Inside the Quark 125, pics.

Originally Posted by BrianG

Well, yesterday evening I got bored (since the site seemed to be down), so I took apart my Quark to replace the thermal pads with epoxy like others have done. I used the regular AS epoxy though - I just took care none of it landed on the board or any electrical contact.

The pads connecting the slab to the heatsink and the FETs to the slab were about 1/2mm thick, so if you take them off, you can't simply epoxy the existing aluminum slab because then the caps are too high and it won't sit flush. The existing slab is ~2mm so I cut a piece of 3mm aluminum instead. It fit perfectly.

I didn't replace the middle slab though. The two boards come part easily enough if you simply desolder the 4 pins holding them together on the motor lead side. The other connector simply slides apart. However, there are a few surface mount device pretty close to that area and didn't want to take the chance ruining them from excessive heat. And anyway, unless you find a way to couple that piece to the external heatsink, it is just functioning as a heat spreader and epoxying it won't help all that much IMO.

Once I got it all together again, I tried it out and now the heatsink acts different. Before, the heatsink took a while to heat up, got warm, and then took a while to cool off. Now, the heatsink heats quickly, gets hotter, and cools off quicker. This tells me there is less thermal insulation between the FETs and the heatsink so there is a more efficient transfer of heat.

All in all, I'm pretty happy with the results and think it's a worthwhile and fairly easy modification providing you don't mind voiding the warranty. Once the epoxy is on there and dries, it's not coming off.

Looks Like you are getting the same results as me. I'm glad it's working for you too. I really think it makes a difference too. I think if I bought another one of these controllers I would make this change right away. It would be nice if they put them together like this at the factory. We wouldn't be able to pull them apart but, we wouldn't need to then either.

[Serum]

Brian, i enjoyed reading your post.

But a few pictures of the on going progress wouldn't have hurt. LOL

You did nothing to the 'middle' piece of aluminum?

What was it that you glued? not the fets to the housing, right?

[Serum]

Thermal connection is the key to success with controllers.

The warrior controllers are not using the aluminum plate in between. i think that's a lack of design. But not using a good thermal connection isn't too keen either..

[BrianG]

Serum: I thought about doing pictures, but there were already pictures of the internals posted previously. I didn't do anything earthshattering; I just removed the internals, then removed the thermal pad on both sides of the spacer aluminum, and epoxied that to the FETs, then to the case. Since the pads were 0.5mm each and the slab was 2mm, I needed to use a 3mm piece instead. And, you are correct, I didn't do anything to the middle piece. Like I said, unless that one is coupled to the external case, nothing done will have much effect.

Cartwheels: I think there are two reasons why they use the pads: ease of assembly and warranty work. If the FETs were epoxied to the case from the get-go, warranty work would be all but impossible. There are other forms of thermal tape which may work better, but the stuff they used is actually better than most tapes I've seen. Unfortunately, it doesn't stay connected under the dusty and vibrating conditions.

[Serum]

Why don't you use a thicker aluminum plate, glued to the case with thermal stuff and use silicone paste or a thin pad?

[BrianG]

Silicone paste doesn't work well if the surfaces move around at all. Pads are not as effective. If there was a way to make absolutely sure both surfaces stay put under use while pressing the surfaces together firmly, then regular thermal paste would be an ideal choise. Since that isn't easily feasible using the stock case, thermal tape would be the next choice, but it's not as effective and still prone to the same problems as the original tape. So, that leaves thermal epoxy. The only downside to that is that you can't seperate them again. In the end, it's all about compromises.

[smhertzog]

Test drove my quark with the angle heat sinks today. It was just over 80 outside. I ran it for just over 12 minutes, none stop and as hard as I could around my track. The Quark was 139 on the hottest part of the angle and 144 were the motor wires go into the board. The motor was 172 so I didnt pop in another set of batteries. I was running a 7xl orion 4800 lipos and 20 66 gears with my buggy diffs so I went to 22 66 to see if it would lower the motor temp. After about 10 laps I flipped it going over a jump and decided to take some temps. The motor was 181 and the angle on the Quark was 151and the wire ends were 150. Looks like I need a better motor.

thinking about it if i was quark i would epoxy heat spreaders to the fets and then drill tap a thread in those plates probably do that first though then use a paste and simply screw the heat speader to the esc case to apply presure for the paste to work.

You could even use a bit of angled alu like smhertzog and bolt that to the side of the case in the same manner with it epoxied to the fets.

This way you can still take the unit apart for maintenance.

[neweuser]

How are you guys getting the Quark apart anyways?

[smhertzog]

It came off on its own were it was taped to the case. It was just flopping around in the case. The heat spreader was on good and took a little finesse. It runs much cooler than it did with a fan and heat sink on the case. I just wish it had an adjustment for the motor timing so I could get my motor temps down. Its always something isnt it?
If the punch control and the brakes had two more lower settings it would be nice too. All in all it is a very nice controller it just needs a little help in the thermal area when your running high amps or high volts on a 10+ truck.
I'm alittle curious about the new Mamba Maxx. I have a brand new in the box Quark and a Quark with like two runs on it. I'm going to Ebay one of them to get some cash to buy a Mamba and check it out. I'm running 4s lipo and a two pole motor which is the only short coming of the Mamba maxx (low volt rating and no 4 pole motors). Its cheeper than buying a Neu motor and I love my Mamba 25.

Originally Posted by neweuser

How are you guys getting the Quark apart anyways?

If you are talking about how to open the case, in the center on of the top plastic piece between the plastic and aluminum casing, stick a small sharp flat sharp screwdriver, opening the aluminum casing just enough for the top plastic piece (there is even a small slot there like it was made for a screwdriver to fit) to come outward.

I'm alittle curious about the new Mamba Maxx. I have a brand new in the box Quark and a Quark with like two runs on it. I'm going to Ebay one of them to get some cash to buy a Mamba and check it out. I'm running 4s lipo and a two pole motor which is the only short coming of the Mamba maxx (low volt rating and no 4 pole motors). Its cheeper than buying a Neu motor and I love my Mamba 25.

I love the Mamba Max, but the Quark I believe just feels better overall for 1/8. When I had the MM in my buggy over all it works pretty good, it just feels smaller. Not quite as much take off power. Occasionally, from a standstill to full throttle would give you just some clicking as if I was holding the buggy. Hard to explain. If you have a light 1/8 setup it would probably be ideal.

just want to point out that as far as i know mamba max controllers dont have a problem with 4 pole motors.
Glassdoctor was running a neu motor on one in his 1/8th buggy conversion and he says it works great i am not sure what settings he uses but i suspect that you need to set the timing settings higher as you normally do.

Yes, maybe I do. I had it set on low timing. I have a secondary 1/8 buggy that I'm now running the MM in. It is an old MBX. It only has enough room for one 4s lipo. I have it setup with the MM and a Neu 1512. It's nice and light and seems like it is very efficient, I think it ran for somewhere close to 15 minutes. I'm going to try the normal timing setting on the MM to see if there is any change, it really is an amazing little controller.