Myth Busted: Thermal Paste Beats Thermal Tape

[notrefined]

While I'm sure it requires significant skill (and developing a skill certainly represents a significant initial investment), it would seem to me that soldering the emitter to the sink has got to be more cost effective than any exotic interface material. So if and when a "basic" thermal paste or tape no longer fulfils the design criteria, my feeling is that saabluster's is the most logical route to improving the thermal path.

 

[AnAppleSnail]

While I'm sure it requires significant skill (and developing a skill certainly represents a significant initial investment), it would seem to me that soldering the emitter to the sink has got to be more cost effective than any exotic interface material. So if and when a "basic" thermal paste or tape no longer fulfils the design criteria, my feeling is that saabluster's is the most logical route to improving the thermal path.

Have you ever soldered an LED to copper? I've done it with thin copper sheets, and it's easy to fail. If I had jigs and timers and wanted to use my toaster oven with my lead-based solder, perhaps I could do it more reliably. My scrap rate using what I have at hand is about 10% of LEDs. It takes planning to do this right. High-temp solder on the wires connected to the LED, then lower-temp solder to mount the LED to your heatsink. Everything in the light has to be able to take the temperature, and you need some way to cool the LED safely (Low-humidity, quickly cool the whole flashlight body). For that matter, how long does it take to get the copper heatsink or whole flashlight to the solder's melt point? Can the LED survive this?

It's harder than it looks. Try it.

 

[The_bad_Frag]

That's true but I'm assuming you mean the foam-type pads? Bondply-100 is a .005 sheet of acrylic and fiberglass, from it's composition I can't see anything that would dry out.



Sounds like interesting stuff I couldn't find any technical performance data. Let me know if you know where to find it. If I used it in my lights I have to rename it to the T-1000

Good to know! I thought that bondply is also made out of that foam-stuff. But now I know better!

But if you really use that liquid metal pad you also need to use that LEDs I linked you. Those are soldered directly to the copper pcb without an insulating layer like on any other star. And at last it needs to be in a copper body. If its worth doing its worth overdoing.

Maybe I find some data sheets. All I know that this works better on cpus than anything else.

 

[fyrstormer]

The fundamental problem with thermal tape is there's a layer of substrate in the way. It might be plastic, it might be metal, it might be some weird kind of super-thin fabric, but it's still there and it still gets in the way. Thermal paste has no substrate to insulate the hot surface from the heatsink. Because of that, it is physically impossible for thermal tape to beat thermal paste.

 

[nbp]

For the sake of not starting a new thread, I'll just ask in this thread as it is about heat management. How long can I run my MC-E Alpha on high without causing a problem? Has anyone run it for an extended period of time? Does the light step down a level if it overheats or anything?

 

[archer6817j]

For the sake of not starting a new thread, I'll just ask in this thread as it is about heat management. How long can I run my MC-E Alpha on high without causing a problem? Has anyone run it for an extended period of time? Does the light step down a level if it overheats or anything?

I think I can take this one I also encourage people to speak from their own experience. As far as I understand, the driver has some sort of passive thermal protection and will reduce the current at some point. This was told to me by someone who should know, but I really have no idea what the mechanism is or how it works. Feel free to address this

I've run one so hot it melted the heat shrink on the positive end of the battery. This was a 30 minute torture test with the light sitting on a table. You should never (ever) let the light get this hot. At this point, extended contact with the head would certainly burn you. Again, I consider this situation abusive and should not be allowed under normal circumstances.

My rule of thumb is, if you can't hold it by the head, then it's too hot. The Alpha should never be run on high and left unattended like I did in the example above. The amount of time that it can run "depends" on a number of factors: ambient temperature, if you are holding it or not, level of charge, etc. Some people opt to move their hand away from the head as it gets hot. This is the wrong approach; keep in contact with as much of the head as possible and your hand will actually cool the light. Too hot to hold? Like I said, shift to medium, turn it off, or throw it in a bucket of water. If you are outside in sub zero temps...you can probably run it until the battery runs out. Same if you happen to be under water. The light will run hotter on a fresh battery because it will draw more current. The light is regulated, but only to the point where the battery Vf (under load) exceeds the LED Vf. After that the current will fall off accordingly. Less current = less heat.

If you want an actual number, I'd say ten minutes on high, but you should really let your hand tell you what is too hot or not.

 

[nbp]

Thanks for the thorough response Jason. Sounds like your rule of thumb is the same as mine: too hot for the thumb = too hot for the light.

I have been using my Alpha at work lately and sometimes it's on high for several minutes as I inspect a holding tank or look for something in a dark warehouse. It gets fairly warm but never too hot to touch with this usage. I just didn't want to assume something about the light's abilities or behavior and then find out I was wrong when I cooked it.