Quark 125B / Sky Technologies

[BrianG]

cadima, you are right; inside the Qaurk there are two thermal pads (about 0.5mm thick) which secures the FETs to the case and also transfers heat to the case. It's not perfect but it works ok. The best connection is direct obviously, but I'm thinking Quark wanted a way to firmly secure the innards to the case (and transfer heat) while allowing them to be removed for any repair work. Thermal epoxy works better by far than these pads, but it considered permanent. No taking it apart after that. Using epoxy would void the warranty.

After a while of track use, one or two people have seen the pads slowly get dusty and eventually seperate from the case. Obviously, this is not good for heat dissipation, but usually happens after the warranty is over. If this happens before the warranty period is up (90 days IIRC), I would send it back to Quark for repair. If out of warranty, you can do the mod if you think you can do it.

On my Revo, I did the mod since it was out of warranty anyway. On my buggy, I didn't touch the insides in case something happens. After a while, I'll probably do the epoxy mod to that one as well.

BTW: If you use normal CPU "goop", which is NOT adhesive, you'll have to find a way to somehow clamp the ESC down so it doesn't move around at all. Any movement will slowly render the goop less effective.

I was planning to have a clamp on the quark, for peace of mind. If I alternatively were to use epoxy how would I get it off the chassis?

[BrianG]

Even though thermal epoxy is considered permanent, you can remove it, but it's quite a bit of work. First, freeze the chassis/ESC in a freezer for about 1/2 hour. Then, gently use a screwdriver to pry up. It should pop loose with a little force. Now this is where the work comes in; use a flat exacto blade to scrape off the epoxy. I would use a fresh one to avoid scratching the surface too much.

I know this works because I've done it twice. It works good on something rigid like the case, but I definitely would NOT try it on anything directly attached to the FETs. Chances are the solder joints on the FET pins (or the pins themselves) would break before the bond would.

Originally Posted by BrianG

Even though thermal epoxy is considered permanent, you can remove it, but it's quite a bit of work. First, freeze the chassis/ESC in a freezer for about 1/2 hour. Then, gently use a screwdriver to pry up. It should pop loose with a little force. Now this is where the work comes in; use a flat exacto blade to scrape off the epoxy. I would use a fresh one to avoid scratching the surface too much.

I know this works because I've done it twice. It works good on something rigid like the case, but I definitely would NOT try it on anything directly attached to the FETs. Chances are the solder joints on the FET pins (or the pins themselves) would break before the bond would.

Can you shoot me a few leads on some of this conuctive epoxy? I'd like to entertain this idea in detail, adn learn of it's transfer properties in comparison to my silver paste.

[BrianG]

I always recommend Arctic Silver epoxy. They have two types; Arctic Silver Thermal Epoxy and Arctic Alumina.

The thermal epoxy has silver particles in it. While not truly electrically conductive, it can exhibit capacitance. So, you don't want to get this stuff on any electronics, like the legs of the FETs. It works fine for the main case though. I say all this, but I did use this on my FETs - you just have to be VERY careful that you use a very thin layer just enough to adhere but not squish out and get on anything else.

The Arctic Alumina is totally non conductive/capacitive. So, this is the stuff to use on the FETs if you decide to go that route.

BTW: The freezer trick works on the first one, but I'm not sure about the Alumina stuff.

I do know that thermal epoxy does not transfer heat quite as well as a good thermal goop, but if the surfaces using the goop are allowed to move around, this would make the goop worse than the epoxy.

I've never messed around inside of a Quark, but I used to do a lot of work with small DC/DC constant current converters. .5" in diameter. These things generated heat and I used thermal epoxy (arctic alumina) to "pot" the converters. Occassionally, I'd have to get back at the current sense resistors or something. What I did was to bake the assembly in a toaster oven set to about 170F. This makes the epoxy spongy, rather than super rigid, and then you can take it off in chucks with an exacto. Like I said, I've never tried this with a Quark, but the heat does have that effect on Arctic Alumina.

If you are going to try this, I should add that you need to preheat the toaster oven to 170F and let it sit for a while, before sticking the PCB, or whatever, in. Ovens tend to spike the temp way over the set temp just to attain its goal temp. Once it preheats, then it will stay at the set temp.

[glassdoctor]

Good info.... but AAngel is now aka Julia Childs :D

[BrianG]

lol, funny GD. :)

So, heat works for the Alumina, and freezing works for the regular thermal epoxy. That's kinda odd. The heat approach relies on softening the epoxy while freezing it relies on making it brittle.