[kennybobby]

Anyone have any predictions as to the likely main bottlenecks here?

.

How to send commands over CAN buss to turn ON charging, and keep it running with voltage and current feedback from the BMS for the pack.

Can you tell who built the OBC electronics, are they tesla-related? Nichicon builds the Japanese Mits and Nissan units, and i have traced out the major circuits for those.

You will need to pull the covers open and post some pictures to get started.

 

[vauron]

Thanks, kennybobby. Meanwhile the OBC from the Kia Soul EV has arrived. Turns out it's a Mando type MCS20, see https://www.marklines.com/en/top500/s500_239 for details of Mando (South Korean automotive company started in 1962 by younger brother of the founder of Hyundai as Hyundai International, renamed Mando in 1980, today 12,000 employees). However Daewoo's brand also appears on the "Main Case" of the unit, making it unclear who actually designed it.

Since GM bought Daewoo when it ran into financial difficulties, conceivably there's some commonality between this 7.2 kW charger and the one in the Chevy Bolt. Since I have both (and also the Chevy Volt battery) I'm in a good position to compare them.

As to what's inside the Kia/Mando/Daewoo OBC, I undid the 19 bolts holding the top on but the seal seems to be holding it down very firmly. I was hoping the seal might soften with heat, but heating it to 200 C didn't change it at all. I'll keep trying in parallel with investigating the commands. If I can look at the OBD of a Kia Soul EV while charging with 240 volts (presumably with the ignition off), see what traffic there is, and then see what effect those commands have on my OBC, I might learn something useful.

 

[kennybobby]

The lid is glued with some sort of RTV on the Mits units also. i use a knife blade to get started cutting thru at a corner, then putty knife and other thin tools to wedge in the gap and work around to free it up. Comes off with no damage or only slight bend at screw holes, little hammer and anvil work and it's GAN.

Quite a bit of merger and acquisition activity with Mando, it could even be a Bosch design similar to BMW.

 

[Sloth]

Hi
Are there any progress? I am thinking of buying such a charger, but I don't know if it can be controlled.

 

[vauron]

Sorry, no news on that front. When my Volt battery arrived it still had something like 345 volts, which was more than enough to verify that the two inverters (one per wheel) could precharge and then spin the wheels on the lift.
Once the car is on the road and using up the Volt's battery I'll have to "fill 'er up", which I may do the first time by trickle charging the individual modules (to 96 volts each, ±24 volts for respectively the long and short modules). But that's only convenient while the body is off, allowing me to easily disconnect the modules from each other.
At some point I'd like to borrow someone's Soul EV and see if I can see messages on its OBD-II connector when an EVSE is hooked up.
In the meantime I've found myself wondering why the charger needs to be onboard. In your own garage it could just as well be installed in the garage as in the car, like a conventional Level 1 or 2 EVSE but outputting 380 VDC instead of merely connecting the car to the grid, but at a small fraction of what's expected of a serious Level 3 charger. Whenever we take my wife's Bolt on anything longer than 150 miles round trip we bring the EVSE in case we can't find a Level 3 charger and have to settle for overnight charging at 25 mph at a Level 2 charger like Blink's.
Now if this EVSE replacement weighed the same as the Kia's OBC, namely 34 lbs, that would be a lot less convenient than packing a lightweight conventional EVSE. But must an OBC be that heavy? Last week I built a two-stage Cockcroft-Walton (CW) voltage multiplier taking 240 VAC in and producing 8 amps out at whatever the battery voltage is, acting a bit like a constant-current source according to Simulink's simulation of it. Without an enclosure it weighs 1.1 lbs; if it output 16 amps it would have to weigh twice that (the capacitors scale with the power and the only massive components in the CW circuit are its capacitors). So far I've only been brave enough to verify that it works as designed with 24 VAC in, I'm still checking the simulation to see whether precharging is needed to avoid back-biasing the aluminum electrolytic capacitors during startup. Also there's a lot of ripple in the outer pumping capacitors that may overheat them at 240 VAC in. If anyone has tried the CW multiplier in this application, or who has a better idea for an "external OBC" (i.e. an off-board charger) that has about the same weight advantages, I'd love to hear from them.