Heat Sink? What Heat?

A couple of days ago I put a 1-watt white LS and 30mm optic on an Elektrolumens heat sink into a Mag 3D, along with a .68 ohm resistor.

Not sure that I wanted this mod to be permanent, I bolted the aluminum star to the heat sink, and used a tiny drop of Castrol Syntec 5W-30 synthetic motor oil to help transfer heat from the star to the heat sink, and another drop between the heat sink and Mag body. Capillary action holds the oil between the two faces of the metal, and it doesn't seem to migrate anywhere else. My first mod beyond overdriving incandescent bulbs. It lights up, and stays lit, which pleases me greatly.

There's about 4.5 volts available from the fairly fresh Eveready alkalines. It draws about 500 mA when running.

Real world nighttime testing: The 30mm is really efficient at harnessing the output of the LS. It has a MUCH narrower beam than the Fraen optic on my Anglelux (which I reviewed elsewhere earlier this month). Trees and other objects that are 40 or 50 yards away are very well-lit. But for lighting the trail immediately in front of your feet, the 30mm's beam is really too intense and too narrow. I think the Fraen or NX05 is more useful for those tasks.

My question is where is the heat? After an hour of continuous running, the Mag body and head are cool, the aluminum heat sink is cool, the aluminum star under the luxeon is cool, and the gold leads that rise up into the emitter itself are cool. It is possible that the resistor is hot, but it's under the heat sink so I can't put my fingertip on it. But I would expect that after an hour some of its heat would travel thru the wire onto where it's soldered onto the LS, and I'd be able to feel it there.

Any ideas?

Thanks.

Something doesn't sound right. Have you checked on the resistor? Is the resistor definately 0.68Ohm? Ohm's law (V=iR) seems to have fallen out the window using your measurements above. Maybe you got a dodgy LS?

Hmmm... strange.

Hmm... Maybe I'm missing something, but I just can't see how a .68ohm resistor can drop current from 3Ds down to 500ma without the resistor producing a ton of heat...

Elektrolumens' Blaster II+ sees around 700ma of current, and I'm almost positive the resistor he uses is at or more than 2ohms (3.3ohms, IIRC)...

One scary thing is, if you don't feel any heat anywhere from the body of the light, and the LS is emitting light and is operating, in the end, something is going to give, and it could be the LS giving way to heat it generates without a way to dissipate it to the body of the light...

The emitter itself not emitting any heat while it's on, and while it's being fed 500ma is just plain wrong... Something's missing...

[ QUOTE ]
FalconFX said:
Hmm... Maybe I'm missing something, but I just can't see how a .68ohm resistor can drop current from 3Ds down to 500ma without the resistor producing a ton of heat...


[/ QUOTE ]

The proper unit of measure here is watts, not tons. I don't see this as much heat. P=(I^2)R which in this case is
(0.5^2)(0.68)=0.17W.
I do agree that something is fishy here. If the LED is really drawing 0.5A there should be an easily sensible temperature rise.

Thanks for the quick replies. I too am baffled.

I'm willing to concede that maybe my experimental motor oil pathway b/t the luxeon's aluminum star and the aluminum heat sink might be an ineffective way of conducting heat (although engines use that oil to conduct heat away from the hottest parts of the engine, like the exhaust valve stems and guides, so I figure it must have some heat conducting capacity).

But that doesn't explain why the luxeon's aluminum star itself isn't darned hot after an hour. I can put my fingertip on it, and it doesn't feel warm. Ditto for the gold leads going up into the LED. If the thing is making heat, it seems to be all staying inside by the die, which it shouldn't be able to handle for an hour.

In contrast, my BB400QL3 gets really warm after maybe 15 mins of operation.

I expected that at least the star and the aluminum heat sink would get like that. But, no.

The resistor is one that Elektrolumens sent me for this project. Its bands are blue, gray, silver and then a gold band at the far right end. That's a .68 ohm, right?

Thanks again for any ideas.

Deleted, erroneous post.

First question, do you have the bin codes for your LS? LS's vary wildly in their voltage drops. I have a Q3 direct driven off 3 C batteries and it draws barely over 500 ma. That emitter replaced one that was drawing only a little more, 700ma, also lame, but not as lame. Primo emitters, with low Vf, would draw way over an amp driven that way.

That there's no significant heat at 500ma in a big flashlight body does not surprise me, although I would expect SOME minimal heat you can feel; it's the same in mine at 500ma. If the heat were not being properly transfered, the LS would definitely be getting hot, the fact that it is not means that your setup with the oil is working well. You might just try unbolting the LS and letting it run at 500ma for 30sec; that should get quite hot to the touch, that would prove that your heatsinking is working.

If it were my mod, I would definitely drop the resistor. That will probably not improve things much, so after that I would get a hold of another LS, and hope for better luck. I think you have a dud, though it might be useable in a 4AA setup...

Kill-O-Zap,

the back of the star said it was an MW1D of Q2J grade.

It may be that the heat is being transferred out so quickly that I don't feel any buildup. I was just expecting the parts closest to the LS to be warmer than parts farther away. But they're all ambient indoors temperature, even after an hour.

Hmm, J is good, I would have expected more current, but in in any case NOT a lot of heat. Do try the free-air experiment, that should convince you that the heat sink is there for a good reason.

I still think 500ma seems low; I would take out the resistor. After all, if it's running so cool, don't you want to push it a little harder? Make the big 3D host pull its own weight /ubbthreads/images/graemlins/smile.gif? It may be that there's enough internal resistance between the switch, and whatever connections you soldered/rigged, whatever, that the resistor is redundant. In fact, it may be worthwhile to double check all the connections, because a high resistance may also mean the connection is unreliable.

[ QUOTE ]
Prometheid said:
Thanks for the quick replies. I too am baffled.

I'm willing to concede that maybe my experimental motor oil pathway b/t the luxeon's aluminum star and the aluminum heat sink might be an ineffective way of conducting heat (although engines use that oil to conduct heat away from the hottest parts of the engine, like the exhaust valve stems and guides, so I figure it must have some heat conducting capacity).

[/ QUOTE ]

Ah, finally something that I know about, cars. /ubbthreads/images/graemlins/blush.gif Engine oil isn't that good a conductor of heat really. Better than nothing, but the only reason it works for cars is because it draws heat by coming in contact with hot parts and then flows away some place where it can disipate the heat. I'd clean it up and try some proper heatsink compound. I use computer CPU grease and it works well enough and is super-cheap.

Yes, thermal grease would definitely be best here, after all that's its entire purpose in life. I know nothing about motor oil, and it in fact may be a terrible heat conductor, but the evidence so far is that the whole LS/oil/heatsink/oil/mag setup is dissipating the heat beyond expectations, definitely better than good enough. I fact, we're looking for MORE heat buildup /ubbthreads/images/graemlins/smile.gif

[ QUOTE ]
marklein said:
Engine oil isn't that good a conductor of heat really. Better than nothing, but the only reason it works for cars is because it draws heat by coming in contact with hot parts and then flows away some place where it can disipate the heat.

[/ QUOTE ]

Right, Marklein, the oil quickly picks up the heat and then sheds it quickly to a cooler area. That's why I had the notion that it made a decent heat-transfer medium. I'm thinking especially of 1990s Suzuki GSXR-series motorcycles that use only engine oil (instead of water) and a radiator to cool their high-performance engines.

Thanks.

My 3C Mag with one of Wayne's stars, sink, RS grease, and a 0.47 resistor draws about 1 - 1.1 amp at the start. Doesn't get terribly hot. A cyan HD in a 6AA double barell direct drive draws about .9 amp from fresh batteries (two sets in parallel) but gets warm fast. The whole sink gets hot. That is about 5/16 thick and of course it is mounted to plastic, not the metal. I think the mag body helps convect the heat away unless left on long term.