This video is a continuation of an experiment that happened during a live stream on the BigCliveLive channel:-
https://www.youtube.com/channel/UClIzWmVzGPm2zhNT2XZ-Rkw
During the stream I hooked capacitors in series with a cheap sealed 10W LED floodlight that is almost certainly baking its LEDs. This reduces intensity, but multiplies its lifespan significantly.
This time I used a light meter to get an indication of how the intensity changed with current, and got a rough approximation of "lux per watt" at a height of 300mm (12") from the front of the light. This was not an ideal testing scenario and is just for comparison on intensity.
One of the most notable things was how the efficiency dropped as the light heated up when running at full power. LEDs are less efficient at higher currents and when they are hot, so modern lights are not only designed to fail fast for profit reasons, but also perform badly as they do so.
Here are the results.
Full power cold 10.7W, 3010 lux, 281 lux per watt.
Full power hot 10.7W, 2800 lux, 261 lux per watt.
1uf, 3.8W, 1460 lux, 384 lux per watt. (Capacitor voltage drop 62V)
470nf, 1.9W, 840 lux, 442 lux per watt.
330nf, 1.5W, 640 lux, 426 lux per watt.
220nf, 1W, 450 lux, 450 lux per watt.
100nf, 0.5W (estimated), 210 lux, 420 lux per watt. (Capacitor voltage drop 90V)
47nf, 0.2W (estimated), 90 lux, 450 lux per watt.
22nf, 0.1W (estimated), 30 lux, 300 lux per watt.
At the lower levels the light's internal leakage shunt resistor would be affecting efficiency.
Even at reduced intensity the light output from a basic 10W light was very good. Useful for ambient or security lighting. At 2W and below it becomes viable to use it as a simple 24/7 light source.
It would have been nice to open this light, but it is glued shut with heat resistant silicone type adhesive. Previous attempts to open this style of construction have resulted in the glass breaking.
Some lights like this may contain an active buck regulator and will possibly strobe at lower power. But most seem to use the cheap linear regulators.
If you enjoy these videos you can help support the channel with a dollar for coffee, cookies and random gadgets for disassembly at:-
http://www.bigclive.com/coffee.htm
This also keeps the channel independent of YouTube's advertising algorithms allowing it to be a bit more dangerous and naughty.
https://www.youtube.com/channel/UClIzWmVzGPm2zhNT2XZ-Rkw
During the stream I hooked capacitors in series with a cheap sealed 10W LED floodlight that is almost certainly baking its LEDs. This reduces intensity, but multiplies its lifespan significantly.
This time I used a light meter to get an indication of how the intensity changed with current, and got a rough approximation of "lux per watt" at a height of 300mm (12") from the front of the light. This was not an ideal testing scenario and is just for comparison on intensity.
One of the most notable things was how the efficiency dropped as the light heated up when running at full power. LEDs are less efficient at higher currents and when they are hot, so modern lights are not only designed to fail fast for profit reasons, but also perform badly as they do so.
Here are the results.
Full power cold 10.7W, 3010 lux, 281 lux per watt.
Full power hot 10.7W, 2800 lux, 261 lux per watt.
1uf, 3.8W, 1460 lux, 384 lux per watt. (Capacitor voltage drop 62V)
470nf, 1.9W, 840 lux, 442 lux per watt.
330nf, 1.5W, 640 lux, 426 lux per watt.
220nf, 1W, 450 lux, 450 lux per watt.
100nf, 0.5W (estimated), 210 lux, 420 lux per watt. (Capacitor voltage drop 90V)
47nf, 0.2W (estimated), 90 lux, 450 lux per watt.
22nf, 0.1W (estimated), 30 lux, 300 lux per watt.
At the lower levels the light's internal leakage shunt resistor would be affecting efficiency.
Even at reduced intensity the light output from a basic 10W light was very good. Useful for ambient or security lighting. At 2W and below it becomes viable to use it as a simple 24/7 light source.
It would have been nice to open this light, but it is glued shut with heat resistant silicone type adhesive. Previous attempts to open this style of construction have resulted in the glass breaking.
Some lights like this may contain an active buck regulator and will possibly strobe at lower power. But most seem to use the cheap linear regulators.
If you enjoy these videos you can help support the channel with a dollar for coffee, cookies and random gadgets for disassembly at:-
http://www.bigclive.com/coffee.htm
This also keeps the channel independent of YouTube's advertising algorithms allowing it to be a bit more dangerous and naughty.
Very illuminating – thanks Clive
Anyone know which live stream this featured in? Thank you
would this cost me less on my electricity bill , as I have to put in some 24 hour security lighting. I don't need much light but have to cover a large area and was going to use 3 of these lights. This would cover the area, but be much brighter than I need so a wast of electricity
Lowering the brightens in this way would give the amount of light I need, and be more cost effective
I could murder a dooby.
Can I ask another question? I know this is an old video but I'm curious, if you had a load of these in parallel could you control them with mosfets so they could be variable brightness, perhaps a gradient from low to high and vice versa?
Where do I find those rainbow connectors?
I have a 10W floodlight like that one but it's too bright for where I have it. I would love for you to make a hack like this to lower the output brightness
I’m curious if a ceiling fan run cap will work. The build quality looks more robust and the leads are thicker and insulated. Anybody care to answer? Thanks in advance
Clive, I tried this, first one lit bright, the next few leds less and less the last few not at all. What's going on?
Does anyone know where to get suitable caps? The ones that I can find on Farnell etc. are only rated for 150VAC, is that enough? (For European electricity grid.)
Why don't the manufacturers do this or is it dependant on your supply.
@bigclivedotcom I'm curious, is there a rule of thumb for expected increased lifetime of an LED as a function of the decreased power? Is the relationship inversely linear (e.g. 1/2 power = 2x lifetime)? exponential (1/2 power = 4x lifetime)?