openinverter forum

I've been observing your projects for a while now and would like to present my own work in progress. But first, a big thank you to everyone who contributed to this amazing wealth of knowledge on reusing OEM parts!

Ideation: In 2019, a friend and me did kind of a charity rally to Malawi. On our way there, we came through some game areas where those jacked-up, heavily equipped Land Rover Defenders of rich white tourists roamed. With some childish admiration we sometimes thought "I want such a thing when I'm grown up". But I was sure that wouldn't happen with a simple plug-and-play approach, and certainly not with a diesel-powered drivetrain.

Vehicle: So, when I came back home and had earned some money, I bought a Land Rover Defender exMoD from 1988 (actually they called them simply "one ten" back then) with an engine issue and comparably few rust holes, following this community's first rule and having learned from previous projects. (Turned out that I had to weld the whole bulkhead) Motor: I also learned about the EV conversion community and the motor options available. I didn't find any suitable OEM motor - Tesla parts were absurdly expensive, the Leaf motor didn't fit, and I didn't many other options -, so I went on Alibaba and purchased a suitable PMSM with controller from the Chinese automobile industry. Somehow I thought this was a good idea... VCU: I also thought I would be able to easily control it via CAN bus (which I hadn't known at this time) and that I'll just write my own code to combine the CAN controls with controls of analogue devices. Back then, I found only the EVSE GEVCU, and closed source Tesla/Leaf VCUs, which did not provide the flexibility I had in mind. Had I searched a bit better or waited a few more months, I would probably have found the openinverter project. Instead, I spent Covid programming my own open source VCU framework based on the ESP32, which I really like. It's an event-driven, non-hierarchical software design, which I can elaborate upon at a later stage. I also just created a little ESD-TVS protected dev board so that the system can be used anywhere in the car (still WIP). This project is progressing very slowly for many reasons, but the most depressing reason is that I'm still unable to spin the motor (and I don't have a proper manual, just a CAN matrix). I have it connected to various CAN sniffers, a 370V battery bench, and my little test bench. It listens and reacts to my VCU commands, but it doesn't switch into drive mode. I'm also unsure about the right start sequence, so I'd appreciate, if anyone could verify that my logic is correct: viewtopic.php?p=67619

Battery: There is a Tesla model S P85 battery pack sitting in my cellar, which I successfully managed to connect to a modified version of the SimpBMS. I will certainly use the SimpBMS code (big thanks to Tom De Bree!), but would prefer different hardware, maybe using my own ESP32 framework and board. Does someone have an overview of SimpBMS-compatible hardware projects post V2.3, which don't rely on the Teensy 3.2? Charging: I fried my first Outlander OBC while connecting it to the pack before I learned how to check the polarity of HV cables in some thread around here. But the next one arrived, and I will perform a first test charge with my granny cable, as soon as my Simp-based BMS is ready. I have got a standard charge inlet to connect the AC lines and CP/PP, which another ESP32 will handle. I also want to add a BMWi3-based CCS system, but that will come at a later stage.

Drivetrain: Giving some thought to the gear/diff ratios, I decided to drop the LT77 gearbox, flange the motor onto the LT230 transfer box with a 1.667:1 ratio, and replace the axle diffs with a 4.75 ratio. That should give enough climbing ability, even in rear-wheel drive. The transfer box motor adapter is a complicated piece of work, as it needs to house the shaft adapter, be long enough to overcome the LT230 diff box, fit into the transmission tunnel, bolt onto the original frame mounts, and hold the LT230 diff/ratio shifter. That goes far beyond my skills, so I have it manufactured by a professional. Other bits and pieces include a battery heating/cooling system with some sort of smart valves to utilise waste heat (or the battery heater) in the cabin, too; an ISA IVT-S, an electrical power steering pump, and a vacuum pump or an iBooster. Do you think the iBooster is worth it?

I will use original gauges and instrumentation for the feel of it, and add a simple BMS screen, maybe with an eInk display.

I'm happy to keep this thread updated, although it might take a while to complete this project.

I like your software approach. Are you aware of : https://github.com/collin80/GEVCU7 . Quite similar approach and the most mature SW architecture I now for EV conversions. There are more https://github.com/mumme74/CarTransporter with good approach for PID, https://github.com/jamiejones85/ESP32-C ... 32-Chademo and my work https://github.com/Modellfan/TeensyVCU.

I would love to merge this together to one Arduino ECU framework being compatible with ESP32, Teensy and DUE controllers. A lot of dry powder beeing used for redoing stuff :-/

Thanks for sharing! I wasn't aware of version 7 of the GEVCU, neither of your approach, which both seem to use some kind of cooperative multi-tasking. All the double work is a pity, indeed, but I think this is mainly due to several people seeing an application gap and starting to work on it more or less at the same time (though some take a lot longer to finish, as is the case for me). A communication problem, in the end, which can be solved by reading forum posts (which I failed to do in the early 2020s). Why did you start your own VCU software?

A hardware-independent ECU framework would be great and could combine all our learnings. Though it would be quite some work, and we would have to use cooperative multi-tasking, which is not always compatible with event-based architectures, if I got it right. Anyways, I'm happy to contribute drivers for ESP32 peripherals and FreeRTOS multi-tasking (though preemptive).

More and more i am convinced your motor uses CAN FD for communication instead of plain CAN bus. Try to see if you receive any reports on 0x105 telegram if you rotate the shaft with inverter started.