We have been spending the last few weeks collecting a lot of information about motors, controllers, batteries, etc. We have all been getting a quick education in these topics. But up to now we have not really made any solid decisions about how to do this conversion.
There are so many things to decide on when you start looking at what is involved in the conversion. AC or DC system? Once you decide that, which specific motor? What kind of battery technology? Which batteries? Motor controller, charger ... it goes on and on.
I want to nail down some of these decisions so we can start making some progress on solving specific problems. So here is what we have decided so far:
Originally I wanted to use an AC system because of the better performance and regenerative braking. But besides the additional expense of these motors and controllers, you also need a higher voltage battery pack. But the biggest decider for me is the much larger amount of information available about conversions using DC systems instead of AC. Remember, I want to get a useable daily driver out of this sometime in the not too distant future.
So, for the motor we will go with a DC system. The question is still open whether to use the ADC 4001 motor or the Netgain Warp9. I am not sure I get the difference yet, if there even is one. But at the moment I would say we are about 80% sure we will go with the ADC.
We are going to use the LiFePO4 batteries. This has been a much easier decision. These appear to be a superior choice to lead acid batteries because of better battery life, lower weight and higher energy density. Even though they are somewhat more expensive at the outset, if they last a little longer than the lead acid then it will be break-even, and if they last as long as claimed, then in the long run they will be quite a bit less expensive. I realize that I am taking some risk because of new and not well proven technology, and also taking a risk with the suppliers. As for the supplier, we will probably use 40 cells of the Thunder Sky(dead link) 150AH batteries. This will give a nominal pack voltage of 128V and 19.2 kWH capacity.
Using the LiFePO4 batteries does mean that the battery management and charging system will be more complex than with lead acid batteries. The management system is still up in the air but we are thinking about building our own.
For the controller, I was thinking about going with the Synkromotive controller. This looks like a good choice, better performance that the Curtis 1231 I had as a placeholder. And I like the fact that it has a computer interface for configuration and diagnostics. Plus built-in cooling fan and contactor. So this is what I was planning on until ...
... I found this site, with an "open-source" motor controller. We are all excited about the idea of building the controller instead of buying it, so I think this is what we are going to use. I am going to modify the controller design to use a Stellaris microcontroller instead of the ATMega that is on the existing design (my company makes Stellaris controllers and I am intimately familiar with programming them). We are going to start by prototyping a controller and testing it with the go-kart. If it looks feasible, then we will build a complete system to use in the car.
How to attach the motor to the drivetrain. The question of direct drive comes up, but from what I have seen the general consensus is the DC motors do not have enough RPM range (or torque over the range) to be use in a direct drive system that will go from low speed neighborhood driving to freeway speeds. So we will mount the motor to the transmission. Which raises the next question: clutch or clutchless? Opinion on the web forums seems to be split on this. I have a hard time being convinced that keeping the clutch is the better choice. It is more mass to spin up every time you start up the motor. Plus it appears to me that the mechanics involved in this kind of mount is more complicated. Amongst us, we are all in agreement that we want to do a clutchless coupling.