless. It's because "as specified" and "as delivered" diverge more often than anyone likes to admit.

Seven steps below, in the order I run them. You'll need a power meter, a PAR meter, and about two hours per batch. Skip a step and you're gambling. Borrow tools if you have to—I've done it.

Step 1: Confirm the model and wattage before the box is opened

Start with the shipping label, the box seal, and the model number stamped on the unit itself. I've seen a Spider Farmer G3000 300 W full spectrum box arrive with an older-generation fixture inside. The label said G3000; the driver plate said something else. Everything matched on paper. Nothing matched in the box.

Match three things before you cut the seal: the purchase order, the carton label, and the physical product plate. If any one of those disagrees, stop. Don't open it, don't test it, don't sign. Send it back as a mis-shipment. Once you've opened it, the return argument gets harder.

What I document at this step

• Photo of the carton label with the PO number visible
• Photo of the product plate showing model and rated wattage
• Driver model number—the driver is where the actual wattage lives

I only started doing this after I approved a batch on the carton label alone. The units went out, a customer measured 270 W on a fixture labeled 300 W, and we ate a $780 credit. That was 2022. I haven't skipped the power meter since.

Step 2: Measure actual wattage at the wall

Plug the fixture into a power meter (a Kill-A-Watt or a clamp meter on the line) and run it at 100% for at least 15 minutes. Let it stabilize before you read the number.

For a Spider Farmer SE1000 1000 W full spectrum LED, I expect the wall draw to land within about 5% of the rated figure at full output. If it's 8% or more under, something's wrong—a downsized driver, a different LED bin, a firmware limit, or a genuinely defective unit. Any of those is a rejection in my book.

Note the ambient room temperature when you test. Driver output drops with heat, and a unit tested in a 32°C room will read lower than the same unit in a 22°C room. I record both numbers. Suppliers will argue the temperature, so I make sure they can't.

Step 3: Measure PPFD at the specified distance

PPFD (photosynthetic photon flux density) is the number that actually matters to a grower, and it's the number most easily fudged on a spec sheet. Use a calibrated quantum sensor—the Apogee MQ-500 is what I keep on the cart, though other calibrated units work too. Set the fixture at the distance the spec sheet claims (I test at both 12 and 24 inches), and measure at the center point plus four corners of the coverage area.

Two things I'm looking for:

• Center PPFD within roughly 5-7% of the stated value
• Corner PPFD no more than about 30% below center

The second number catches more problems than the first. A fixture can hit its center spec and still have tired edges—which means the outer plants in a canopy get a different growth pattern than the middle. There's something satisfying about catching that on a first pass, because a grower won't see it until week four of a cycle.

Step 4: Verify the spectrum claim, not just the category

"Full spectrum" is a category, not a specification. Two fixtures can both say full spectrum and deliver very different red-to-blue ratios. If you have access to a spectrometer, run a quick scan and check the R:W ratio against the supplier's claims. If you don't, do a visual comparison against a known-good reference fixture of the same model.

I keep two reference units on a shelf—one from each product line we stock. When a new batch arrives, I run a side-by-side visual against the reference. Any visible shift in tint (especially toward green or away from deep red) is a flag. It's not science; it's triage. Anything that flags goes to the spectrometer for confirmation.

"Everything matched on paper. Nothing matched in the box."

Step 5: Run the Zigbee products through a real pairing test

This is the step most people skip, and it's the one that costs the most when it fails. Plug in the controller or gateway, pair every light in the shipment to it, and then leave the whole set running for 48 hours—or rather, 72 if the shipment is going to a hot-climate customer.

I'm checking three things:

1. Every fixture pairs on the first attempt
2. Mesh connections hold after 48 hours of continuous operation
3. No fixture drops off the mesh without being physically disconnected

We didn't have a formal re-test procedure for returned Zigbee modules until 2023. It cost us when a "refurbished" controller kept dropping mesh connections after about 36 hours. Now every returned unit runs a 72-hour soak test before it goes back in the box. The 36-hour failure mode is the specific reason I test for 48 hours and not overnight.

Step 6: Physical build and thermal check

Run the fixture at full output for a minimum of two hours and feel the driver housing and heat sink by hand (carefully—some surfaces run hot). I'm looking for hot spots, buzzing from the driver, flicker at low dim levels, and anything that sounds like looseness inside the housing.

The heat test runs for 48 hours—or rather, it should be 72 if you're shipping to a hot climate, as I mentioned. I keep a thermal probe on the driver during the first hour. Anything more than about 15°C above ambient at the driver surface is worth a second look.

Also check the physical build: cable strain relief at the entry point, solder joints on the driver board if the housing allows inspection, and the weight of the heat sink. A heat sink that feels noticeably lighter than the reference unit has less aluminum in it. That matters for lifespan, not just for the first week of use.

Step 7: Verify facility lighting placement separately

Grow lights are only part of the shipment. Most commercial grow facilities also take in auxiliary lighting for work areas, loading bays, and the occasional LED chandelier for the office or meeting room. Those fixtures have their own failure points, and they get lumped into the same shipment, so they get checked separately.

For decorative pieces like an LED chandelier, I check dimming consistency across the fixture (no visible banding), color temperature consistency between the fixture and the published spec (Delta E under 3 for brand-critical applications, based on the Pantone matching system guidelines), and the mounting hardware against the installation surface.

For track lighting in a facility—work areas, walkways, propagation rooms—the placement check is the part that catches problems. How to position track lighting in a grow facility is actually a different exercise than in a retail space. You want:

• Heads angled so the beam pattern doesn't spill into the plant canopy—light pollution from auxiliary fixtures is a real problem in photoperiod-sensitive rooms
• A minimum 4-foot offset from any plant canopy edge, measured horizontally
• Dimming controls that go low enough (under 10%) to keep the room usable during dark cycles

I walk the installed track with a light meter and a tape measure. It sounds excessive. It saves a call from a grower at 2 a.m. during week three of flower.

What I check before the sign-off

Final pass before I sign anything:

• All seven steps documented with photos and numbers, not just checked boxes
• Any deviation recorded as a percentage of spec, not just "pass" or "fail"
• The batch sample size—I test 10% of a shipment, minimum five units, plus one unit per pallet
• Driver and LED bin numbers logged against the PO, so future batches can be compared to this one

Common mistakes I've made (and watched others make):

1. Testing only one unit from a batch and calling it a representative sample. It isn't. A single good unit proves a single good unit.
2. Skipping the 48-hour soak because the shipment is urgent. Rush orders are where the process gets skipped, and they're the shipments most likely to fail in the field.
3. Trusting the carton label. The label is a claim. The power meter and the PAR meter are the facts.
4. Forgetting the auxiliary fixtures because they "aren't the main product." The LED chandelier in the meeting room is still a branded product. If it flickers, the customer remembers.
5. Not writing down the numbers. If the next batch is 12% dimmer than this one, you need the previous numbers to prove it. "It felt dimmer" doesn't win a supplier dispute.

None of this is glamorous. The process exists so that the boring work happens before the expensive work does. That's the whole job.