2
\$\begingroup\$

I have an electric motorcycle (built two now) that I'm having an overheating issue with. This is a fully from-scratch build (minus the motor, controller and accessories), and I am learning a lot along the way.

Here's my issue: I'm having a hard time mounting the controller (the brain of the bike) right up against the front of the "electrical enclosure", which would ensure a direct contact and potential thermal transfer of heat from the controller to the aluminum shell, allowing air to cool it.

I currently have a 2 cm gap from the heat sink fins of the controller to the front of the enclosure. I do not want to mount this outside the "enclosure" because 1. aesthetics and 2. I want to keep all electronics away from water and dust – it's essentially waterproof. See pics:

Controller in its mounted position with the gap I'm trying to fill:

enter image description here

The heatsink of the controller:

enter image description here

The inside of the "enclosure":

enter image description here

The outside of the "enclosure":

enter image description here

First question: How can I fill this gap on these two builds to help with heat dissipation? A thermal sheet mounted to another heatsink and then another thermal sheet to the enclosure? What do you recommend?

Second question: On my next builds, would you recommend air cooling? Cut two vents high on the box, install a filter (for debris) and vents low down on the box, forcing air in and out, while allowing water a place to drain out. Or can I get away with keeping it sealed and using a heatsink (my preferred method)?

\$\endgroup\$
7
  • 1
    \$\begingroup\$ This is kind of an unanswerable question as you've bounded it -- you'd need to get as much thermal conductivity to the outside world as that thing has now. I don't think there's a practical way to achieve that within the bounds of what you're contemplating. It's clearly made so that you can insert it in a hole in a flat plate with all the fins exposed, and seal it up from behind. I'd do that, to the bike you have now, and just leave the fins sticking out in the air. Your enclosure will be sealed, and with luck your cooling will be adequate. \$\endgroup\$
    – TimWescott
    Sep 12, 2022 at 19:18
  • \$\begingroup\$ Better yet would be to arrange it so the fins are vertical -- but that has obvious disadvantages. Some baffles that tend to direct air across the thing when you're in motion, yet still give you some vertical convection when you're sitting still, may end up helping. \$\endgroup\$
    – TimWescott
    Sep 12, 2022 at 19:19
  • \$\begingroup\$ The case has those molded-in fins for extra surface area so you are going to have a difficult time getting good thermal contact between them and any sort of interface material. Not difficult as in "interesting challenge" but difficult as in "prohibitive". Once you get heat out of the fins, you need to dissipate it out of a flat plate that has - at best - half the surface area of the original case. You might be able to get something going with heat pipes over a much larger surface area, but I think this is more of a learning experience at this point. \$\endgroup\$
    – vir
    Sep 12, 2022 at 19:25
  • 3
    \$\begingroup\$ I arrived at "can't do that" because I'm pretty sure you'd need ribs that'd interdigitate with the existing ribs, and thermally conductive glue, and then an even bigger heatsink on the outside of the enclosure, to make up for the thermal gradient across your hacked-together thermal path. At some point, making the enclosure three inches shorter and cutting a hole in the front starts looking mighty attractive. \$\endgroup\$
    – TimWescott
    Sep 12, 2022 at 19:52
  • \$\begingroup\$ mount it so that the fins are outside the case. seal around it and put a cover over it if you want it to look better. \$\endgroup\$
    – Jasen
    Sep 12, 2022 at 20:02

2 Answers 2

3
\$\begingroup\$

2cm is quite a lot of gap, so you could attach another piece of sheet metal to the enclosure to make up some of the space. You could coat the sheet in thermal paste and bolt or weld it down. After that, a large thermal pad to the enclosure will probably work fine.

Assuming the dissipated thermal power is moderate, you shouldn't need any additional heatsinks - the surface of the metal enclosure has a lot of surface area already, and if the bike is moving then you've got forced airflow already. Power dissipation in the controller will be lower when you're stationary or moving slowly, and higher when you're moving fast, so the airflow scales as the thermal demand does.

Thermal pads are usually sticky (but not adhesive) on both sides, so they're easy to fit on a plate like that. Very large ones are usually expensive, so you might want to buy two or three smaller ones to fill the space.

They come in variable thicknesses. In general you want the smallest thickness possible that'll reliably provide contact, because a thicker pad has lower thermal conductivity. The thermal conductivity is measured in W/mK (watts per meter-kelvin) and a higher number means more efficient heat transfer.

This approach is cheap enough that you can just try it and see what happens. You can evaluate the solution by getting a temperature logger with a PT100 thermocouple probe and sticking the probe under the controller when you mount it. Use the bike for a while and see what the temperature looks like under various conditions. Check the safe operating temperature range in the controller datasheet to see if you're within spec.

There's no way to answer your second question without knowing a lot more about the specifics of the controller's power dissipation and the thermal dissipation properties of the enclosure. You'd need to experiment with this yourself - try the cheapest solution (thermal pad) and measure the results. If it's no good, you might need an alternative approach.

\$\endgroup\$
0
0
\$\begingroup\$

Gap fillers are normally not good thermal conductors. They probably won't solve your problem with overheating. Conducting heat over distances is always a hard thing to do.

  • Thermal conduction is slow and the thicker the material the more resistance you'll have.
  • The bigger the surface area the more heat you can get away.
  • The more airflow, the more heat you get away.

For your case I see the following solutions:

  • Get the heat from the controller to the case with heatpipes (but you still might run into trouble) to the enclosure.
  • Force convection inside your case with a fan to distribute the heat evenly in your box so it can dissapiate over the surface of the whole box (maybe even mount some heatsinks on the inside walls of your box to increase surface area.
  • Get rid of the ribs of the heatsink on your controler and mount it flat against the sheet metal of your enclosure with a good thermal paste.
  • Create a watercooling loop with an external radiator (which are available for bikes with internal combustion motors.) You'd have to find a way to get the heat from the controller into the water loop but that seems possible.

I'd suggest the last solution, because it adds the lowest number of additional thermal interfaces of which each adds thermal resistance.

It still might be not enough and you'll need to increase the surface area that is exposed to outside air.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.