openinverter forum

Edit 17.01.23
This introduction no longer really reflects my build but it shows where I started!, while I have tried to maintain doing everything on as tight budget as possible I have ultimately spent a few more quid than Damien’s £1000 ev challenge!…………


Hi all Welcome to my budget RX8 build thread, I’ll start by highlighting the budget theme, so don’t expect any Tesla parts here, if it wasn’t super cheap, free or I could build it myself it pretty much hasn’t been an option, it started of as and attempt at Damiens €1000 build bit is now progressing to include to use of bmw phev battery modules I have managed to get a good deal on,

Spec as built was:

13” series wound dc motor, new brushes bearings and armature balanced to 6000rpm

Direct coupling to gearbox

13 x 75ah flooded lead acid batteries (always destined as proof of concept only)

Prius gen2 dc dc convertor salvaged from a damaged £20
inverter

Chinese 3.3kw charger

Home built igbt chopper motor controller evbmw driver board arduino uno controlled logic

Currently I am working on battery boxes, bms and building a paralleled pack out of 3 complete bmw 330e phev hybrid battery packs total of 15 modules arranged 5s3p for a total of 24ish kwh

So work on the battery conversion has been ticking along, and the front pack is done and ready to be mounted, I’m using bmw330e modules 3 whole packs worth in a 3p5s configuration, paralleled at a module level with big bus bars using the original connectors soldered on, also paralleled at the cell tap level so each 3 modules is on one cell monitoring slave, have tested on a quick and dirty stripboard simp bms and everything is behaving

Have been working on mounting and connecting the last 9 bmw modules, still to finish the bms wiring and re mount the charger before I can test things

Nice project!

Thanks

Ok so it would seem I’ve found the limits of the gen2 Prius dc convertor, hooked up the batteries with 3.6v per cell 288v pack total, precharged, contractors close, no 14v at the 12v battery! So now I either need to step Voltage down to 200 for just the converter or find something else

Apparently some resistor can be swapped to extend the limit. Hope somebody chimes in who knows which one.

That would be amazing if anyone has that info!

I don't know if this will help or not as it's for a Gen3 but maybe worth a watch.

Thanks for the link, I think I will have to pull the gen2 off and get it apart and start hunting for similarities, see if I can find the resistors to change, I presume they are voltage dividers for the h/v reference coming into the converter

Ok so after playing about with the gen2 dc converter over the weekend and some helpful bench testing by others I found my problem, I had originally had issues with wiring my dc converter straight to my controller dc bus because during pre charge when voltage went over 100ish the converter would go active and the bus voltage would stop climbing, what I did was separate the converter from the bus with another contactor which came on after pre charge, this setup seemed to work find for up to 200volts, but what I should have done was change the value of pre charge resistor to allow enough current flow with the dc converter going active and still finish the pre charge, with this done everything works as it should!

What if you don't enable the DC-DC until DC switch is closed? With that 12V enable signal I mean.

I think it just wasn’t enjoying the inrush from being behind a contactor, either way it’s working now so on to the next headache

Damm I let the smoke out

So the battery upgrade went well and all seemed to be working fine, when the controller popped, ultimately my fault as I was changing so stuff in my controller code, change...test...change...test....change, forget to let precharge complete before mashing the throttle to hit current limit main contactor closes inrush current pops the 2 drive side igbts and takes out the gate driver board also.

So aside from having to rebuild the controller and add an interlock with a small delay to the auxiliary contact, first impressions of the battery upgrade amazing!

Ev8 wrote: ↑Mon Apr 12, 2021 3:19 pm Damm I let the smoke out

Good memories! Myself i let the smoke out at least 6 times.
If you want to make DIY power section you need to have:
a) desat protected drivers so you can calibrate your dead time and boost.
b) You need to stack DC rails and phase outputs as much as possible.

check this out...This one lasted for a year or so... https://mazdamx3ev.wordpress.com/2015/0 ... r-za-300v/

Then we designed proper DC rails. 3 years uninterrupted driving with Mazda DIY inverter speaks for itself.

Very nice work there, so it turns out a software implemented over current protection can’t match a gate driver with desat, however I’m not even sure that would have saved it given the inrush

So, as a result of the magic smoke coming out I have decided to give making a dc controller from a Prius gen2 inverter a go, it would appear a few have tried and failed before me, specifically burning out the low side igbts, so I have a crazy idea about just using the all of (mg1&2) paralleled high side igbts for switching and then use a monster diode module mounted where the boost inductor would be as a freewheeling diode, effectively building a high side chopper rather than an half bridge, watch this space.,.

I have built a written code for a prototype board, with a teensy 3.2 microcontroller, incorporated mg1/2 control, current feedback from Toyota hardware, mg1/2 fail status, contactor and precharge control, brake and reverse switch inputs allowing different throttle profiles and current limiting in forward and reverse, 700amps and fast throttle ramp was interesting in reverse with the old controller! Oh and a temperature gauge driver, cos why not

Will release code and schematics when tested, but that won’t be for a while as I’ve just had elbow replacement surgery!

So a smallish update is probably due, just before my surgery I got a frame built to physically mount a gen2 inverter under the bonnet, and started a new box that’ll bolt on the side housing the positive contactor, precharge etc and the prototype board I have built to control it all. I have now got to the point of bench testing systems and things are going ok well I have control over mg1 and mg2, able to read current sensors, temperatures, voltages and fail pins, other inputs include a bms interlock, brake and reverse switch inputs, code includes different throttle ramps, rpm limits and current limits for forward and reverse modes, hopefully I will get to install into the car in the next couple of weeks

That inverter cover looks really tidy with a bit of paint on it. Might have to get the rattle can out for mine this week.

Just out of interest whats the air intake for?

The fan cooling the big DC motor underneath! This is a Frankensteined gen2 dc controller

I just can’t help myself from painting things!

Ok a bit of a update, I have finished the gen2 Prius inverter as a DC driver install, upgraded my motor cooling fan and charged the battery pack, have just completed an 18.5 miles test drive to shake it down, inverter is behaving well but I need to work on my code a little as the throttle limiting based on current doesn’t work great and it seems to often overshoot and hit my hard current limit,

Simp bms seems to freek out and reset battery SOC to 100% if I go over the 500amps range of of the current sensor, need to look at that
Also the teensy controlling the bms decided to stop working and melt down literally could cook on it! No idea why this was the main reason for the test drive being only 18 miles

More testing today, I have replaced the teensy and found an issue with a short on the bms board, changed some things in the bms software to stop the strange behaviour under high load, no more resetting soc to 100%

I have managed 33miles of testing including an 80mph cruise, battery soc is down to 52% so hopefully I should see 60 miles range which seems pretty decent for 22kw worth of used bmw phev modules given that dc motors aren’t exa the most efficient!