Questions about building a BLDC power stage for 48V, 22 AMPS

youngfg

1 µW
Joined
Jan 30, 2014
Messages
2
Power stage design for BLDC hub motor, 48 V, 22 Amps.

You may wonder why I would want to build a controller with such a normal power rating.

I am helping a high school team on their Shell Eco Marathon car.
shell-eco-marathon-mater-dei.jpg

mater-dei-prototype.jpg

The car uses a bicycle hub motor in the rear, it's a fairly common motor, 500W with hall sensors, and we have used a 10 cell Li-ion battery pack and a off the shelf 48V, 22amp Chinese controller.
With this combo we have won several competitions an set some new records.
Now I think Shell is trying to limit the number of electric cars in the field, so this year they have changed the rules, they are requiring the controller to be built by the teams.
No off the shelf controllers, no modified controllers, and to prove you built the controller the circuit board must bear the SEM, Shell Eco Maraton, mark.

I have Lebowski's chip on order, and I think we may also try using a MC33033 chip, from the Simple BLDC controller thread.
So from the forum I have a good idea on how the "brains" of the controller work, but I can't find any info on a simple power stage.
The threads I have found are for stages that a way overkill for what we need.
I need a efficient, relatively low power stage, that hopefully is simple so I can have the students build it.

The way we use the car follows.

We start from a dead stop, no pushing allowed, accelerate up to about 19 mph, then cut the motor and coast down to about 14 mph. We repeat this process for the entire race which is 25 minutes long.
We do 11 laps around the track, which is about 1 mile around. We may only use the motor two are three times per lap. At the end of the last lap we coast to the finish, brakes are never used.

Last year our best run used 47841 joules from the battery, and I think the maximum current draw was 12 amps.

Any guidance on component selection, general controller design, or links to existing designs would be a big help.

Thanks
Gene Young
 
Hi Gene !

i have been in talks with Robert Neisen on this project for a number of years, this latest rule is a nasty one to push on high school students, and worse yet, shell wants their brand stamped on the boards and likely claim ownership on all documentation submitted to it's specs and innovations.

We talked about possible Brushed motor vs Brushless to simplify controller design, but brushless is more efficient overall, so if it's possible, it is the best bet.

a 6 fet controller should be plenty, heck, even a 3 fet should work for the simplest possible layout.

Do you know if any other teams are managing to come up with a working solution ?
 
I don't think Shell plans on "Stealing" anything, they just need an easy way in the tech line to tell if you built your controller.
With 1200 cars to tech it's a big job.
I can just imagine someone in the tech line saying "Yeah EB709XC, that's what we decided name our controller"
At least with this rule all they have to do is look for "SEM" and you pass.


Do you have any schematics of a 3 or 6 FET controller?
 
Back
Top