LED stoplights not emitting enough heat to melt snow covering the lights | Hacker News Reader

LED stoplights not emitting enough heat to melt snow covering the lights

twincities.com

51 pointskyleblarson14 comments

This was news years ago. Cities were adding heaters to lights to solve the problem. Here's one article from 2009: https://boingboing.net/2009/12/17/led-traffic-lights-d.html

And there's a commercial off-the-shelf solution:

https://products.currentbyge.com/transportation-lighting/led...

(which only uses 30W for the heater, so it's still far more efficient than the old-style ~150W incandescent.

If the Minnesota DOT wasn't aware that this could be a problem, they are incompetent.

It's even a problem with little used incandescent lights... on a street where the light is nearly always green, the red light can still ice/snow over and not have time to melt during brief red intervals in cold weather.

Yeah I was going to say, when hasn't this been a problem? But there are a multitude of solutions

The 30 watt heater should also last the life of the LED bulb, saving the maintenance every year or so to replace the traditional lightbulbs.

For most climates, designing in a temperature sensor to keep the light a little above freezing would pay back in well under a year of power savings.

Update: Did a quick search for LED traffic balls… looks like about 10 watts when on, and I don't see any of them integrating heaters. Market opportunity! Make LED traffic balls but add a tiny micro controller, thermistor, and a heating element, maybe attached to the existing heat sink and you are set. (Yes, attaching a heater to the heatsink sounds counterproductive, but you will be running it when the heatsink is nearly freezing so the LEDs will still have plenty of thermal differential to dump their waste heat.)

I'm not surprised there are no LEDs with heaters. Seems like it would be integrated into the housing/lense.

This is such a clear and obvious solution I'm surprised this is even a story.

It'd be a bit complex, but why not spin the lens with a motor at high speed, like a clear view screen on a ship?

https://en.wikipedia.org/wiki/Clear_view_screen

Actually, maybe it would be simpler than I first thought:

1. Mount the lens using v-shaped rollers on the edges (three would be sufficient)

2. Along the edge of the lens, attach permanent magnets - a ring of magnets with alternating polarity, NSNSNS...

3. Coils along the edges (might even be able to do it with printed circuit coils) energized, interacting with the magnets to spin the lens.

Essentially, you're making the lens operate as the rotor of a high-speed brushless rotor. One could probably trivially implement a prototype of this using hardware store parts over the weekend.

In fact, this seems like such a simple solution, I wouldn't be surprised if there isn't a patent already out there describing it.

If not - consider this submission of mine "prior art" I guess - anybody at YC want to fund a startup? /lol

Makes sense, but wouldn't it be cheaper to stick a motor on the light to vibrate and detach the snow/ice?

In cold weather a car windshield can ice over at 60mph without defroster heat, I don't think a vibrating light housing will stop freezing rain from building up.

There is one thing that is proven to work in a wide variety of situations, is easy to supply, has no moving parts: heat. And there are already solutions out there to provide heat.

Moving parts are typically a very vulnerable point of failure. Traffic lights must have high uptime and require low maintenance.

Why not? It would freeze between the spinning lens and the LED units. It would require maintenance or break down.

You could just use a small heating element to reproduce some of the heat of the old bulb.

Or, use a heated panel. Reposting a link another commenter added:

https://products.currentbyge.com/transportation-lighting/led...

It is a problem on modern cars as well, since most have LED tail lights. They get covered in snow and become invisible.

Reminds me of this: https://xkcd.com/1172/

Reply on news.ycombinator.com