Hi all,
finally I got my motor and everything setup. I run the whole thing on 48V batteries. The motor uses approx. 80A constant with 120A peak.
So judging by the specs the controller should be able to handle it. However, going uphill the controller gets too hot and shuts down.
Am I overusing the controller?
If so, would you have a recommendation for a good but buyable controller with 120-150A constant?
Thanks a lot,
Marc
Most controllers that can run that much current are going to need some cooling aids. A good way to do this is to mount the controller to a heatsink with externally exposed cooling fins, this usually solves most heating issues
I did a thermal analysis on the old and new version of this. Basically the new one with the aluminum PCB is amazing and the old one is one of the worst performing controller thermally that Iāve seen. The only one worse is the original 4.12.
5 Likes
Thanks a lot for the quick replies. So there is hope!?
This is the version Iām using:
Can you elaborate on the heatsink and fins stuff a bit? Ideally with a picture or reference. I have no clue what to do. 
Thanks a lot, Marc
48 is not a multiple of 4.2 or 3.65 or 3.6.
4 Likes
There are some good pictures and examples if you scroll through that firmware thread.
Since these are through hole mosfets, you could try to find some lower RDS On mosfets and swap them out with a basic soldering ironā¦ but I think that package just has its limitations. Probably better off selling it and getting the aluminum PCB version. Still a great price at 130. Not sure if itās also a stolen design thoughā¦
1 Like
Thanks guys. Iāll rethink my design, try and put the controller into direct airflow and also glue some heatsinks to the back. Iāll keep you posted.
If it doesnāt work, Iāll switch the controller.
The new version looks still much better than a suitable Kelly controllerā¦
Changing to lower Rds-on FETs usually means higher gate charges and that increases switching losses and heating.
Itās often a delicate balance between gate charge, Rds-on, switching frequency, gate drive current, turn-on and off times, etc. We canāt always say lower Rds-on = less heating.
2 Likes
Itās a multiple of 4
Letās just call it 12s.
2 Likes
yes, itās based on two 12s4p.
I read some more and learned that the controller should 2x-3x the capacity of the motor. This would man 160-240A continuous is that true!?
This is 50.4V 
2 Likes
Precisely.
I know people and some companies calculate using 4v per cell which is easier but itās the same with 1GB being 1000MB which is just not true. Itās 1024MB.
The good thing is you say 12s and people know what you mean.
3 Likes
The ā48Vā is not a calculation or number, itās a name. Just like George is a name. Thatās why itās called ānominalā. You canāt use names in place of measurements when engineering vehicles.
2 Likes
Ok okā¦ but if the battery is empty, itās less than 48v. 
Can we try and focus on the heat issue and/or alternative controller?
Thanks a lot.
IEC 80000-13 specifies 1GB as 10^9 bytes or 1000MB, while a GiB is 2^30 bytes or 1024MB. But I digress
5 Likes
3 Likes
Because retailers had a different language than engineers, this caused a ton of strife which culminated in the official designations of Gigabytes vs. Gibibytes to differentiate between marketing speak and engineering language.
3 Likes
Alright, weāll call it a draw.
Sorry for the OT.
No the engineers straight up lost the data nomenclature battle.
I just pray we can win the battery voltage fight.