You know these guys are helping me. They wouldn't let me make a mistake.
And really it is so easy. I just need to get them ready and follow the instruction. What could possibly go wrong?
1 - put the COBs on the drivers with the thermal interface compound
DONE!!!
...these black wire looking things...are what? Oh! Do I have take them out of the package?
View attachment 3389645
I always worry that 'somehow' the TIM by itself will fail to maintain contact when heated. Probably has something to do with the tactile smoothness of the compound when applied (which 'sets' after exposure to enough heat). I also worry, even when using fasteners, that, somehow, the TIM will flow and thin when heated, thereby reducing the previously attained fastener torque, thereby reducing the pressure keeping the thing affixed to the sink... and then, that, somehow, it will 'fall off' and potentially break the circuit while i'm not available to immediately intervene, should a breach occur. I wouldn't want anything arcing due to a failed TIM dropping an emitter and popping a wire, causing a potential high voltage arc, and all the scary stuff included in that scenario.
So i went overboard with kapton tape the first time, which worked very well. And my cobs did indeed slide a bit after they heated... but thanks to my McGyver taping, it was only like 1mm away from where it started. It didn't want to stick to the aluminum very well at first, but after a solid first run, i wasn't worried anymore. (but then i totally popped a weak solder on one of the points while moving it around one day, had to take it down and fix it... but no issues after that)
But yeah, i can confirm (as others have) that TIM is enough to keep it attached to the sink, as long as your wire tension isn't pulling it away... as long as you get a near-perfect spread (thinnest possible amount required for full mating surface coverage). Too little and it'll come off; too much and it'll slide more, and maybe even drip into something it shouldn't.
The problem is one I anticipated. It is a design choice to go with square tube. Round is so much easier to make connections, of course.
i drew up the tangents for flattening a spot on an aluminum round pipe. But, to get a 35mm square, flat cut makes the pipe awfully thin in the middle of the cut or awfully large in diameter.
And it is a lot of material to remove.
Really the best idea I may tackle later, is to create (or buy) small aluminum saddles that bolt onto a water pipe anywhere I want. Those bolts can also attach the drivers on the top of the pipe.
It is just much more fabrication than I can manage right now. But, not too difficult and certainly the most modular approach.
View attachment 3389962
^Nice.
Hmm... square rod, round bore? Probably also not exactly "easy."
Square Tube with closed ends, with threaded openings on each end?
Or... imagine having a tube of enclosed heatsink fins, with water flowing through it. Like if you made the whole thing into a modular waterblock, fins and all. Look at how CPU water blocks are made; imagine doing that in the shape of a pipe, instead. Could maybe even do something like... braids of fanned-out copper wire, inside a water pipe, attached directly to the spot directly behind the emitter contact point. Just throwin' it out there. ^^
