Q-III heatsink isn't enough.

I accidently ran my Q-III long enough for the body to be VERY warm (it got turned on while in the pouch somehow). Not hot enough to be uncomfortable to hold, but pretty darn warm. It is just a TWOK running on the regular board and with a basic CR123 lithium.

However, I've now noticed some distortion of the phosphor and a couple of bubles under the dome. I estimate that it was running for about 20 minutes or so.

It still lights up, but I'm afraid that the damage is already done.

If this is the case, I guess I'll have to go with a higher fV LED.

I have a Q3 with a TW0J that I run on a 3.7V R123... it gets warm, but I've probably gone through at least 30-40 charges and the light, star, emitter, dome and circuitry all look fine. I mean, if your light appears to have changed, then something must be wrong, but I don't think that if you ran 100 Q3s with primary cells for 20 minutes, all 100 of them would be damaged.

My Q3 runs a TWOK off of r123`s .
Have been on several charges as yet without any damage or loss of light-intensity!
It gets warm on it`s high mode (2-stage switch) but not really hot.
I don`t use a modded heatsink, just stock.

Wolfgang

Hi Orion,

The heatsink is only one part of the equation. It simply allows the heat from the emitter to get passed on to the body of the light. After that, the body of the light must dissipate the heat to somewhere else, usually your hand or the air.

If it was in a nylon or leather pouch, then there is nowhere for the heat to go, so it simply builds up in the heatsink and body. Since it was in the pouch and insulated from the air for so long, I suspect that this is why the heat built up to a point where such noticable damage occurred. While the stock Q3 heatsink may not be the best, it would seem that it is reasonably good enough to transfer the heat to the body in most cases.

pb

Orion said:

I accidently ran my Q-III long enough for the body to be VERY warm (it got turned on while in the pouch somehow). Not hot enough to be uncomfortable to hold, but pretty darn warm.....

Click to expand...

If it wasn't for the LED, it would generate even MORE heat. Why? Luxeon Led's have great light output, but the bulb itself doesn't seem to get hot. I have a 12v Ultrafire with a compressed Xenon bulb. Uses FOUR cr123's. That's 14.8v when you measure the batteries direct.

That was left on in a plastic bag laying on top of papers. The emitter end melted a hole in the plastic, and probably would have melted around the body if it was not sitting in folded paper. I am darn lucky it didn't ignite the papers or my car seat while I was driving. In short, I discovered it when I got home 35 minutes later...the batteries were nearly depleted and it was TOO HOT TO HANDLE. I let it stand a good 30-45 minutes before I could touch it and was STILL very warm.

I wonder what could happen if I had "magic batteries" that run for 12 hours if the darn thing would start to get red hot. You feel the heat coming off that bulb from a couple of inches away....something I have yet to see LED do.

In short, heatsinks on these things aren't meant for long term use. The perfect flashlight would have a Peltier (sp?) heat sink, as used on some pc cpu's. The Pelteir affect is caused by a certain sandwhich of metals. Flow the current through them on direction, the metal natually gets hotter. Reverse poarity and watch that sucker get COOLER. No cooling fluid, no compressor, just metal and electricity. It's frickken great.

Hrmmmm, a Peltier on a flashlight... I'm going to have to think about that one.

But, don't Peltiers require their own power? That would diminish runtime even further though I have absolutely no idea by how much.

cratz2 said:

But, don't Peltiers require their own power? That would diminish runtime even further though I have absolutely no idea by how much.

Click to expand...

Yes they do. But consider everything that can be done with computer controlled flashlights to minimize the drain.

Firstly, the heat sink won't be huge, so that is less drain. A fan has to have at least 10x the drain.

A heat sensor and microcircuit would only kick in the cooler when it senses excess heat. If normal operating temp is reached, it can save more power by only pulsing power through. The entire heat sink doesn't have to be peltier. Maybe just a ring or plate between the heat conducting surfaces.

Dang, someone is going to steal this idea.

the problem i see with that is that the peltier also needs somewhere to dump the heat. if you just layer it, it won't do you any good, and will produce even more heat since it isn't 100% efficient. you would need an even larger heatsink to take care of the heat from the peltier. i don't see it as a very prudent thing to put on a flashlight, or even a cpu for most people. it does keep the cpu cooler, but you also need a much beefier fan at least, and usually a watercooled setup is required to run peltiers properly. the electrical drain for peltiers is insane too, it would be similar to doubling the electrical needs of the light, just like it does in a computer.

I wasn't worried about total runtime, I was just thinking what sort of crazy *** light you could build with active cooling like that. Just having a mad-scientist moment.

Perhaps the star wasn't pressed good to the QIII area that it does heatsink. Was there thermal paste? Also I noticed that stars always have rough spots where they are broke/bent off from the rest of the starts. Feel underneath and you'll feel a sharp lip on 2 sides of the star "arms". I grind those smooth when using a star against a HS.

There is the possiblity that these raised parts were the only contact and kept the rest of the star from making thermal contact....

Yeah, I haven't been into the whole overclocking scene for quite some time but I was under the impression that the power demand was considerable. Also, without a backup cooling system of some sort, if any aspect of the Peltier device fails, it acts as a very effective insulator and would likely kill the LED or circuitry in short order.

But, we are an extremist forum so why not? I think it would probably need to be a from-scratch build rahter than modifying an existing light such as a Mag. There could be a second cell in a seperate section of the light for the Peltier, possibly SLA.

Who knows... maybe someone will do it. I had thought of using an active liquid cooling system when I first got into modding and realized how much even a Lux III on a LiON in a small space can generate.

zespectre said:

I wasn't worried about total runtime, I was just thinking what sort of crazy *** light you could build with active cooling like that. Just having a mad-scientist moment.

Click to expand...

Well, my dream would be to slap a Pelteir on one of those 15 million CP spotlights with the 12v lead-acid battery. Then replace the the typical 7amp battery with a serious of 10 amp rechargable D cells of NiMh.

As many of us know, Lead Acid has near continuous voltage drop from the moment you use it. Levels off at 10.5 volts for a while, and begins to bottom out. Better battery chemistries like Lithium and NiMh sustain higher voltages for longer periods of time.
Now THAT would be one slick light. <drool> :wow:

gnef said:

the problem i see with that is that the peltier also needs somewhere to dump the heat. if you just layer it, it won't do you any good, and will produce even more heat since it isn't 100% efficient. you would need an even larger heatsink to take care of the heat from the peltier. i don't see it as a very prudent thing to put on a flashlight, or even a cpu for most people. it does keep the cpu cooler, but you also need a much beefier fan at least, and usually a watercooled setup is required to run peltiers properly. the electrical drain for peltiers is insane too, it would be similar to doubling the electrical needs of the light, just like it does in a computer.

Click to expand...

As I understand it, the Pelteir "metal sandwhich" litterally cools the metal surface. It shouldn't conduct heat as much as cool the other metals it touches. The hot plate will attempt to transfer the heat to the pelteir plate, which in turn will simply blow it off electrically. Convert that heat energy into electricity and maybe we can sustain longer run times. :naughty:

Maybe I just need to re-read up on the Peltier as I may be entirely mistaken. But going from memory, I though it consisted of two metal plates and a ceramic material between them. Somehow, heat was pulled off of the CPU with the first plate, through the ceramic and to the second metal plate. The heat would then be disapated with either a fan, liquid circulation or, I think actually liquid nitrogen in one case.

But this is from a long time ago... But I specifically remember the power being a big issue... I think on a couple of the websites, they committed a second powerline to the power supply for the Peltier for both secondary heat purposes and to allow for plenty of drives to be powered by the main power supply.

Cratz,

You are entirely correct. A peltier junction does not bleed off heat electrically. Instead, it simply moves the heat very quickly from one side of the junction to the other.

Peltier junctions also take a significant amount of power to do their job. Furthermore, since the other side of the peltier junction gets pretty hot pretty quick, you need significant heatsinking to dissipate that heat into the surrounding environment.

In fact, you can use a Peltier device in reverse.. That is, use it to heat an object as opposed to cool it. You do this by either flipping the device over or reversing the current flow through it.

Now, consider a module which has the current estimated maximum theoretical efficiency ratio of 1:1, which means 1 watt of heat transfer for 1 watt of power applied to the device. Using a peltier to cool a 3W Luxeon, that would translate into 6 watts of power needed to run both the Luxeon and the Peltier. Plus the addition of a larger, more efficient heatsink.

You're going to need a bigger battery.

pb

tron3 said:

Yes they do. But consider everything that can be done with computer controlled flashlights to minimize the drain.

Firstly, the heat sink won't be huge, so that is less drain. A fan has to have at least 10x the drain.

A heat sensor and microcircuit would only kick in the cooler when it senses excess heat. If normal operating temp is reached, it can save more power by only pulsing power through. The entire heat sink doesn't have to be peltier. Maybe just a ring or plate between the heat conducting surfaces.

Dang, someone is going to steal this idea.

Click to expand...

The laws of physics state that you can't win and you can't break even.
More heat will be generated with a peltier (look at the cooling fins on them) and the light it is cooling than with just the light. Even if this were not true the peltier uses a lot of energy.

The term "sink" is a very good description of what a heat sink does. You can pour large amounts of water/heat into a sink but unless the drain hole is the same size as the input sooneer or later the thing will overflow and stop getting rid of the heat/water properly. In a real sink you design for maximum normal volume and make the exit hole as big as needed. In a heatsink it's surface area that serves the purpose of the hole.

pbarrette said:

In fact, you can use a Peltier device in reverse.. That is, use it to heat an object as opposed to cool it. You do this by either flipping the device over or reversing the current flow through it.
pb

Click to expand...

That information lead me to my "invention". Presented to the right people it is interesting to see their reaction.

Casually mention that you are working on reversing the polarity of a microwave machine so that it will *rapidly* freeze hot things. If that works you can then say that the meals will be still be hot when thawed.

Nobody's bought the latter idea yet but the reverse polarity sets some wheels turning.

Anyway, even though the LED still technically worked, I put a new LED in the QIII, one of a slightly warmer color. I also put a 20mm reflector in it. It's good to go again!

since it was 'ok' to mention far out things, the peilter idea was interesting perhaps just as far out some mentioned but quickly disregarded water cooling

i am not totally convience this is all that bad, perhaps not limited to just water, but i'd be some kind of simple convetion might go on in a closed liquid raditor system w/nozzels creating a high/low pressure 'naturally' a light the size of 2 123 batteries like my L2 for ex. would certainly dissapate enough heat not to make the handle more than a bit warm (depending on total watage ) but w/o doing the math i am sure 2 5w leds would be easy to cool

the cost and difficulity of manufacture would be the downside but bet it would work super maybe a small company custom making a few for a gb might work

just my crazy thoughts to the mix