I’ve been needing some excellent lights that I can use for photography and also a work-light. I know work-lights usually don’t need to be great, but I often spray paint in the backyard after dark, and great lighting is a must for that.
I’ve built 2 lights with some of the highest praised LEDs on BLF:
Triple “Nichia NVSL219BT D220 4000K 90+ CRI LED” from MtnElectronics
Quad E21 “NVSWE21AT 5000K sm503-D240-M1-R9080” from Virence
Now, the great test results quoted for these lights are always for low power, say 700ma. I had been assuming since I haven’t seen test results for higher powers, that there wasn’t a noticeable difference. I was wrong. The lights I built perform great when run at low power, but go to heck when I go up to 2 amps.
After building the 219b triple I was split between some COBs on a WTS from a new user “rngwn”, or some E21’s from Virence. I chose the E21’s because the COBs had no reviews. Well after the E21s disapointed me, I wanted to give up. I felt my endeavor was just bleeding money. Eventually I reached out to rngwn and asked if he would be willing to send me a pair in exchange for a honest review. He accepted, and this is that review.
Here is the link to his WTS thread where he is selling them: Budget 5700k CRI 95 100w COBs [Items in Stock]
They shipped from Thailand and arrived on day 20 after placing the order. They came packaged well enough. It was a plastic mailer with the COBs foam-wrapped together inside. Stapling the thin plastic mailer shut seemed risky compared to just taping it, but he said their post office tells them not to use tape.
I mounted the COB on a what I call a “Headlight”. It’s a CPU cooler inside a red mounting ring. This lets me bolt it to things like my custom built light frames. Or even just a tripod. It has a 1/4” threaded hole.
I did a chip test. That’s when you use low power to make all the LEDs on the COB glow and compare the brightnesses. I was pleasantly surprised to see they were all quite even. Only one near the middle was a tiny bit dim. And dim is really no problem. If one was too bright, it would burn out later and kill a whole row of LEDs.
Like with all LEDs, the point you measure the output from matters. For all the readings I put the meter 1 foot in front of the COB and at a slight angle of 10 degrees. With the emitter bare I got these numbers:
From straight-on the CCT goes up about 200 and from a near 90-degree angle it goes down about 200.
That’s a lot higher CCT than the advertised 5700K. rngwn said that the seller actually listed them as 5700-5900K. That was news to me.
So I made a reflector and covered it with D-C-Fix in hopes of pulling warm light from the sides and correcting it.
The CCT was still cool at about 6125K, but the Tint, CRI, and TLCI numbers are excellent and consistent for all power levels! Yay!
—0.25A 26.5v— (6.6 watts)
—0.50A 27.02v— (13.5 watts)
—1.00A 27.75v— (27.8 watts)
—2.00A 28.88v— (57.8 watts)
—3.00A 29.80v— (89.4 watts)
—3.34A 30.10v— (100.5 watts)
Now if we plot that we see that they are very well clustered and neutral along the BBL. Perhaps being up along the Daylight Locus would be better? IDK. But no LED I have tested to date is up any higher than this.
I am very impressed with all the metrics except the CCT. Luckily CCT is one of the few things you can fix. I ordered some filter plastic to drop the CCT from 6125 to about 5625. That is inline with the daylight coming in my windows so there will be no wierdness in my photos. I should also be able to double the plastic and get down to about 5125 too if I like, but that might make a lot of heat.
It’s also very bright, putting up brightness numbers at 1 foot distance that are about half of today’s daylight. That’s pretty good since I am not focusing the light into a beam.
All in all, these things really delivered for me. So last night I placed an order for 6 more of these COBs.
Check out the complete data here:
COB Review Data.zip