InfinitySink Development Thread - 2020 Community Influenced Universal Heatsink

the pg 7 is 3-6.5mm wire, chillin

1 Like

My battery box will have my esc in it too so no preference for me

3 Likes

Any updates?

1 Like

There are updates in my vendor thread. Plates are done, I am just waiting on them to arrive. They will be here shortly.

5 Likes

For the life of me, I cannot find your vendor thread.
Sorry man.

1 Like


here u go

4 Likes

What testing do we want done?

@surfnacho was kind enough to send over one of the InfinitySink V1’s with its mounting plate and hardware and I’m going to test how the temperature of the ESC(s) would be affected by how the sink is used.

Here were my initial thoughts. I’d love to hear what yours are!

  • Power resistors will be used as stand-ins for ESCs. They can easily be adjusted to different temperatures and held there for hours without having motors screaming at me if I used actual ESCs.

  • A power supply will be used to bring the resistors up to 70°C when mounted only to the plate in still air. This is the “standard” the other setups will be compared to (using the same power supply settings each time).

  • Test #1: Record the temp when the plate is mounted to the sink w/o thermal compound. The sink will be horizontal in still air.

  • Test #2: Record the temp when using thermal compound between the plate and sink.

  • Test #3: Record the temp with the sink completely within a small enclosure and no air movement.

  • Test #4: Record the temp with the fins sticking out of the enclosure and a moderate amount of air flowing past the sink using a fan.

Any other tests that would be useful?

11 Likes

I’m personally questioning a mental experiment that you could help me out with.

The way my esc is mounted into the enclosure would have the infinity sink mounted like red below.

That area isn’t getting the convection that the top of the box is. If my esc was mounted parallel to the flow of air, (green) how much more heat would be removed?

Second part, what if the red, perpendicular-to-air orientation was used but on the front of the deck where it would get hit by air straight on? How would that compare to a perpendicular exposed layout.

2 Likes

The front or top of the box would be better but I can’t say by how much or if it’s even necessary to keep your ESC temps down. I haven’t tested the InfinitySink yet and just don’t know how well it can shed heat. It’s surface area is low compared to sinks with longer fins but does that make a 3°C difference or a 30°C difference? I don’t know.

So much depends on how hot your ESCs are inside and how well (thermally) they are connected to the sink. Assuming a decent thermal connection just a little air on the sink can do a lot versus still air.

You’ll probably have to test your setup with it temporarily set up in different positions. My guess is that front mounting will be a little better than top. There’s a lot of air swirling behind the box and it might not perform much worse there though.

Just too many unknowns now to say which is best.

2 Likes

I have a completely different use case but maybe still worth mentioning.

As you might know, I’m still running 4.12s in a top mounted box above the rear axle. It gets directly cooled by the fresh air while riding.

Brian @Creavenger is running neoboxes in their case mounted similar to your drawing. Although i don’t remember if he had them in the front, side or rear of his top mounted box. (also with a cutout i think…?)

We were getting roughly the same temps on a quite hot day with me probably weighing 30kg more.

Brian:

Mine

4 Likes

Yeah I run mine on the side now, but only a small 40x40mm heatsink is exposed, which is better than nothing and It does catch airflow being on the side which is good. I think since I installed the heatsink it runs 3-4 degrees celsius cooler.

Before I ran my vescs on the back of a 3dprinted enclosure. While in that orientation the airflow wasnt great it did still do a good job to keep the vescs cool. I remember when I switched to the fully closed enclosure and noticed higher temps.

Marcel has them fully exposed by having a slit which lets air in which must help a lot better with cooling for sure.

2 Likes

As air flowing over the top and side of the box It creates a low pressure zone behind it which creates eddies. I imagine the swirling air pattern would still be effective at removing heat from the plate

2 Likes

does anyone have the correct vesc6 mounting pattern? The trampa one is not correct. I would like to fix the files and get an order in, so I can do a nice black friday sale for yall.

2 Likes

I can measure for you tonight.

@Battery_Mooch
Solid set of tests. Are you using thermal compound between the resistor and plate?

I’d also argue with the size of the resistors. I feel that much larger area would be more representative of the esc housing attached to the plate. Although that is probably not going to hugely affect things. Maybe doing 4 resistors total would be slightly better. Probably won’t affect final temps, but i do think hysteresis will be different

1 Like

Yes, thermal compound or epoxy will be used for the resistors. But that won’t matter as I am only going by temperature changes from an initial setup based on heat sink orientation and air flow. It will make it easier to heat everything up though and spread the heat more evenly.

I will probably find the power level that brings the resistors up to a steady 70°C case temp with the sink mounted horizontally in still air. Then I will hold that power level steady and try different orientations and moving air to see what affect that has on the resistor temp.

I don’t think using four resistors will make for much of a change. On the ESC the heat is all localized to the row of MOSFETs, making for two strip sources of heat for two ESCs. The ESC does have a heat spreader plate but I don’t think adding another plate between the resistors and the mounting plate will have that much of an effect for the steady state stuff I’m doing.

If we were trying to model the response times and temps for non-continuous heating (like when riding) then I think you could be right and I would want to use another plate between the resistors and the mounting plate. There are just too many tests to do for this kind of modeling though and I just don’t have the time.

Having said all this, if I can find another piece of 1/8” (or so) aluminum I’ll add it between the resistors and the mounting plate. Even if it doesn’t make much of a difference it doesn’t take much to add it to the setup.

1 Like

Yeah I essentially agree with all that. I think the test is valid. Just make sure your heater vent isn’t blowing at it haha

1 Like

We have determined the ac cycles for our labs down to a few ms now. Trying to do thermal tests is annoying af when the ambient isn’t steady

1 Like

yo your site needs the heatsink holes and dimensions

2 Likes

use this for now:

3 Likes

danky

1 Like