MagLED Cooling?

tibim · Oct 11, 2006

Everyone knows about the issue with the magleds regulating down to less output due to heating up. I just purchased a 3D mag w/ magled yesterday and was noticing there is quite a gap between the magled and the aluminium wall where the magled sits in the flashlight. Since air is not a very good thermal conductor I decided to generously cover the magled with thermal paste before putting it back in. Thermal paste/grease is used alot in computers for attaching heatsinks to processors. It fills the gap between two peices of metal and helps transfer heat more effectively. You can buy it at computer stores or even radio shack.

So I've done this and put the light back together. It's been sitting here running for about 25 minutes and it looks *to me* like there has been no drop it output. I took off the head and felt around, the retention device for the bulb(led) is REALLY hot. That is a good sign, heat is being transfered.

I'm not sure if anyone has done this before or if it even helps but maybe someone will try it out and record the results more scientifically. Perhaps it may be possible to add some more metal to the screw on retention peice that holds the LED in, resulting in more surface area and better cooling.

Just some ideas if anyone wants to experiment.

-tibim

Varroa · Oct 11, 2006

Thermal grease tends to flow when it gets hot so if you put too much on it could flow to the bottom of the socket. Also, the thermal grease is not non-conductive (it can conduct electricity) so you COULD possible get short if the grease moves from the sides to the bottom. Just my $0.02 worth but otherwise a great idea.

tibim · Oct 11, 2006

That depends on the type of thermal grease you use. I used silicon based thermal grease which if remember correctly is non-conductive. However I know some of the metalic based thermal greases are conductive as you've pointed out. As far as the thermal grease flowing I'm not sure but I'll run the flashlight standing up for a while and see what happens.

Handlobraesing · Oct 11, 2006

tibim said:
Everyone knows about the issue with the magleds regulating down to less output due to heating up. I just purchased a 3D mag w/ magled yesterday and was noticing there is quite a gap between the magled and the aluminium wall where the magled sits in the flashlight. Since air is not a very good thermal conductor I decided to generously cover the magled with thermal paste before putting it back in. Thermal paste/grease is used alot in computers for attaching heatsinks to processors. It fills the gap between two peices of metal and helps transfer heat more effectively. You can buy it at computer stores or even radio shack.

So I've done this and put the light back together. It's been sitting here running for about 25 minutes and it looks *to me* like there has been no drop it output. I took off the head and felt around, the retention device for the bulb(led) is REALLY hot. That is a good sign, heat is being transfered.

I'm not sure if anyone has done this before or if it even helps but maybe someone will try it out and record the results more scientifically. Perhaps it may be possible to add some more metal to the screw on retention peice that holds the LED in, resulting in more surface area and better cooling.

Just some ideas if anyone wants to experiment.

-tibim

What you did wouldn't really do squat. The retention ring gets really hot stock as well. The bottleneck is the thermal transfer from the socket to body.

tibim · Oct 11, 2006

I was under the impression that the light would drop down to lower output after 15 minutes. I've run it up to 30 minutes now with no noticable drop. Just did this run to test if the thermal grease would flow down into the socket when heated, and it does not.

I'm just experimenting here and posting my ideas. If it really doesn't do squat then no big loss.

I think my next step will be to remove part of the reflector by the retention ring and possibly try to(and I haven't figured out with what method yet) add more material to the retention ring or possibly connect it to the body somehow.

etc · Oct 11, 2006

It is a bad design which cannot be fixed with patches.

Create a whole new head. Go custom. Their LED solution is a patch. An affordable patch, good for the masses but not for everyone. I only got my conversion unit because I got it for $16, had it been more than $25, I wouldn't have gotten it.

PlayboyJoeShmoe · Oct 11, 2006

A patch? Probably.

For the masses? I'm a flashaholic, and I think these are GREAT!

I've used a 3 cell M*gled in an attic for more than 20 minutes and noticed no major drop. Does NOT mean it didn't happen. Just that I didn't notice it.

Al · Oct 11, 2006

PlayboyJoeShmoe said:
A patch? Probably.

For the masses? I'm a flashaholic, and I think these are GREAT!

I've used a 3 cell M*gled in an attic for more than 20 minutes and noticed no major drop. Does NOT mean it didn't happen. Just that I didn't notice it.

Have to agree ... not the BRIGHTEST, but as a utility light that runs "forever"
they are hard to beat.

Mike L. · Oct 11, 2006

I was just going to post about the Magled cooling issue and if anyone had found a cheap workaround. Thanks!

--Mike L.

daveman · Oct 11, 2006

One can always wait till Mag release a heatsinked version of MagLED. Eventually...

Illum · Oct 12, 2006

i just use copper foil.... :ohgeez:

tibim · Oct 12, 2006

Someone with machining capabilities could make a new rentention ring out of steel with heatsink fins around it. The bottom of the reflector can be removed to make room. I wonder how significant the temperature drop at the LED would be with the thermal paste + heatsinked retention ring. It's got to make a difference.

WNG · Oct 12, 2006

I did the same thing to my Dorcy/Craftsman Endurable Lux-1 drop in PR mod.
I used Dow Corning silicone heatsink compound. Non-conductive, thick as heck compared to cpu heatsink compound.
Since this PR is only running with a resistor, I felt it definitely needed all the help it can get in removing heat. I have an old 3D Mag, which looks to have more metal to the bulb carrier. So far, so good.
But I do have some hotlips and emitters to do a proper mod later. The lux-I or III will be overdriven given the extra heatsinking.

evil.genius.todd · Oct 16, 2006

tibim said:
Someone with machining capabilities could make a new retention ring out of steel with heatsink fins around it. The bottom of the reflector can be removed to make room. I wonder how significant the temperature drop at the LED would be with the thermal paste + heatsinked retention ring. It's got to make a difference.

I'd love to see some diagnostic information on this hack?
Do you have access to any testing equipments? temp. sensor thermocouple, DC power supply(other than batteries for some consistency between runs), or no contact IR thermometer. Anecdotal "it looks brighter to me" is encouraging. But nothing makes me more confidence than a nice unarguable analytical test.

Someone else higher up in the thread suggested using copper foil. I wonder if Aluminum foil(since everyone here probably has some) would be an effective gap filler? Sure it's conductive. But it's not like it's going to flow anywhere it shouldn't.

strat1080 · Oct 16, 2006

I was under the impression that the light would drop down to lower output after 15 minutes. I've run it up to 30 minutes now with no noticable drop. Just did this run to test if the thermal grease would flow down into the socket when heated, and it does not.

You would have to measure it with something to tell rather than just making observations. Beam shots at different time intervals might be useful. It actually takes a significant drop or increase in light to actually notice a difference. If you are just making observations you aren't really going to notice a 25% drop in output or something of that nature. Many people can't even tell the difference between a SF P60 and P61 lamp assembly. The human eye just isn't that keen to pick up these differences. They require measurements.

wasBlinded · Oct 16, 2006

The dropoff in output is smooth, starting from when you first turn it on. You will have a hard time noticing without a light meter to tell you. Now, if you compared one that was just turned on, to one that had been on for 15 minutes, I'm pretty sure you would notice the difference.

pilou · Oct 16, 2006

Based on the various graphs I have seen, the dropoff seems fairly progressive. So for the casual user who won't use the light consinuously for more five minutes, close to full brightness will be enjoyed with the LED Mag or the LED drop-in module. I suspect that large lights like the 3D Mag are used for brief periods anyway, so the 50% dropoff within 15 minutes may not be that big of a deal to many people. If you need to carry a light for a long time, a Mag3D may not be the light of choice anyway :laughing: .

Not that reflector ... Has anyone tried some cheap modifications, or not completely cheap ones like the M2 reflectors?

MagLED Cooling?

tibim

Newly Enlightened

Varroa

Newly Enlightened

tibim

Newly Enlightened

Handlobraesing

Banned

tibim

Newly Enlightened

etc

Flashlight Enthusiast

PlayboyJoeShmoe

Flashaholic

Al

Enlightened

Mike L.

Newly Enlightened

daveman

Enlightened

Illum

Flashaholic

tibim

Newly Enlightened

WNG

Enlightened

evil.genius.todd

Newly Enlightened

strat1080

Newly Enlightened

wasBlinded

Flashlight Enthusiast

pilou

Enlightened

Similar threads

Latest posts