Any not-totally-crappy paste will work. That MX-4 is fine.
Typically you’re dealing with small amounts of heat (low heat flux) over an inefficient, almost completely choked-off thermal path with these ESC’s. Lots of surface area typically too.
Different pastes will make almost no difference IMO as long as they’re applied decently.
Make sure the surfaces are flat and there are no burrs. Don’t glop on the paste, you are only filling in the low spots. Any paste between the ESC and the sink just increases the thermal resistance (increases the temperature).
If you need to insulate (from the heat sink) the portion of the ESC you are trying to cool then you will need more than just paste.
I’m also considering removing the DV6 from it’s existing aluminum case, especially if the thermal path is that bad. It’s tough to tell from looking at photos, but I may have to take mine apart to really investigate.
My main issue is the ESC gets quite hot, and all my electronics & batteries are housed inside a single top-mount nanuk case. I’m less concerned about the ESC actually; more concerned about the molicels. The goal is to pipe that ESC heat out of the enclosure rather than cook everything inside.
Which components tend to be the biggest source of heat on these ESCs?
I was going to smooth and diy match lap the surface of the heat sink plate to the ESC’s case to maximize surface flatness and minimize thermal resistance.
@MysticalDork for $10 it’s a fun experiment, haha. But i get your meaning.
The thermal path is fine, that’s not the issue.
The issue is that you have a heat-generating component in an insulated box.
As soon as you give it ANY path for heat to get out of the box, it’ll be totally fine, even if the path to get out is completely terrible by computer heatsink standards.
You gotta keep in mind that a typical desktop computer is moving somewhere between 50 and 250 watts of heat across a ~1.5-2 square-inch contact patch, whereas an ESC, even a dual being pushed hard, is likely generating less than 10-20 watts of heat at most, and you’re transferring it across probably 6-10+ square inches of contact with the heatsink.
The contact patch is not in any way the bottleneck, like it is in a CPU cooling scenario.
I respectfully disagree.
These ESC’s typically take the heat from the epoxy case of the FETs (high thermal resistance) via a thermal pad (high thermal resistance). The only thing saving those FETs…sometimes…is there typically isn’t a lot of heat and/or the low duty cycle for the heating.
But, as you mentioned, the heat flux is low and the contact area is large so the (IMO) choked off thermal path isn’t a huge issue for moderate use of a properly spec’d ESC.
We have no idea how hot those FET junctions are actually getting though. Board temp readings are pretty useless unless significant testing was done to derive the junction temps from the board temps at various power levels.
The MOSFETs.
Removing heat from them via the black epoxy cases, through a thermal pad, is very inefficient though. Not sure what can be done about it for a reasonable amount of effort, or if it’s even worth it.
The VESC 4 was probably the worst design in terms of heat dissipation. The PCB was sandwiched between the mosfets. The only place to add a heatsink was on the outside of the mosfet’s epoxy housing where thermal transfer isn’t great.
Pretty significant temperature change. I think the lesson here is that while the thermal transfer may not be ideal in some designs, airflow is critical in keeping temperatures down.
OK so my 12S3P p42a pack is dying.
I’ve got a fancy new 12S4P p42a pack that @A13XR3 made me.
I was gonna put it in my second board but I think I’m gonna try and cram it in my existing board so I can get using it right now.
Here’s my question: how can I juice up my vesc config to get the max out of this battery?
I kind of got humiliated by a Meepo Hurricane going up a hill yesterday and I can’t let that happen again.
The current settings (provided by Propulsion Boards):
Motor Current Max (Amps): 80.0
Motor Current Max Brake (Amps): -70.0
Battery Current Max (Amps): 40.0
Battery Current Max Regen (Amps): -12.0
Absolute Maximum Current (Amps): 400.0
Maximum Braking Wattage: -1500000.0
Maximum Wattage: 1500000.0
My first guess I just leave the motor numbers as they are (Demon 2.0 has 2 X 6374 190KV Motors) and pump up all the battery numbers by 25% since the new battery has 25% more p-groups.
Assuming the battery’s wiring and other components are spec’d to handle the full discharge current of the cells, then that new pack could theoretically put out 120 amps or so. Given that, you could set your battery current as high as 60A per side.
You could also play with your motor current, increasing that can give a significant performance boost, especially at low speed. Keep an eye on temperatures though, if you go too far you can start to push the motors into magnetic saturation.
You can also bump up your battery regen current by 25% as well.
If that doesn’t markedly improve things, then I would suspect you might be running into thermal throttling, and need better cooing for your ESCs and/or motors to allow sustained operation at high loads (such as hill climbs).
Hello! I am using VSS with my Flipsky 6384 motors and don’t have the sensor cables plugged in for temperature but it still works. How the reliability would change if I had them plugged in?
Ah cool. So presumably if I pump up the Battery Current Max by 50% from 40 to 60 I would want to also pump up the Maximum Wattage from 1500000.0 to 2250000.0?
Real quick question (I think):
If I have 2 63100 flipsky motors at 190kv and run a 14s battery, how many do I need in parallel to get the full power in terms of motor current?
Not as quick as you might think…if I understand your question.
“Full power” can mean a lot of different things.
How many cells you would need in parallel depends on the cell you choose, the current you need, and might be limited by the space you have available.
How much current do you want to be able to draw?
Are you limited to a certain cell size?
I think I would use molicel p42 cells, but I am not limited in terms of size.
If I know how much current I need to draw, I can figure out the amount of parallel cells myself, the question really is: how much current could those motors draw max? Is that limited by the ESC? (it is a spintend ubox)
The ESC has complete control of the motor (hence the acronym, Electronic Speed Controller). The motor will only ever receive as much current, and the battery will only ever see as much draw as you allow in the ESC’s settings.
This is the wrong question IMO. “full power” is ambiguous, and you probably don’t want “full power” as I would define it.
Motor current and battery current are very different, and they are rarely equal under normal riding conditions. Lots of discussion on this topic already, so I’ll let you go dig for that if you want, to avoid clogging up the noob thread more than I already am. Suffice it to say that it’s entirely possible, even commonplace, to have your maximum set motor current (e.g. 80A) flowing through your motors, and be drawing less than 20A from the battery.
Increasing your battery current does have an impact on performance, but not as much as you might think at first. Motor current primarily affects maximum torque, and is felt the most under low speed, high load situations, like accelerating hard from a stop. Battery current affects your high-speed torque and acceleration - think like being able to goose it and pass someone quickly once you’re already going.
To answer the spirit of your question rather than the letter, 3-4p of P42A should give you far more than enough power to get into trouble with. 2p is still plenty peppy, but your range will be somewhat more limited.
I read already that battery current is WAY lower than motor current, but I had some trouble with the maths and really just wanted to know how if the battery I have (14s4p) will work of if I need a bigger one due to the large motors.
So your answer gave me pretty much everything I need, thank you! =)
