Making Solderable Aluminium Baseplates.

[pixpop]

I'm familiar with metal loaded epoxies, and rejected them a long time ago. I'm not after brute strength and heat tolerance, and it should be obvious from the kinds of stuff this forum is about, what properties I want this stuff for. Mainly, ease of getting stable optical/electronic structures with good heatsinking cheaply and without a lot of drilling and threading work.

Interesting that you rejected the epoxies. Care to share why? I've been using the Arctic Silver epoxy to attach LDs to Al or Cu heatsinks. It seems to work fine, giving a good mechanical and thermal bond. It doesn't conduct electricity though.. is that the problem?

[The_Doctor]

I might look at that stuff. Got any specs for thermal conductivity?

The main reason I rejected metal-loaded epoxies was on the advice of a sales represntative of Marlow (or Melcor, I forget which) who make TEC's. You can use it, but nothing beats a direct atomic bond between metals. No exceptions. One of my designs is to use high density TEC's and the thermal coupling is one of the most critical points to get right, it allows low power to get very good cooling, the efficiency drops fast with any weakness of coupling waste heat out of the TEC. Where arctic silver might help is if a thin copper layer spreads the heat over a wide enough area to compensate for the reduced coupling of arctic silver, but it's likely to demand a greater area than I'll have room for. The TEC's I got came with pre-tinned metallised surfaces too, so I need to make the best of what I've got.

[pixpop]

Here's their website.

http://www.arcticsilver.com/arctic_s...l_adhesive.htm

it says, among other things:

Made with 99.8% pure micronized silver.
62% to 65% silver content by weight.
Superior thermal conductivity. Greater than 7.5 W/mK

The material is quite fluid, so it's easy to get a thin film with it. To use, it's similar to regular 5 minute epoxy, though takes a bit longer to set.

[The_Doctor]

Ok, thanks. I could have looked, I guess, it was that easy.

Still, try regulating that thickness precisely. Not easy. And all that aside, aluminium has a thermal conductivity of 237 W/(m*K), and HTS-2000 is very close to that, though the web site doesn't specify a value. Convinced? To get similar performance with arctic silver, I'd need a copper heatspreader with over 30 times the area, and more if I couldn't insulate it well enough against other sources of heat in the system. More again if I couldn't minimise the epoxy thickness to a molecule or two. To match the performance of a laser the size of a a kitchen box of matches, it would have the bulk, and the internal thermal padding of a smoke machine! My lasers will be small enough to fit two or three in that match box, by volume at least, and contain at least two diodes on TEC's with lenses, drivers, supply regulation and a PBS cube or other means of beam combining.

[frank_1257]

Makes apparently some of the worlds best thermal adhesives and have some of the lowest thermal resistances seen anywhere.

However their best silver based adhesive is somewhat weird to work with it will either come on dry ice in a syringe so after it is mixed they freeze it, and ship it to you.

Another way to buy is in a tub which they suggest to place on of all things a hot dog roller to mix before usage.

This particular model cement has a high temp cure at 100C+ so not everything will be agreeable to this curing.

I learned about this supplier while working with thin disk YVO4 laser material.

It would seem certain that they would have some lower performing well thermally adhesives that would not require as high of curing temperatures.

I have just this past weekend while playing with indium discovered its ability to adhere very well to a glass light bulb used to heat and reform some wire.It is my experience it will not stick to polished glass well, especially if molten indium contacts cold glass. The best bonding occurs with ground or acid etched glass heated beyond indiums melting point.

As yet I have not played with eutectic I believe it is called bismuth alloy indium this is supposed to have melting points at near 100C.

Recently I have discovered laser designs that employ soldered pedestal mountings as mentioned previously in Coherent Lasers.

Another cool way I am seeing to deal with tiny optics is to place a peltier on top of a pedestal and an optic that has been precoated in indium around its perimeter is simply tweaked while current is applied to the peltier to melt the low temp indium alloy. Many peltiers are soldered and assembled with indium so I am sure there is a tight rope act to doing this correctly.


I wish there were pen applicable plating solutions I have seen these for copper that are gold based, something like this would be sweet for aluminum?

[mixedgas]

email the folks at Caswell Plating and see if they can do electroless nickel on AL. You'd strip the oxide with a abrasive, and quickly surface etch with sodium hydroxide and quickly wash it with di water and then submerge it in the plating solution.

http://www.dalmar.net/aluminum_plating.htm

seems much easier then the thermal method to just send the substrate out to a shop that plates AL.

BTW, the Al brazing rods are NOT indium, they are tin with something added to strip the oxide. I have a friend who uses them to make 19" rack cases.

I've seen copper plated al, but it doesnt like temperature extremes, you'd get orange peel at really high temperatures, but your not going that hot.

Steve