Does this still run hot enough to melt snow & ice off the front ? IIRC, the Japanese introduction of LED traffic lights was delayed by having to retrofit heaters...

If this can be adapted for grow-ops...

Does this still run hot enough to melt snow & ice off the front ? IIRC, the Japanese introduction of LED traffic lights was delayed by having to retrofit heaters...

Really depends on the LED's used. I have a flashlight with 7x XML2-T6 LED's and these things can get very very hot, they only use about 19w-22w @ 3.7v-4.2v They are incredibly powerful as well.

Trucks and larger cars that sport these super bright LED headlights are a menance, able to dazzle oncoming drivers the way accidentally leaving highbeams on can, even from behind, into the driver's mirror. They should be banned. No one needs blazing head lights in normal urban driving.

I have a flashlight with 7x XML2-T6 LED's and these things can get very very hot


If a LED light gets hot like that, it's poorly designed. LEDs don't tolerate high ambient temperatures, and start to lose performance and age fast above 50C. If the housing for the light is hot to the touch, the LED inside is even hotter.

That's a problem with many LED products, especially those designed to fit standard sockets, because the body of the lamp can still heat up to 80 C and above, and that kills the LED sooner or later. That's why they don't recommend installing them in recessed cans or other enclosed spaces.

The absolute efficiency of a LED is still actually very poor, on the order of 25% which means 75% of the input power is still going to heat. The trouble is, if the junction inside the diode heats up to 125 C or above, it gets destroyed, and the thermal connection from the tiny junction to the outside is quite poor.

This design belongs into obsolescence management.


Hardly. There's nothing wrong with the edison socket or similiar. All the other alternatives have failed to make ground because of their contrived designs, fragile parts, cost, etc.

It's not the socket that is the problem, but the LEDs and their auxillary components. For example, solid state capacitors exist with lifespan of 100,000 hours at 105 degrees C. They're not used because they cost more, and the factories churning out lightbulbs want to push the manufacturing costs closer to 50 cents than $5 per bulb.

The EM noise out of a properly regulated LED bulb is also neglible. Again, cost is an issue, and manufacturers often just let the LEDs themselves do the rectifying with a simple current limiter in line, which results in lights that strobe - but hey, got to meet that $3 price point somehow.

LEDs will become better over time, but currently there's no incentive to improve since the competition was oulawed.

In the end, it doesn't matter if the LED bulbs only last for a year as long as you can buy a pack of 10 for just a couple bucks, like you used to with standard bulbs.

The issue is that they outlawed standard bulbs because they were getting too cheap to make by just about anyone, so there was no longer a profit margin to speak of. Phillips et al. introduced and lobbied the lightbulb bans to make themselves elbow room on the market against competition by introducing artifical complications that the competitors weren't prepared for.

And as the CFL bulbs were a resolute failure hated by just about everyone, there are now no real alternatives to LED lighting, for better or worse.

One of problems is, the heat sensitive components are placed inside this socket above the source of heat. This socket is perfect except it was designed for incandescent lamps before hundred years - not LED or fluorescent ones.


One of the most heat sensitive component, and the one that's actually producing most of the heat, is the LED itself, that cannot be placed anywhere else for obvious reasons. The rest can be made to handle the heat, cost permitting.

No socket by itself can really do away with the heat problem as long as the whole business has to be integrated into the bulb itself, and moving the other components out of the socket to a separate ballast introduces just more complication and more cost to the consumer because of the different requirements of different LEDs.

For example, you could make a current limited DC bus in your house to drive a bunch of bare diodes, but then those diodes must be of a specific type to work with that particular system

So it's no longer a question of, buy a bulb, screw it in, flip the light switch.

You'd actually need two sockets per light, one to house the actual light and the other to house the ballast for that light. If you buy a system now, and the bulbs last ten years, then ten years down the line when you buy a brand spanking new bulb you also need to overhaul the ballast system to match your new bulb.

Then you'd get problems of people fitting new bulbs with old ballasts, or new ballasts with old bulbs, so to combat that issue they'd have to make the sockets incompatible across generations... which by business practicalities becomes incompatibility across different manufacturers(!), so you'd be married with Phillips instead of Osram and vice versa.

That's another reason why the standard screw socket is a nice thing to have.

Or, if you go for a really techy solution, you could imprint each LED with a standard chip which communicates with the power supply over data lines, and tells it exactly how much current and voltage it requires, what sort of dimming is permitted, how hot it is running, etc.

But then you need a whole computer system with network cabling running through your house just to keep the lights on, and that has the problem of all built-in home automation systems: cost, maintenance, management and obsolescence, planned or otherwise.

Meanwhile, you can still buy an old antique desk lamp that you like, and screw in a new bulb, and it just works. There's no questions about whether the lamp and the bulb speak the same digital protocol, or worries about hackers getting into your IoT light switch and burning down the house.