Quick intro:
I'm Thomas. I live in Melbourne, Australia and am studying my undergraduate degree in EE right now.
I have worked on an almost scratch built e-bike project in the past which sorta led me into the world of diy EVs, when i randomly stumbled upon the OI wiki whilst looking for controllers suitable for ebikes.
Given how car-centric this damn city is, its not really viable to get around without a car. I think I'l embrace it with as much sustainability and budget a (broke) uni kid can, and build my own DIY EV!
Many thanks to the community at large for sharing all of your hard work and ingenuity. I hope to contribute something here too!
The plan:
1. I want to convert the GT86, its a very common car here in Melbourne and has always been my dream first car.
2. Will utilise Gs450h drivetrain.
3. 8kWh LTO battery pack, to be tied directly into the Gs450h inverter HV bus as usual. The brand of cells is "TianKang" which I am importing from China with purportedly 93% SoH - the seller says they came out of an electric train which is pretty awesome.
4. Make use of zombieverter for vehicle control.
Thinking:
1. It handles well, is light (1300kg), is RWD and should be very aero and thus have good kms/kWh.
2. The second time I came to the forums was when I found Phillip Schuster's GT86 conversion through his Youtube. It's an awesome conversion, brilliant workmanship. But the one thing that troubled me was how little space the car has, which meant Phillip had to use 4 or 5 seperate battery compartments which seems a huge PITA, and could be unreliable. Finding out about the Gs450h, the choice was obvious, as the engine bay is opened up for a large battery - though to a limit as the weight distribution would probably be horrible if I jammed it full of batteries. Conservative upper limit of 120kg of batteries which is the mass of the engine it would replace.
3. The concept of an unkillable chemistry appealed to me very much - immortal battery pack. Furthermore, in the GT86 chassis it seems impossible to fit a very large amount of battery easily. If the capacity is limited, C-rate must be high enough to support the 200kW Gs450h. I weighed up my options, and LTO seemed to be the safest, most impressive cycle life option. Cons are that they have poor energy density, but the main design criteria I am aiming for is power instead of range. So this isnt a huge problem, at 25C, 8kWh is enough. Which is about 143kg, 70L.
4. de facto for Gs450h. Thanks to all the work done by Damien, johannes, xp677, and more whom I'm not aware of.
Progress so far:
1. So, the project is finally kicking off because I have finally pulled the trigger and bought the car I wish to convert. It's my first car
It is written off though, and needs some repairs before it can be registered again. Once the conversion is done another certification will be needed before it becomes road legal with the EV modifications. This paper-work seem scary.2. About July 2024 I sourced a Gs450h tranny and inverter from a nice bloke who is a retired panel beater. He was stripping the car for parts and got it for $1500AUD which seemed very good.
3. These cells were purchased for a mind blowing $500AUD/8kWh. W H A T. They are shipping via sea freight right now, I am incredibly lucky my parents operate a small business in customs & logistics which makes this fairly simple.
4. I assembled my own zombieverter, with board from JLC and some missing chips from digikey. It has been programmed and hooked up to the inverter. After a bit of troubleshooting (it was surprisingly very well behaved), the Gs450h is now spinning! The project is practically done then... right? whats an EV but a big box built around an electric motor.
Future:
An idea for the future with this project (and potentially any Gs450h project, though it makes most sense for my choice of chemistry) is to make use of the boost converter in the Gs450h inverter as a "range-extender". I can make use of a more power dense chemistry that doesn't necessarily support the same power density as the LTO pack. And recharge the LTO pack at up to 30kW which is plenty for any legal cruising speed. The big challenge here is controlling the boost converter which afaik no one has figured out. Some others have been curious about this, I was heavily dissuaded when Johannes himself said something like "that will be hard, start with something simpler". I haven't been able to put down the idea entirely, so the system should now be modular and work as a 'bog standard gs450h conversion" with the LTO pack only, albeit with very poor range. With opportunity to increase that range.
I plan to share progress updates on youtube, and will use this forum thread as a blog for more details.
https://www.youtube.com/@tamashamas6193
Here are some photos of the things mentioned here
