Paul & Sabrina's cheap DIY 144v motor controller

[MPaulHolmes]

Hey Matt, fixed a couple mistakes on the code chunk. I'm going to test parts of it in the simulator when I get back from school. Dang I'm late! Then I'm going to pulse the throttle from 4 amps to 50 amps to 4 amps to 50 amps...etc..., while the rotor is locked (emergency brake). I'll be pulsing it 4 times a second, and then I'll gradually turn up the P (with I starting at 0). The oscilloscope should be able to lock on to that signal, so I can watch if there is overshoot. Then I'll back off the P a little, and gradually increase the I, doing the same thing. So my loop will keep chasing 50 amps, and then turn around and come after the 4 amps. I picked 4 amps instead of 0 because 0 is a special case (ben hurtling down the street to his death).

[Christ]

Sounds like the sensors could be gotten from any ABS vehicle...

[bennelson]

Originally Posted by MPaulHolmes

(ben hurtling down the street to his death).

Come on guys! I'm right here!

(Ben leaves room to purchase 5-point racing harness...)

[MazdaMatt]

Okay, back on topic. Paul... this may sound crazy... but when you do your PCB design for the controller, could you connect an SD card slot to the SPI (Serial Periferal Interface) lines on the mcu? Does that mcu HAVE SPI lines? It would be AWESOME to datalog these numbers. You pop in a card, drive around for a while "trying stuff", then go back home and see the numbers in an excel chart.

Also, please do that pulsing test at dead stop, low speed and high speed as the motor will react differently due to its increased EMF at higher speeds. I eventually will redesign your controller to account for the dI/dD (rate of change of current with respect to the rate of change of PWM duty cycle... you're the teacher, did i note that correctly, or should it be dI/d^2D?) Anyway, the point is, it would help the system response for the system to be able to figure out how quickly it needs to adjust.

To be clear, the system says "i have 0 current and pedal is at 100, hocus pocus, output is 10", "i have 5 current and pedal is at 100, hocus pocus, output is at 200 because it saw that current went up 5 from output change of 10, so 2:1 meaning we need to shoot for 200 to get 100", "I have 100 current and pedal is at 100, hocus pocus, maintain output of 200"

Obviously we wouldn't be that agressive.

[Tango Charlie]

Originally Posted by MazdaMatt

...the system says "i have 0 current and pedal is at 100, hocus pocus, output is 10"...

Aha! I KNEW all that computer stuff was magic! Now I have proof!


Sorry, back on subject...

[Blue Bomber Man]

I agree that the concept could be pretty simple, but we need to remember that wheel speeds do very even when the vehicle has complete traction. During a turn wheel speeds vary to each other depending on the steering position. The best way for me to show exactly what I mean is with this graphic from carbible.com



The system will probably need some tolerance built in plus something to account for the differences in radii.

[MazdaMatt]

Make a new thread re: awd once bed has produced a few of these controllers Nothing wrong with 100hp per wheel, is there? TEEHEE!!!! Think of how fast we could kill Ben.

[Christ]

I dont mean to be a party-pooper on the "kill Ben" thing... but if something did happen to him, via the controller or not, you guys are really bringing attention toward a conspiracy... LOL.

Carry on!

[Christ]

Originally Posted by Blue Bomber Man

I agree that the concept could be pretty simple, but we need to remember that wheel speeds do very even when the vehicle has complete traction. During a turn wheel speeds vary to each other depending on the steering position. The best way for me to show exactly what I mean is with this graphic from carbible.com



The system will probably need some tolerance built in plus something to account for the differences in radii.

Of course wheel speed varies around radii, but a simple input sensor on the steering system could control whether or not the traction control took effect, or limit exactly how much "give" there was in the duty of the traction controller (obviously, you don't want your TC to completely let go while you're going around a corner, but maybe limit it's effect so it's not so aggressive about making all the wheels spin the same speed.)

I think the right logic would allow the system to "understand" that you were cornering though, and compensate, since the values for one wheel vs the other wheel on a corner (with a given value for the steering correction sensor) will remain mostly constant.

The programming has to have some "give" anyway, for things like uneven tire wear, etc.

[Blue Bomber Man]

FYI my name is also Ben

Anyways... I think it would be a lot of fun to continue to work on projects like these and keep them open source. Once Paul is finished with his design of his controller we could start deciding on how we could feasibly create a traction control/AWD system and make that cheap as well. The individual motors for each wheel might remain expensive, but if someone wanted to use small motors for each wheel and depend on their total power, the only real limiting factor to cheap electric car would be the cost of the batteries. And we all know that those prices are probably going to plummet over the next 5 years.

I think the price on Lithium Ion batteries is around $700-900 a kw-hour. A vehicle that was light weight and had a low CdA could get by with maybe 15 kw-hours for a 100mile + range. I foresee all these new cars that could be using the Lithium tech driving the price down to maybe $200-400 per kw-hour.

An electric motorcycle could cost 10 grand brand new and have 120 mile range or better. The possibilities are limitless.