
I've had an on/off interest in EVs for about 10 years now, but I am now at a place in life where I can really start exploring that interest. With that said, my first project isn't an EV project at all. It's a hybridization of an ICE car using a GM E-assist motor and an inverter that I picked up for $80 USD...actually, I grabbed 2 of them because it would be foolish not to at that price. Why hybrid? Because I enjoy the idea of bolt-on hp and torque. It also allows me to get my feet wet in the EV world and build up a solid knowledge base. Ok, enough of the intro.
This thread is intended to be a log of the hacking a 9th Generation Honda Accord Hybrid inverter. I have no clue what I'm doing, so the learning curve looks more like a wall. I'll need the communities help.
Let's go over some specs for this inverter:
It is VERY similar to the Toyota inverters in its construction (MG1, MG2 and buck/boost converter).
Rated for 124kW in the HAH.
Boosts voltage up to 700VDC.
Can take a battery input voltage of 320VDC nominal.
Liquid cooled.
Power module is integrated with the heat exchanger, eliminating the need for thermal paste and making thermal transfer more efficient.
https://info.ornl.gov/sites/publication ... b52422.pdf
Info grabbed from here, starts at page 21.
https://www.energy.gov/sites/prod/files ... _o_web.pdf
The beautiful thing about this inverter is that it seems all of the control signals are available at the 20 pin connector from the control board. On page 12, it looks like the "ORNL Controller" connects to the connector on the right side of the driver board. That's it.
So I did a bit more research...$25 and a few hours later I find that the 17 pins on that connector are for the upper and lower switches for the Traction Motor and Generator motor (total of 12 wires) and 5 wires named SCIGV, SCIVG, SCIMV, SCIVM, and GDEN. I'm assuming that GDEN is Gate Driver Enable? I'll figure that out as the thread continues. Honda has great wiring diagrams, btw. The OEM control board is nice, but I have no use for it. It has double of everything for the Generator and Motor sections, 2 micros, 2 resolver-to-digital. I technically *could* use the CAN signals to control it as I do have the dbc file, but it still a ton of work as it's around 1600 lines worth of signals to decode. I'll upload some photos later on.