Prius Gen 3 Inverter Logic Board Support Thread

[FFMan]

OK - just making a start on this. Got the V1c board from damien, soldered up the connectors and powered it up, got flashing led and web interface is up - all good so far. So i installed the board in my gen 3 inverter as per damiens 3 part video complete with gorilla glue option

I ran a few tests (connecting mg2 fwd and start to 12v) and was getting two errors, the tmpsmax one (maybe temp sensor type not set correctly) and the precharge fail. not surprising as there was no HV connected. So not wanting to dive into connecting the HV just yet, i got to watching a latuta on you tube, i liked his small steps approach, so watched him load stm32_sin to test the pwm with no HV etc. When i loaded this firmware, the blinking led no longer blinks. I tried reverting to foc (I guess this is what damien loaded before he sent it to me), but still no led blink.

I note that I appear to have no parameter list, and when i download the json file it is empty.

Question, does the firmware upload wipe the parameters and if so where do i find the relevant parameters to start with. I found the 'database' of lists but could work out which was a good match. Tried to load the set i thought was closest, but whilst it appears to load, when i then request download from the board, the json file is empty, so i guess it's not loading correctly.

any help much appreciated - my wife will murder me if that 3 series touring we bought for the project doesn't find a use some time !

[chrskly]

OK - just making a start on this. Got the V1c board from damien, soldered up the connectors and powered it up, got flashing led and web interface is up - all good so far. So i installed the board in my gen 3 inverter as per damiens 3 part video complete with gorilla glue option

I ran a few tests (connecting mg2 fwd and start to 12v) and was getting two errors, the tmpsmax one (maybe temp sensor type not set correctly) and the precharge fail. not surprising as there was no HV connected. So not wanting to dive into connecting the HV just yet, i got to watching a latuta on you tube, i liked his small steps approach, so watched him load stm32_sin to test the pwm with no HV etc. When i loaded this firmware, the blinking led no longer blinks. I tried reverting to foc (I guess this is what damien loaded before he sent it to me), but still no led blink.

I note that I appear to have no parameter list, and when i download the json file it is empty.

Question, does the firmware upload wipe the parameters and if so where do i find the relevant parameters to start with. I found the 'database' of lists but could work out which was a good match. Tried to load the set i thought was closest, but whilst it appears to load, when i then request download from the board, the json file is empty, so i guess it's not loading correctly.

any help much appreciated - my wife will murder me if that 3 series touring we bought for the project doesn't find a use some time !

If you have no LED blink and no params, then there's something wrong with the firmware. If you can't get it to reprogram through the web interface, then you can try reprogramming through and stlink.

[FFMan]

thanks Christian - after some forum digging it seems i'm not alone and sometime the OTA load doesn't work as intended.

So i spent an hour or so getting stlink to work on the PC (mfc140.dll issues) and this morning have reprogrammed the stm32 with the bootloader and the foc firmware and all that seemd to work ok.

However, still no flashing led, though it was working before i starting any firmware loads so unlikely to be a hardware issue i hope.

What can i try next ?

[chrskly]

thanks Christian - after some forum digging it seems i'm not alone and sometime the OTA load doesn't work as intended.

So i spent an hour or so getting stlink to work on the PC (mfc140.dll issues) and this morning have reprogrammed the stm32 with the bootloader and the foc firmware and all that seemd to work ok.

However, still no flashing led, though it was working before i starting any firmware loads so unlikely to be a hardware issue i hope.

What can i try next ?

Sounds like the programming is not working to me. AFAIK if you don't have a flashing LED the programming didn't work. Are you following the steps on the wiki? https://openinverter.org/wiki/Toyota_Prius_Gen3_Board

[FFMan]

https://www.dropbox.com/s/y21rr7p5tf7g1 ... s.PNG?dl=0

yes following those instructions and programming reports as having worked for both bootloader and foc/sine fw

i checked all the stlink settings as per the article and they all match up.

does the blinking led come on any other time - what if the led isn't now working - can i verify whether the firmware is running some other way though i think i'm clutching at straws here.

the web interface still has no parameters.

i checked all the voltages as per the article and they're all there too.

[Bigpie]

If you attach a serial cable to the rx and tx instead if the wifi module, you can see if the bootloader is running, itll print out 2F a drain number of times. If not the bootloader isnt running, if it keeps repeating indefinitely the firmware isnt running and its staying in the bootloader. Make sure it's a 3.3v usb to serial.

[FFMan]

ok - so for clarity, this is what I have tried:-

configured my usb-serial to 3.3v, run up realterm and flashed the rx line to ground and i get spurious characters showing in realterm and led flashing on my converter - crudely kinda proves it works and i've identifed the rx ok.

removed the wifi module and hook my usb to serial into the wifi header pin 2 (Gnd) and 3 (TX) to my rx. To check i've read the schematic right (I'm more used to building my own stuff rather than working with someone elses) if you hold the board up with the idc connector on the lower edge, the headers pins

1 6
2 7
3 8
4 9
5 10

when i power up the V1c nothing is received, tried a few speeds but rx light isn't flashing and no garbage (which i'd expect at some speeds even if not the correct one). i also tried connecting my rx to pin 4 (rx) just in case as rx/tx is a matter of perspective and seeing as i set my converter to 3,3v no harm could some.

firstly, is that the way to run the test ?

If i'm getting nothing out, what does that suggest ?

appreciate the help

[FFMan]

tried erase chip and reload of bootloader and fw but still the same. I can see the stlink chatting away to the STM and it loads ok it just doesn't seem to run. Does this mean the stm is ok, but its a s/w issue ?

What blockers are there to the led blinking. Is this a very basic heartbeat, or does it only blink if some checks are passed ?

[FFMan]

I'm not quite sure where to go with this now, i have a board which was functioning at first, but when i attempted to update the fireware has since refused to do anything useful although it appears to accept the firmware load.

Is there anyone willing to take on some paid support,or is it somehow bricked and i either need to purchase a replacement or accept defeat. If i'd done something silly to damage the board, i'd shrug and purchase another, but without clear indication as to what has happened i'm a little less bullish about this approach though still quite invested in building my open EV.

is it possible/likely the stm32 is damaged or locked somehow, and would replacing that IC (assuming i can get one) worth a try, or does the fact that it accepts a download not indicate this ?

any help much appreciated

Alex

[LeonB]

is it possible/likely the stm32 is damaged or locked somehow, and would replacing that IC (assuming i can get one) worth a try, or does the fact that it accepts a download not indicate this ?

I’m pretty sure your STM32 is ok. When you reprogrammed it with stlink, are you using .bin or .hex files? .hex is fairly straightforward but with .bin you have to use the correct start address.

[FFMan]

i tend to think the stm is ok too.

You've made me question something. when i load the boot loader it loads at address 0x08000000 (as per the wiki instructions) and when i load the sine or foc firmware i guess i need to specify a different address. The wiki instructions suggesting loading the main f/w over the wifi, but as this seemed to cause the issue, i was trying to load the f/w via the stlink, but i didn't change the load address. would this cause the f/w to overwrite the bootloader ?

Should i specify a different load address for the main f/w load ?

Is loading the hex file easier or likely to be more reliable/less error prone ?

thanks

[LeonB]

but i didn't change the load address. would this cause the f/w to overwrite the bootloader ?

Yes, that is probably your problem. Just use the .hex files and you should see the LED blinking again

[Dylan Witt]

Question, why are we only using the buck boost input for the gen three inverter for power delivery? I mean the connection side that is posted in this forum, the main power connection for the MG2 is pictured, I was using my kia hybrid starter generator to trace out the power connections, and found that these two connections are the only ones connection directly to mg2 igbt, with the capacitor installed we are running power though the tiny copper busbars on the caps, my question, why not go directly here and still keep the factory caps connected but now we can actually push some power though the inverter, my testing using a 12volt battery load tester and using the web interface to monitor amps, at 400rpm Turning the generator with a drill I'm getting 34.2 amps before I'm having to use my feet to hold everything from spinning. This Busbar connection pictured. I'm going to end up using this side of the inverter for my main drive motor power delivery. But with the factory caps connected

[FFMan]

Yes, that is probably your problem. Just use the .hex files and you should see the LED blinking again

I see now the hex files have the address specified in them, and the boot loader is at 0x08000000 and the f/w at 0x8001000. Certainly when loading the binary i did not change the address so this was for sure an issue.

However, now using the hex files and I did not restore my blinking led when i try and program via STLINK. I removed the wifi module for this due to comment in the wiki about the the LDO.

I tried loading the boot loader via stlink and the main f/w via OTA as per the wiki instructions but the connection keeps dropping . The wiki mentions stengthening the regulator for the wifi modukle but the comment is :- TODO: does this also affect the latest revision boards? so not sure if this needs doing though these symptoms would suggest it still does. I have ordered the part shown i the wiki, 800mah @ 3.3v

However, LDO aside, loading hex files via stlink should work shouldn't it and stlink reports successful loads and handshakes happily continuously with the stm32.

Any other thoughts or things i can try. this isn't making much sense at the moment.

thanks

[FFMan]

Just looking around at options and in STLINK>Core>MCU i get the below. Does it matter it says 'core halted' ?

https://1drv.ms/u/s!Al_kMZ0uQpMxg-EZ9pi ... g?e=WNXeCS

does anyone have one to compare with. Clicking run etc doesn't produce the flashing LED BTW.

Also my option bytes look like this if anyone can verify.

https://1drv.ms/u/s!Al_kMZ0uQpMxg-Ea-87 ... Q?e=tfV5uP

thanks

[4markowen]

Just double checking before I start soldering the tricky little chips that came with the kit.

IC14 being the hardest to see, but there appears to be a notches to how I've orientated and placed it?

IC12 looks more straightforward and was confirmed earlier, but just wanted to double check.

Check twice, solder a few times, swear a lot

Here's to ruining my eyesight and making sure I don't inhale those solder fumes...

[FFMan]

Just looking around at options and in STLINK>Core>MCU i get the below. Does it matter it says 'core halted' ?

https://1drv.ms/u/s!Al_kMZ0uQpMxg-EZ9pi ... g?e=WNXeCS

does anyone have one to compare with. Clicking run etc doesn't produce the flashing LED BTW.

Also my option bytes look like this if anyone can verify.

https://1drv.ms/u/s!Al_kMZ0uQpMxg-Ea-87 ... Q?e=tfV5uP

thanks

Seem to have reached a bit of a dead end here, can only imagine the STM is fubar is some way so will seek out a new chip (not that i can find any reliable stock at the moment) or an entire PCB, either way I give up on this one, it loads but doesn't appear to actually run. If anyone in the UK fancies a go at it, let me know. cheers

[Bigpie]

Where abouts in uk are you? I'm in rotherham if you want me to take a look.

[FFMan]

Where abouts in uk are you? I'm in rotherham if you want me to take a look.

Thank you - that would be useful before i go to the expense of replacing it. I'm in Bicester but can post it to you. Another set of eyes on it would be helpful in case i am being a muppet somewhere.

Can you PM me an address and I'll find some sensible packaging. I'll probably dpd it given the post is intermittent, so if you cold include a mobile phone no that is helpful as they'll text when they will deliver etc.

Alex

[Bigpie]

@FFMan, pmd you my address.

Anyone know about the temphs? Chasing my cutouts I noticed the temperature feedback seems to have a ceiling of 100 degrees

Graph shows where inverter shutdown current and fstat tanks as auto resume

[Jack Bauer]

The inverter will not cutout on overtemp but it will limit. The only source for a cutout is the IGBT desat detector outputs.

[Bigpie]

It looks to have cutout at lower currents on the graph, unless something spiked and Mr Toyota said now before Johannes gets the info.
2.5 Throtcur, if I'm reading this right, it's happy at about 250 amps then at 380Hz ish Mr Toyota sits down.

4.5 Throtcur, current 100 amps higher, but again at 380Hz ish Mr Toyota sits down.

[Bigpie]

So wondering if it's DC max current? Anyone able to post some graphs of idc, iq, id and stats? Are you able to get over 100 amps DC?

[johu]

I don't think the inverter even knows the DC current. Rather looks like some event that happens at full throttle and 380Hz. Could be the point where the amplitude is so high that the IGBTs can stop switching for parts of the period. Actually that's a good thing because it reduces switching losses. But perhaps Mr. Toyota don't like it?

You could plot "amp" to see that

[Bigpie]

I shall plot that next time.

Setting to derate at 100 amps has stopped cut outs for the most part. Stamping on it from 0 mph will still do so, but I can now floor it after rolling a little. Removed fmax and can now get the rpm back up high again. It's even allowed me to bump the throtcur further. Pretty pleased.