ever feel like you've been being lied to?
So I was poking around for new LED decorative bulbs - specifically, clear bulbs, because the selection there has been a problem for a while, and no frosted bulb looks good in a ceiling fan. Just for example. And I stumbled across something on Amazon, or more specifically, two somethings, both made by the same company originally: LED filament bulbs, in various tiny wattages.
I noticed that the 4w bulb was being claimed as a "40w replacement," and honestly, while I've seen improvements in output per watt, that's less than half the power draw of any legitimate LED or CFL replacement I've seen at the claimed output lumens. So I imagined that really, it's putting out more like 200 lumens or so, and the output is whiter, which makes people think its brighter. The one customer photo made me think the same, but that'd be okay anyway, so I ordered a couple in different bases, to test.
Which, of course, I have done.
Test setup
When I got the bulbs out of their boxes, I immediately felt the lightness - they felt like incandescent bulbs, not LED or CFL. The weight isn't there. And they don't look like LEDs. I mean, seriously, look at this thing.
What the hell.
The electronics are down in the base, such as it is. They're visible from the top. There's no vacuum inside; the globe is to protect the electrics and such.
I'm going to cut straight to the chase: IT'S NOT A LIE. I kept remeasuring output and power consumption levels to revalidate it. They're 2700K as claimed. The 4w bulb actually draws between 3.6 and 3.7 watts; the 2w bulb draws 2.1-2.2. But compare for yourself.
Here's a side-by-side of the outputs. Bulbs in same location, camera set to manual and the same settings (ISO400, f3.5, 1/30th second exposure) in both cases.
Side-by-side
This photo is mostly the incandescent, but with horizontal rectangles of the LED Filament bulb overlaid. A couple of them are obvious, where the LED's output is more obviously brighter; a couple of the others are harder to find.
Stripe Overlays. 3.8w source image, 40.1w source image
The 3.8 watt LED Filmament bulb, in the fixture, and also, a photo of the consumption reading (in watts) on the metre.
And the 40 watt incandescent (actual draw 40.1w) used for comparison purposes, with its load level as well.
As you can see, the spread is wider on the 3.8 watt LED bulb, but I think that's mostly a function of the fixture and relative location of filaments. If I had a way to do true equivalent spread, the 3.8w output photos would be even brighter in comparison to the 40.1w decorative clear bulb.
I kept thinking something had to be fraudulent here, but... I can't find it. I kept checking back on the power metre to see if it had jumped up somehow, just because this is literally twice the best efficiency per watt that I've ever seen.
The bulbs are rated in the 15,000-hour range. One source says 15,000 hours; another says 25,000 hours. I'm going with the lower to be more cautious.
I checked for strobing, too. I have one way to test for that only. According to it, the LED strobes less than the incandescent.
And as far as most places are concerned, this doesn't even seem to exist. I can't buy it anywhere reasonable, I have to order it from obscure sellers online. And they aren't brand new; Amazon reviews go back a while. They're $8 each, cheaper(!) than the frosted and less-efficient LEDs I've been buying, even on a per-lumen sort of basis.
So yeah, genuinely, it's like finding a little alien artefact my mail. Seriously, what the hell?
These aren't dimmable. That's explicit on the packaging. Make them dimmable and the excuses for incandescents are pretty much gone. So we're not there yet - but we're close.
eta: By request: closeup of the filaments, powered, but low exposure for detail. This is actually a different bulb of the same type - this one is 2w (a different but same model drew 2.2w in testing), with output similar to a 25w incandescent.
Also posted to ソ-ラ-バ-ド-のおん;
comments at Dreamwidth.
I noticed that the 4w bulb was being claimed as a "40w replacement," and honestly, while I've seen improvements in output per watt, that's less than half the power draw of any legitimate LED or CFL replacement I've seen at the claimed output lumens. So I imagined that really, it's putting out more like 200 lumens or so, and the output is whiter, which makes people think its brighter. The one customer photo made me think the same, but that'd be okay anyway, so I ordered a couple in different bases, to test.
Which, of course, I have done.
Test setup
When I got the bulbs out of their boxes, I immediately felt the lightness - they felt like incandescent bulbs, not LED or CFL. The weight isn't there. And they don't look like LEDs. I mean, seriously, look at this thing.
What the hell.
The electronics are down in the base, such as it is. They're visible from the top. There's no vacuum inside; the globe is to protect the electrics and such.
I'm going to cut straight to the chase: IT'S NOT A LIE. I kept remeasuring output and power consumption levels to revalidate it. They're 2700K as claimed. The 4w bulb actually draws between 3.6 and 3.7 watts; the 2w bulb draws 2.1-2.2. But compare for yourself.
Here's a side-by-side of the outputs. Bulbs in same location, camera set to manual and the same settings (ISO400, f3.5, 1/30th second exposure) in both cases.
Side-by-side
This photo is mostly the incandescent, but with horizontal rectangles of the LED Filament bulb overlaid. A couple of them are obvious, where the LED's output is more obviously brighter; a couple of the others are harder to find.
Stripe Overlays. 3.8w source image, 40.1w source image
The 3.8 watt LED Filmament bulb, in the fixture, and also, a photo of the consumption reading (in watts) on the metre.
And the 40 watt incandescent (actual draw 40.1w) used for comparison purposes, with its load level as well.
As you can see, the spread is wider on the 3.8 watt LED bulb, but I think that's mostly a function of the fixture and relative location of filaments. If I had a way to do true equivalent spread, the 3.8w output photos would be even brighter in comparison to the 40.1w decorative clear bulb.
I kept thinking something had to be fraudulent here, but... I can't find it. I kept checking back on the power metre to see if it had jumped up somehow, just because this is literally twice the best efficiency per watt that I've ever seen.
The bulbs are rated in the 15,000-hour range. One source says 15,000 hours; another says 25,000 hours. I'm going with the lower to be more cautious.
I checked for strobing, too. I have one way to test for that only. According to it, the LED strobes less than the incandescent.
And as far as most places are concerned, this doesn't even seem to exist. I can't buy it anywhere reasonable, I have to order it from obscure sellers online. And they aren't brand new; Amazon reviews go back a while. They're $8 each, cheaper(!) than the frosted and less-efficient LEDs I've been buying, even on a per-lumen sort of basis.
So yeah, genuinely, it's like finding a little alien artefact my mail. Seriously, what the hell?
These aren't dimmable. That's explicit on the packaging. Make them dimmable and the excuses for incandescents are pretty much gone. So we're not there yet - but we're close.
eta: By request: closeup of the filaments, powered, but low exposure for detail. This is actually a different bulb of the same type - this one is 2w (a different but same model drew 2.2w in testing), with output similar to a 25w incandescent.
Also posted to ソ-ラ-バ-ド-のおん;
comments at Dreamwidth.