Project Aims

  • <$10,000 conversion
  • Top speed of +100kph
  • Range of +100km
  • Make a bike that is able to be registered
  • Make the bike look like the standard petrol version

Design Specification

  • Regenerative braking
  • 1990 Suzuki Across
  • 216VDC @ 15A (~3.2kw) VRLA battery pack (~90kg) for proof of concept
  • 3.3kw 3 phase induction motor

Wednesday, January 14, 2009

I have now completed the initial schematic for the 3-phase motor controller. I have to update some footprints and then it will be time to start on laying out the PCB. You can download stuff from here

Schematic PDF
gEDA project
gEDA symbols

The schematic is done on gEDA, the PCB will be done on PCB the PCB editor that is part of the gEDA package. The software will be written for avr-gcc. Both gEDA and the c complier are open source projects, and very good at what they do. Currently gEDA only works under Linux, but the compiler will work under windows with wan-avr(Just a side note I am designing and writing all the software for this project on a asus eee 901 with gentoo installed on it, a fantastic piece of hardware).

To get the gEDA project to run you need gEDA suite installed, uncompress the project and uncompress thje symbols. You need to make the gschem point to the schematic symbols directory. To do this you can simply uncompress the symbols tar into your linux home directory and everything should work. Otherwise you could just copy everything in the /lib/schematic directory into the /hardware/mc directory. Otherwise have a look in /hardware/commonrc and update/add the schematic directory to the component directories. One this is done change directory to the /hardware/mc directory and run make edit, this will open gschem. Have a look in /hardware/makefile for a list of other things you can use make to do.

Please if you do know anything about 3-phase motor control I would appreciate any feedback on the design, and I will post updates and corrections as I make them.


Kermit T. Car said...


I enjoyed looking at your project. It looks like a lot of fun. I'm worried about the rating of the IGBTs you have chosen, though. Figure 3 of the datasheet shows that you are well outside of the SOA for these devices, if you want 15A @ 300V. You also need some capacitance on your DC bus, to snub its stray inductance (I learned this the hard way-- you need more than you think. I'm adding 100uF to my car)


Bradley Jarvis said...

Hi Will,
Thanks for the feedback, with the IGBT I will be pulsing it so I can go outside the constant drive curve(I need to check the sample code I have to see what the pule width is). However I am probably pretty close to the limit of the chip so I will do some more research and find a more suitable one. I was thinking of adding the caps to the battery management where there will be a bit more room(I was contemplating if I actually need them, I don't have long cable runs but for the sake of some caps I don't think it's worth the headache of blowing some IGBT's). I hope you get your car back on the road, what IGBT's are used in your inverter?

Thanks, Brad

Kermit T. Car said...

Hi Brad,

I wouldn't count on the pulsing to give you more headroom. Rather the opposite, as there are switching losses. The safe operating area plots usually specify a really low duty cycle (<1%) for pulsed operation. I'm using the SKM145GB066D from Semikron. The inverter I'm using originally had SKM195GB063DN, but I couldn't find replacements. These are halfbridge style modules, with both high side and low side switches, as well as the freewheel diodes. The Semikron site has a lot of good information, particularly about thermal design. The only problem is the parts are kind of hard to get here in the states.

As for the caps, it's important to put them as close as possible to the switches, as there is enough inductance in even short leads to cause pretty high voltage transients. I'm using three of these:
Allied electronics part # 613-0601
and one of these, a little farther away:
Allied electronics part # 862-1278
A cap this big causes all kinds of inrush current headaches, though. I'm still working on dealing with that.

I'm just learning all this myself, so I'll keep you updated as I learn more :)


Bradley Jarvis said...

Hi Will,
I have designed the schematic with addition of caps in mind, you'll see that the board has some MOSFET drivers to switch 3 relays a charge, run and discharge. The microprocessor will handle swithing the relays, and since it has a Vdc IN monitor it would be able to detect when it is safe to engage the run relay(for this I have an EV200 relay). And the discharge was going to be used only when I need to do work on the motor control/ battery so I don't electricute myself on the caps. I guess the only issue is getting a resistor and relay combination that gives you a good charge time and durability without costing a lot.I may look at putting the caps in the inverter case inline with the cables, this gives me the flexibility to change this part as required, I will however add space for the snubber caps for each phase as close to the switches as I can get them

Thanks, Brad

Steve said...

Hello Bradley,

Awesome project mate. Especially the Variable Speed Drive. I've been watching your project for many months now. I have you "book marked" and I check back every couple weeks to see how things are progressing. Not often is there an EV project posted on the Net who's design is starting from the ground up. Well done.

I'm interested to know what will be the MAX Frequency = ? for your VSD. Standard 50 or 60Hz.?

Steve in Australia

Bradley Jarvis said...

Hi Steve,
Thanks for taking an interest. I am using the AT90PWM3, which has inputs and outputs designed to control among other things a 3-phase motor. I have not calculated what the maximum frequency generated will be but the limiting factor will be the motor and supply voltage. The VSD will easily do 60Hz.