VESC HW 4.12 temperature comparison

[hexakopter]

Hey guys,

I have done a comparison how much airflow is affecting the temperature rise of the VESC 4.12 hardware.
You can see the video here:


To give you an more detailed insight of the setup:
To the VESC itself. It is self soldered with a self build heatsink on top of the IRFS7530 MOSFETs. I was aware, that the heat transfer between the D2Pak plastic housing and the aluminum heatsink wouldn't be great but I did it anyways and also put thermal paste between them. The settings were at 50A max battery current and 120A max motor current. The MOSFET temperature limit start was set to 80°C and the end to 100°C.

I was running the test on my eBike with a MXUS 3000 v3 motor and 10000mAh Multistar 11s battery. I compared the temperatures when the VESC was inside my bike bag (so no airflow at all) and with the VESC + heatsink was hanging out of the bag.

I was logging all the VESC data with the bluetooth module from https://metr.at and the metr.at iOS app. The app is also available for android and there is also the perimetr app available to change the VESC settings on the fly.
Here you can see my logs from the metr.at logging app:
https://metr.at/r/jTArX (short run on hill with a big incline without air-cooling)
https://metr.at/r/9LggY (short run on hill with a big incline with air-cooling)
https://metr.at/r/QgKsx (longer run on hill with a smaller incline without air-cooling)
https://metr.at/r/Pk4XU (longer run on hill with a smaller incline with air-cooling)
This data was then overlaid to the video I have done with my Xiaomi Yi and a StorM32 BGC brushless gimbal.

I have also created two graphs of the two runs where you can see the different temperature rise better and also see that the current limiting caused by the temperature limit on the longer run also leads to a longer run while the distance was exactly the same.

So lets summarize what I have learned from the test. It is very important to use the airflow when it is possible for you to mount the VESC where the air is flowing. When mounted correctly the VESC 4.12 hardware is able to take a lot of current without getting to hot. I was aware that the airflow has a big influence to the temperature development, but not as big as seen in the graphs. Impressive. A lot of builds of electric skateboards I have seen dont use the airflow from driving at all. When not hitting the temperature limit thats ok, but if you need more power you know what to do.

If you have any question feel free to ask.

[Leon]

Cool video. It shows the software is able to protect the hardware sufficiently by ramping down the output power. Works as expected and nice that it just doesn't power off at a certain threshold.

Nice editing btw
//Leon

[hexakopter]

Thanks. Always nice to hear positive feedback.

I also mentioned two graphs in my last post, but I forgot to upload them. So they are coming with this post.
I also forgot to mention, that the elevation data isn't that accurate because I have no barometer in my phone, so they are just from the GPS data. You can see that they look quite different, although it was the same route.
The voltage of the run without air cooling was always a little bit higher than the run with air cooling, because I rode with the VESC inside the bag first and than the same route with the VESC hanging outside.

[benjamin]

Very nice comparison, thanks for sharing! It really shows that the airflow has a huge impact on performance.

What I'm missing from the video however is the motor current. The motor current is what causes the heat in both the motor and the FETs. The battery current and the wattage have very little to do with heat generation other than in the battery. What this also means is that since you use LiPo batteries, they are probably fine with a bit more current, so you can increase the battery current quite a bit as well if you want full torque at full speed.

[hexakopter]

Cool that you like it. Hearing that from you means a lot to me.

Next time I will try to adjust the overlay to my needs, so put the elevation data out and the motor and battery current in. But this time I was just using the default overlay available for the metr.at data. The motor current is just visible on the four metr.at logs I have posted.
In theory my Lipos should deliver 100A continuous current from the data hobby king provided. But I saw myself that this data are dream values, because the voltage drops a lot. I also read on other forums, that most people recommend using just 5-6c what means 50-60A.
Because I don't want to stress the batterys to much I just use 5c for discharge and 0.4c for charging (to just 4.15V per cell), so the battery lives longer. Putting the battery and motor current a little higher was planned with the VESC 6 when regenerative braking is build in.