Artic alumina as potting compound?

[reefphilic]

I've just received some madmax+ converter board and Lux3 from the sandwich shoppe. /ubbthreads/images/graemlins/grin.gif
Wayne had advised me to use Devcon 2ton epoxy as the potting compound but I already have some artic alumina thermal epoxy laying around. Can I use the artic alumina instead of Devcon? What the purposes of the potting compound? /ubbthreads/images/graemlins/confused.gif /ubbthreads/images/graemlins/blush.gif

 

[EvilLithiumMan]

Somewhere buried deep in this forum (25-30 pages ago), someone stated that a BB750 / MM+ could run hot enough that the solder connections from the emitter board to the convertor board may actually soften enough so that the assembly would 'crush', causing it to lose it's intended design spacing of .3 of an inch. The epoxy is primarily intended to retain this spacing and not for heat dissipation. You could use the AA, but it's not really worth the extra cost.

[Edited}
Well, I've gone through all 38 pages in the forum twice now and can't locate what I thought I had read. I did find one post from Wayne that does mention the possibility of a sandwich collapsing, but it makes no mention of epoxy.

http://www.candlepowerforums.com/ubbthreads/showthreaded.php?Cat=&Board=TheSandwichShoppe&Number=364624&page=18&view=collapsed&sb=5&o=&vc=1

 

[Doug S]

Actually ELM, it was the epoxy that was softening at high temps.

 

[pedalinbob]

though i have never tired it, some have stated that they have used JB Weld (not JB Kwik ) as potting compound.

i think it is rated to 500f.

Bob

 

[dat2zip]

Any epoxy will suffice as long as it can withstand temperatures as high as 200F and preferably remain hard to 160F.

AA and AS adheasive glues tend to get soft just beyond 100F and I would test AA or AS before using it as potting compound.

I recommended 2ton as I have tried this and it has worked fine. There are a zillion other epoxies out there that will probably work fine too.

To test the epoxy before using it, I would create a test coupon filling a small container or cup. The thicker the better, but, it does not really matter. A very small test coupon could be made as long as you control the mixture precisely. This test will need to performed at least twice to get the required information.

Let the coupon cure longer than the recommended cure time (usually 24 hours).

At room temperature test the coupone for rigidity and firmness. Is it rock hard? It should be. Now put the test coupon in an oven and bring it to 140F. Restest for firmness at 140F. If the coupon is no longer hard, and you can can poke it or squish it, then this epoxy fails to meet the hardness test for use with filling modules.

Create a 2nd test coupon. This time after it sets up and gets semi-hard and before it is fully cured, post bake it in the oven at 140F for 15 minutes or until fully cured.

Test while hot. Is it rock hard now? If so, you will need to post bake the epoxy to raise the HDT. If it is rock hard at 140F it is probably safe to use this epoxy for module filling.

The other thing that can be done is to use a high temperature solder when soldering the stick on to the converter board and the emitter board. This will reduce the chance of the solder getting soft when hot.

Many epoxies technical data sheets are not available. If you can find the technical data sheet, find one with a HDT higher than 140F.

Post curing and the effects of post curing can be found with a quick search on Google.

Wayne

 

[Klaus]

[ QUOTE ]
dat2zip said:

snip

AA and AS adheasive glues tend to get soft just beyond 100F and I would test AA or AS before using it as potting compound.

snip

Wayne

[/ QUOTE ]

Hmmh - this is straight from the manufacturers site for AS:

Temperature range: - 40C to >150C

Huh ? /ubbthreads/images/graemlins/confused.gif /ubbthreads/images/graemlins/thinking.gif

Klaus

 

[McGizmo]

Klaus,

An epoxy can continue to function as a thermal bond beyond its HTD. In a discussion I had with a fellow who is a R&D chemist with a company who makes all kinds of urethanes and epooxies for thermal as well as structural applications, he pointed out that as an epoxy softens (reaching HDT), its thermal conductivity suffers as well. Granted, this may be over kill and this guys company makes blends that need to be post cured at high temperatures. In addition to the hardner and resin, they provide a third chemical that can be varied for changing the HDT as well as required post cure temperature. Fortunately our mods don't require this level of sophistication!

I believe Wayne's concern in the use of thae Arctic epoxies here is of a structural concern and not the thermal capacity. I use Arctic Alumina as a potting in my PR builds but I am not relying solely on it for thestructural bonding ofthe PCB to the E-can. I first glue the PCB into the E-can with crazy glue. Between the two, I think the structural requirements are met. I also post cure the AA, presumably raising its HDT.

 

[Klaus]

Don,

I´m sure you and Wayne have much more experience with AS and AA than I will make in my lifetime but the extreme discrepancy between Wayne´s statement and the manufacturers spec really made me wonder. The 100°F out of Wayne´s post is less than 38°C and probably its just a typo.

But what do I know /ubbthreads/images/graemlins/blush.gif

Klaus

 

[dat2zip]

Klaus,

I'm not disputing the specifications as you have listed.

The AA or AS does not specify the HDT (Heat Distortion Point??) as other technical publications for other epoxies are listed.

It's just a point of reference where the epoxy starts to soften or HDT, not where the epoxy fails.

100F or where the epoxy gets soft in no way means it is beyond it's useable range.

I'm only pointing out that some epoxies have a lower HDT point and that the higher the HDT, the better mechanical strength the module will be. Using a low HDT epoxy may result in the module crushing over time due to heat.

I'd hate to see anyone go through all the trouble of building a sandwich and have it collapse some days/weeks/mo's later.

There is no real reason to use AA in the Sandwich module. Cost per oz of epoxy, AA is very expensive. A two tube Devcon 2ton or others you can find at the HW store cost similar to AA but can fill 3-10 modules. That and the fact the epoxy is there more for mechanical strength and not thermal characteristics.

In fact, I think Bondo is good epoxy for temperature. I used to use this a lot when I built model airplanes and if I recall, Bondo didn't show any signs of HDT at all. Maybe, that is because of the filler being used. If you can goop some bondo as the filler it should be fine too.

When it comes to mixing epoxies the ratio is very important to the outcome of the properties of the cured epoxy.

Even 2ton won't cure if the ratio deviates too much.

Epoxies especially when mixed in small quantities are much harder to control the ratio of partA to partB. This resultant effect could weaken or change the HDT point dramatically.

I use a pump base system that dispense a precise partA partB and a combined resultant of 1oz worth of epoxy. The pump system is very accurate and the resultant epoxy cured will be similar from batch to batch. Even then, I post cure all my modules to try to raise the HDT as high as possible. The 1 oz BTW will fill approximately 30 modules and that's the quantity of my builds. I have tailored my module assemblies around this amount of epoxy as a single run to maximize the process.

Wayne

[edited]
Found a glossary term on the web:

Heat Distortion Point The temperature at which a test bar deflects under a given flexural load and a prescribed amount of heat.

http://www.matweb.com/reference/deflection-temperature.asp

 

[Klaus]

Wayne,

how can I question the sandwich masters recommendation

No offense was intended

Best regards

Klaus