If you're sourcing LED drivers for a commercial retrofit or a new build, you know the feeling. The fixtures arrive. The driver is in there, sealed, and the spec sheet looks fine. But you don't actually know if it will last, or if it will flicker, or if it's actually delivering the current it's supposed to. I review roughly 200+ unique lighting items annually for quality compliance, and in our Q1 2024 audit, we caught three different driver defects that would have caused failures within six months. This is the checklist I use. It's not theoretical. It's what we do on the bench before signing off on a batch.
This guide assumes you have a basic multimeter and a few minutes per driver. If you don't have a bench setup, skip straight to step 4—that's the one most people miss.
When to Use This Checklist
Use this before you accept a large shipment of LED drivers, especially if they are from a new supplier or a new model. We use it for any order over 50 units, and for every new driver model that enters our inventory. If a project has a critical uptime requirement—like a parking lot at a hospital or the lighting in a data center—I'd test every single driver in that batch.
There are five steps. Each one is pass/fail. If a driver fails any step, we flag it. If more than 5% of a batch fails any step, we reject the whole batch and request a replacement.
Step 1: The Visual Inspection
Start with the obvious. Look at the driver casing. Are there any dents or cracks? Are the wire leads stripped and tinned properly? Check the input and output labels—do they match the order spec? I once rejected a batch of 200 drivers from a vendor because the label said '120VAC input,' but the spec on the purchase order explicitly required '120-277VAC.' The vendor had run out of the right labels. A lesson learned the hard way: trust, but verify.
Check for:
- Physical damage to the housing
- Loose or poorly crimped wire connections
- Label legibility and accuracy (voltage, current, and model number)
- Date code. If the driver was manufactured more than two years ago, be cautious. Electrolytic capacitors age, even on the shelf.
This step takes about 30 seconds. It's not sophisticated, but it catches about 15% of the defects we see.
Step 2: Input Current and Power Factor (The 'Does It Even Work?' Test)
Plug the driver into a line power source. Use your multimeter to measure the input current. Compare it to the spec. If the spec says '0.5A at 120VAC' and you're reading 0.7A, something is wrong. A high input current could mean a short, a failing rectifier, or just a poorly designed driver.
Most buyers focus on the output—lumens, watts—and completely miss the input side. The question everyone asks is, 'Is it bright enough?' The question they should ask is, 'Is the driver efficiently converting line power into LED power?' A driver that draws more than spec is wasting energy and generating excess heat.
What I look for:
- Input current within ±10% of spec
- Power factor (if you have a power meter) above 0.90 for commercial drivers. We rejected a driver from a low-tier vendor because it had a power factor of 0.78 at full load. The seller claimed it was 'within industry standard.' I've been doing this for years. Industry standard for commercial use is 0.90+. That driver would have added to our building's reactive power charges and generated heat in the distribution system.
This step takes about 2 minutes per driver.
Step 3: Output Voltage and Current (The 'Is It Set Right?' Test)
This is the core of the test. Connect a suitable load—either an LED array that matches the driver's specified output or a set of high-power resistors. For a standard 350mA driver that drives a 36V array, use a load that simulates 36V at 350mA. Don't run it without a load. It can damage the driver.
Measure the output current. This is the most critical measurement. LED lifespan is inversely proportional to drive current. A 50mA overdrive on a 350mA driver can halve the life of the LEDs. This isn't a guess. In 2022, I ran a blind test with our team: the same LED board with a correctly set driver versus a driver that was 15% over-spec. 80% of the team identified the over-driven board as 'brighter,' but within 2,000 hours, the over-driven board had a noticeable 15% drop in output.
The tolerance we accept:
- Current: ±5% of spec. Anything above that is a reject.
- Voltage: ±10% of spec. Voltage has more tolerance because the LEDs themselves have a forward voltage variance.
In one batch of 100 drivers, we found three that were outputting 410mA instead of 350mA. That cost the vendor a redo and delayed our launch by a week. But it saved us from having to replace the LED arrays in five years.
Step 4: The 'Unlit Flicker' Test (The One Everyone Forgets)
Here's the step most people miss. You've tested the driver with the load and it looks fine. But now, disconnect the load. Just the driver, powered on, with the output leads shorted or open (check the driver spec—some require a load). If you have an oscilloscope, look at the output waveform. What you're looking for is a clean DC signal with very low ripple.
Why this matters: A driver with high output ripple can cause a visible flicker in the LEDs, even if the average current is correct. This is especially true for drivers that use a low-quality capacitor in the output filter. The 'flicker' is not always visible to the naked eye, but it is to a camera. In a warehouse, you won't see it. In a video production studio or a showroom with stroboscopic lighting? It's a disaster. Faster, stronger. This is about consistency in the DC output.
I'm not 100% sure of the exact threshold, but in my experience, a ripple of more than 5% of the average output current is a problem. I'll flag any driver where the ripple is visually 'noisy' on the scope.
What we do: If a driver fails this test, we check the capacitor rating. If it's a cheap, generic capacitor rated for 85°C instead of 105°C, we reject the batch. The driver will probably work for a year. Then capacitors dry out, ripple increases, and you get flicker. A lot of drivers—probably the majority—fail because of this. Why do drivers fail? Because of heat and cheap capacitors.
Step 5: The 'Hot Start' Test (Simulating Real-World Conditions)
Unplug the driver. Let it sit for 60 seconds. Plug it back in. Does it start cleanly, or does the output flash? This is a 'hot start' or 'power-on' test. A good driver should power on to its steady-state current within 500 milliseconds. A driver that flashes indicates a poor inrush current control circuit. This can cause dimming protocols to fail, and it stresses the LED array on every power cycle.
I learned this one the hard way in 2020. We were testing 100 drivers for a new series of recessed lights. It looked clean. After the visual and electrical tests, they all passed. Then I did a power-cycle test: on, off, on, off, twenty times. Five of the drivers flashed on the third power cycle. It was a bad batch of start-up capacitors.
Quick check: If you don't have a scope, just use your multimeter in 'MAX' mode on the output. Set it to measure the maximum current. Power on the driver. If the max current is more than 200% of the spec, you have a flash issue.
Before You Reject a Batch
Not ideal, but workable: if you find a defect, document it with photos and the measurement. Then contact the vendor. A good vendor will replace the batch at their cost. If they push back, you have the data. This is exactly why we include a testing clause in our contracts now.
A final note on cost and value: The cheapest driver that meets spec isn't necessarily the cheapest overall. To be fair, a budget driver might work for 10,000 hours. But a quality driver from a reputable source? 50,000 hours, no issues, consistent output. For a parking lot light running 12 hours a night, that's about 2 years vs. 11 years. The cost increase per driver might be $5. For a 200-unit order, that's $1,000. The cost of replacing 200 drivers in the field? In my experience, figure $50 to $100 per fixture, including the truck roll and labor. Do the math.
This isn't about being perfect. It's about being consistent. Use this checklist, document the results, and you'll catch the bad drivers before they end up on your ceiling. That's the whole point.