openinverter forum

Great solution, and less noise.

I had some doubts about how well it would work, but testing showered me with high pressure atf so seems up to the job

At last i can no longer put off the control electronics!

Will the lack of availability of Modice enclosures and the fact I want to drive both mg1 and mg2 in this inverter I have squeezed 2 gen2 Prius boards into and old vw ecu housing

I have also found resolver pin outs for the rx400 transaxle so that should help take the guesswork out of resolver wiring

Well a little testing with bits of inverter and motor on the smallest table possible has confirmed resolver wiring is good

https://youtu.be/7AnE_7L9ZMk

I’ve had time to do most of the car wiring, up to a point where I could test/tune mg2, foc tuning with 2 motors on the same shaft seem more difficult, mg1 is currently disconnected at the phase wires but when performing foc tuning the area range of values between reserve rotation and forward is a lot I get from reverse rotation at 10a at 7000 and forward direction 10a at 14500 everything in between is no movement with 100a so I’ve set it to 10500 and it seems to behave well enough for some low speed testing, a little juddering when first setting off, I will tune further when I have a working cooling loop and power steering, please excuse the Rusty brakes and flat tyres the car has been neglected

A couple of other random pictures of recent things Vacuum pump control circuit squeezed into the vcu case
Prius power switch for start signal and manual regen pot

A quick question for anyone interested, currently UDC is measured by mg2 vcu and this controls the contactors, mg2 vcu also takes the analog pot feeds which are mapped across can to mg1 vcu, this works fine however if I try to map across udc via can I get nothing.
I was reluctant to connect the vh pin on the inverter to both vcu as it might require some adjustment to the udc gain? But it not possible via can this is the way to go.

Ok interestingly with udc mapped over can from mg2 I can see if I plot udc on mg1 that the data is getting across because at random intervals the plot jumps to battery voltage, instead of the adc -84v the floating pin gives. I wonder if tieing that pin to ground would make a difference?

A little forum digging found the udc answer
https://openinverter.org/forum/viewtopic.php?t=1428
Useful info for anyone else wishing to run 2 boards/motors
I will connect the inverter vh pin to the mg1 vcu as well and report back if I need to rescale anything.

Hmmm, so if there's no CAN and run the 2vcu's in parralel, there could be gain issues? Will await your findings.. I was thinking of putting diodes to all inputs running in parallel, with 'dummy relays' for precharge and dcsw on 2nd vcu.

Currently my set up is mg2 vcu does contactor control, takes analog pot inputs shares pots over can to mg1 vcu, all +12v logic inputs from the car ie, brake,start,forward,reverse are connected to both vcu, although have since discovered I could have shared this via can also.
I think I’m happy for mg1 vcu to have no influence over the contactors, if there’s an mg1 inverter fault condition it’ll shut down the igbts anyway it’ll be down to me to restart both vcu to reset. But I will let you know how that goes, gen2 inverters are cheap!

Also interesting side note syncof or rx400h mg2 also appears to be 10200 much like oth have found on the Prius

[media] [/media]

Mg2 only still but progress 3rd gear, conservative parameters, fmax300, throtcur3, throtramp10,

Needs further tuning on ramps as I get a lot of driveline clatter at slow speed when transitioning between throttle and regen,

Nice going. Bravo to get the wheels turning!

Nice! much of vibration was fixed in later firmwares, which one are you running?

I just updated to the 5.17 before this test yesterday, I think the issue is with physical slack in the driveline and to aggressive ramps currently, I exprerienced similar harsh chatter before with my dc motor before I got that dialed in better with software

So I’ve had a bit of time to tune mg2, and it’s getting close to drivable, I unfortunately have had one or 2 instances of unwanted regen and accelation, which have mostly been solved with tweaking syncofset (I had a 1000point wide deadband when foc tuning)

I’ve also run into what others have experienced with Prius motors and mgr’s in that torque falls flat past around 2500rpm, no amount of field weakening makes any difference, so I look forward to helping test alternative field weakening strategy

Side note for retro zero I have no issues paralleling the inverter vh pin into both vcu boards, no need to change scaling

Thanks for that

So I’m order to carry out more testing I needed to get the power steering and dash working without the original engine ecu, having discovered that I couldn’t use can messages straight from the open inverter board thanks to the way it sends 16 but values across 2 bytes being the opposite of what the car wants, I had to build a little arduino board to take messages from the vcu and on one can Id change the byte order and spit them back out on the id’s for the dash and power steering.

I also added a couple of small fets for driving things, such as switching on and off the auxiliary oil pump,

Placed it inside the case from the original ecu,

Ps yes I know the power supply is overkill but I had it in the spares box

Thanks to some other code I found online it also takes can messages from the abs wheel speed sensors and drives the speedo as this function was also done by the engine ecu

Nice. I got brakes working and bypassed converter. Still flat lines around 2500rpm..... Let's start testing...

Yep I should get to start tuning mg1 in the next week or so, I’ve been super busy at work