A Peek into Ryan's Builds: Heatsinking a Hot Linear Driver

[jimdt7]

Very nice build man :beer:
What lens did you use ?

Jim

 

[tsteele93]

Every material has a Thermal Transfer Coefficient that determines how quickly heat moves through it - copper (398) is almost twice as good aluminum (237) for this.

Aka thermal conductivity.

See the second chart and note that the alloy is VERY important.

Thermal Properties of Metals, Conductivity, Thermal Expansion, Specific Heat - Engineers Edge

Aluminum can run as low as 120 W/m*C at 795 Specific Heat J/kg-C up to 220 W/m*C at 900 Specific Heat J/kg-C.

Copper is also very alloy dependent with C11000 at 388 W/m*C and 385.0 Specific Heat J/kg-C.

The molar heat capacity of copper and aluminum is the same but copper has 3 times the specific heat capacity of aluminum which means for each gram, it takes 0.3 Joules to raise the temperature by 1C - for aluminum it takes 0.9 Joules (approx)

I think you are saying that backwards. Aluminum has about 2.3x the specific heat of Copper.

specific heat

In reality, Copper and Aluminum both do a very good job for our needs and very likely they are close enough for either one to do just fine. I was glad to see our copper modules come in at C11000 because you really do want a "wick" or "fast moving pipe" up against the diode to move the heat away into the sink.

 

[steveclv]

Yes - how about inverse 3 times LOL

The biggest problem is that for most of us, knowing the alloy for what we can purchase 'over the counter' is very hard. For example we have no idea what alloy is used for the host body. And yes, as you say, it can make a huge difference to the effectiveness.

Also black hosts are much better than shiny ones

 

[tsteele93]

Or, you can make your own molds and pour exactly what you need!

1 lb pound (16 oz) .999 Fine COPPER bullion Shot Chunks for Melt

 

[ped]

@ Ryan

Think of it in the terms of an oven, as the diode heats the main heatsink, the air temp is going t rise inside the cavity, along with the heating of the driver itself, its gonna get warm quick.

In a sealed host, you need MASS , or directly 'sinked to the main body.

Google the term "thermal runaway".

 

[RyanElectro]

Very nice build man :beer:
What lens did you use ?

Jim

G-2

@ Ryan

Think of it in the terms of an oven, as the diode heats the main heatsink, the air temp is going t rise inside the cavity, along with the heating of the driver itself, its gonna get warm quick.

In a sealed host, you need MASS , or directly 'sinked to the main body.

Google the term "thermal runaway".

I know what you guys mean! but the mini copper heatsinks can't be ground down any further or it won't press into the top of the pill! Plus it's better to have more mass on there.

It works very well this way too I'm sure it's already very effective, seeing as this laser runs rock solid stable on my LPM for 3 minutes.

 

[ped]

One of the vets...either rhd , or wannabu...err SJLL...JLS..cant remember what he's changed his name too, the one with the muppet charactor burning his hair with a laser, has done a few builds with the driver directly heatsinked to the body of the laser, I'll see if I can find them.

 

[lasersbee]

I couldn't grind the height of the heatsink any lower as it wouldn't press into the pill then.

edit: I like to think of them as "copper pipes"... that transfer heat as they get cooled moving heat into the pill. No airflow, yes, but still surrounding air & compound helps dissipation of heat nonetheless.

The surrounding HOT non moving air acts like an insulator.
And if you increased the number of "Copper pipes" (solid
copper) the heat would transfer more efficiently (faster)..


Jerry

 

[BShanahan14rulz]

I think there is a misunderstanding here. He used those heatsinks not because he thought that the pins would increase performance, but because they were copper colored and thus might hold more heat.

 

[RyanElectro]

The surrounding HOT non moving air acts like an insulator.
And if you increased the number of "Copper pipes" (solid
copper) the heat would transfer more efficiently (faster)..


Jerry

Yeah, it's not possible to get it solid unless I grind it all the way down which is pointless since I'll just be taking away more mass and also it won't press into the pill nicely

But if I do find some solid copper sinks, I'll definitely use them. Even still, I'll probably just fill the gaps with solder in the future.

I think there is a misunderstanding here. He used those heatsinks not because he thought that the pins would increase performance, but because they were copper colored and thus might hold more heat.

Yes, correct, also to transfer heat faster. But I'm quite sure they are real copper, not just copper-plated as I think you're implying. At least it's for sure not brass, as it's polished and doesn't look anywhere near golden. I suspect it is a high-percentage copper alloy.

The performance and stability from this setup speaks for itself though. I could probably further the performance like some have suggested by filling the gaps with solder. I'll try to find some silver-bearing solder to do the job

 

[Stix62]

..... unless I grind it all the way down which is pointless since I'll just be taking away more mass and also it won't press into the pill nicely

What about if you grind them all the way down and then, with a very thin layer, glue 2 back to back?

 

[RyanElectro]

What about if you grind them all the way down and then, with a very thin layer, glue 2 back to back?

Haha that's plausible but it will require lots of grinding and would be pretty wasteful won't it?

Filling the gaps with silver-bearing solder will probably be just as good

 

[Stix62]

Haha that's plausible but it will require lots of grinding and would be pretty wasteful won't it?

Filling the gaps with silver-bearing solder will probably be just as good

I only asked because it was something I was thinking of doing with some finned heatsinks I have, and I'd only be doing one or two so the waste doesn't matter.
The solder idea sounds a lot easier, I must admit

 

[justinjja]

IMO, There is no point in heatsinking the driver any more (edit: more than you have already heatsinked it)
Because it can run for 3 mins without the driver kicking into thermal protection mode

More heatsinking on the driver would more heat from the driver to the diode...

 

[RyanElectro]

IMO, There is no point in heatsinking the driver any more
Because it can run for 3 mins without the driver kicking into thermal protection mode

More heatsinking on the driver would more heat from the driver to the diode...

I have to disagree with this. My laser LPM's much more stable when I compared my heavily heatsinked 1.8A vs. my lightly heatsinked 1.6A.

And I really don't think a stock driver w/o heatsinking @ 1.8A can run for 3 min w/o thermal protection kicking in. Some people already have problems running it for longer than 60s before it starts pulsing.

 

[justinjja]

Sorry, what i meant was there is no need to get a solid piece of copper for the driver,
because the finned one is clearly working fine