ZombieVerter VCU Support

[86elcamino]

Thanks Damien, I since found a (really good) schematic with all the pieces buried in the thread, albeit for the old version of VCU. And indeed, we don't have the shunt...

If anyone can access the wiki page for the GS450 and add this pdf to it, it could save newcomers an awful lot of time! I can't log in there, the sign up link is broken.

viewtopic.php?p=12105#p12105

I'm almost ready to start wiring up my GS450 project. Any help with a wiring diagram and explanations of the functions of the various inputs and outputs in layman terms would be great.

Thanks to everyone that's putting in so much time on this project.

[Bratitude]

If anyone can access the wiki page for the GS450 and add this pdf to it, it could save newcomers an awful lot of time! I can't log in there, the sign up link is broken.

viewtopic.php?p=12105#p12105

Just sign in with your openinverter credentials

[et0]

If anyone can access the wiki page for the GS450 and add this pdf to it, it could save newcomers an awful lot of time! I can't log in there, the sign up link is broken.

viewtopic.php?p=12105#p12105

Just sign in with your openinverter credentials

That wasn't working but I tried again after resetting my password and it did!

I've added the schematic to the old VCU page, and a mention of the required shunt as well.

I'm almost ready to start wiring up my GS450 project. Any help with a wiring diagram and explanations of the functions of the various inputs and outputs in layman terms would be great.

We have a system all connected up on the bench just now, so I will update the old VCU schematic to Zombieverter and post that once I've got it working and confirmed some of the guesses / cribbing from other sources is correct. You're right, it's pretty baffling to come at it fresh!

[et0]

OK, some progress but stuck again - appreciate any suggestions.

Shunt is connected up to CAN EXT.

“Start inverter in manual mode” produces a message “unknown command sequence”. What should it do?

Pulsing the start signal operates the precharge relay, which stays on for about 5 seconds. However the main relay doesn’t close and there is a message “STOP-PRECHARGE”

Supply voltage is 65, all minimums are set to 45.

Here’s the plot of that happening. INVudc shows up (that’s read from the inverter?) but not udc.

And the parameters I have set:

params.json

(637 Bytes) Downloaded 216 times

What should I be looking at to debug this, is there still a serial output?

[et0]

Here’s the plot of that happening. INVudc shows up (that’s read from the inverter?) but not udc.

Looking at the code, seems a reasonable guess that it's not communicating with the shunt, as that's where udc comes from. So maybe the shunt is not initialized. I can see code for that in the project but it's commented out. Should I go down the rabbit hole and try to build the source and call that initialize function or is there an easier way?

What should I be looking at to debug this, is there still a serial output?

I see parts of a serial interface in there but it doesn't look complete? How are people typically working with this HW? It would be great to have some basic live info like whether comms with the various modules is established, view of the states and variables etc.

Lots of questions into the void...

[Jack Bauer]

Isa shunt should be connected to can ext 2 (this is the ev can) can ext is vehicle can. If your shunt is new then yes it needs to be initialised. I'll add a menu option to do this.

[et0]

Jeez-o, what was the reason for changing away from Arduino again?
Some firmware building breadcrumbs to maybe help the next person:

Code: Select all

You will need the arm-none-eabi toolchain: https://developer.arm.com/open-source/gnu-toolchain/gnu-rm/downloads The only external depedencies are libopencm3 and libopeninv. You can download and build this dependency by typing

make get-deps

Now you can compile stm32-vcu by typing

make

But you also need Gnu Make, as it's not included in the gnu toolchain (!?)
To get that you need something called "chocolatey" https://chocolatey.org/install and then do 'choco install make'

You also need Python. Just typing "Python" in the shell went and opened the windows shop and got version 3.9 for me, but there are other ways to get it.

edit: I needed to use the bash shell (installed with GIT?) to run make

Then do the above steps and it should generate a .hex and a .bin, the .bin can be loaded via the web interface under "UART Update".

[et0]

Isa shunt should be connected to can ext 2 (this is the ev can) can ext is vehicle can.

OK.

If your shunt is new then yes it needs to be initialised. I'll add a menu option to do this.

Thanks Damien!

[et0]

Here's a list of connections and where they go to, or at least my best guesses. I've added it to the wiki since it's better than nothing, but will update as I learn of further errors.

[Jack Bauer]

So to refresh some memories the idea of this project , and the move away from the arduino platform, was to build on the work of Johannes and others on the open inverter project to build a VCU that would ultimatley remove the need for people to do such things as compile software, figure out how to load it etc. Rather just connect some wires, select a few menu items from the web interface, calibrate a throttle and go driving. Now we are not at this stage yet BUT are not far from it. I have now 3 cars running on the release hardware and have accumulated close on 10k miles.

My software skills are terrible. Thanks to others and most recently Johannes taking on the task of cleaning up my mess and making it into a properly formatted C++ project we are at a point that people will not have to go through this pain. Now, if folks with the skills want to get into the programming and make changes then that's great also and is one reason I make things opensource.

The chip shortage has hit hard but I am still making kits available despite a lot of extra work needed on my part nor am i increasing prices. Its the most I can do to move the project into more people's hands.

We do need documentation. I know. Its not a fun or sexy job and I REALLY appreciate the efforts being made. I drove 140 miles today in the grey goose. CCS fast charged 3 times. Ran a Gen 1 leaf inverter, Model 3 pcs, Ampera heater, isa shunt etc etc. So I do believe the blocks are there, the structure is well on the way now thanks to Johannes (oh yeah he's also adding regen!). Once the chips shortage is no more I'll be offering fully boards again.

I know there are lots of questions. Lots of things people want. Lots of things people don't like or want. Thats cool. I'll acomodate as much as I can when I can.

Two things that are on the way are 1)Zombie expander (can based i/o expander with BIG power capabilities) and 2)First commercial application. Number 2 allows me to pump some revenue in and get a test rig built. Then fully tested Zombies will be available.

[et0]

Everyone appreciates your efforts Damien, your work has made a lot of projects possible. Thank you! And also to Johannes.

And I am sure the shift to the new platform makes sense in the wider view. It just seems rather impenetrable at the stage it's at now. I am especially missing an obvious debug facility: even when the project is 100% bug free, you're always going to get setup issues with bad wiring or broken / wrong version external modules and a way to see what's going in real time is very useful.
The web interface feels like wearing boxing gloves to do SMT soldering to me, but that's probably just showing my age...

[86elcamino]

Here's a list of connections and where they go to, or at least my best guesses. I've added it to the wiki since it's better than nothing, but will update as I learn of further errors.

zomb-con-et.png

This is great! Hope that the other connections can be filled in soon, along with maybe some definitions of the abbreviations, and for what purpose the others may be used.

[EV_Builder]

Question: Where can i find the WebGUI source/Firmware for the VCU v1?

[Jack Bauer]

https://github.com/jsphuebner/esp8266-web-interface

[et0]

With some pressure to get this up and running, I tried to get the shunt initialization done.

I was going to add a terminal command to do it, but it doesn't look like the terminal is working, maybe the UART or pins aren't configured?

So then I added it in the main function and tried to upload that using the web interface, but unfortunately (or inevitably) that has bricked the VCU. The web interface is still there but nothing coming from the VCU.

Help! How can I reflash it?

[Jack Bauer]

So its quite hard to actually "brick" the stm32 as the bootloader is not overwritten by a firmware update. Just power cycle the vcu, reconnect to the wifi and upload again.

[et0]

So its quite hard to actually "brick" the stm32 as the bootloader is not overwritten by a firmware update. Just power cycle the vcu, reconnect to the wifi and upload again.

I tried that 3 or 4 times. The upload progress bar doesn't appear, as it did the first time.
Should I press "erase flash" or anything else?

[Jack Bauer]

Sounds like the partial upload is blocking the bootloader from recycling. Power up the vcu, connect to the wifi, select the bin fine but dont hit upload. Locate C12 on the board. Use a small tweezers to short it out. Hit upload then release the short and the bootloader should fire up and catch the upload.

[et0]

I'll try it tomorrow. What about the terminal, should it be working or no?

[et0]

OK, the tweezer trick worked. Assuming that's the reset line for the STM, I might solder a button on there.

And the shunt appears to be programmed (it's reporting udc at least), and the VCU seems to be back to life.

Problems I'm having:
1) InverterPowerLSSW does not pull in the relay to turn on the inverter, so I bypassed it for now. Can hear the relay buzzing slightly.
2) tmpm was 6400, which produced an error. I zeroed it out in the code, and the error has gone, but that seems like an issue.
3) Now I have "run" and no error, forward selected etc, pot signal is good, 0.6A pulled from the HV supply but no sign of the motor turning.
4) no terminal still

What should I look at next to get the motor spinning please?

[Jack Bauer]

Due to an "upgrade" on Olimex wifi modules I can no longer supply them as open networks. By default all vcu kits will have SSID : inverter PASSWORD : inverter123

If someone could please update the wiki with this would be much appreciated.

[Jack Bauer]

Please don't bypass components like that. If the relay does not turn on its for a reason. What motor/inverter setup are you using? Are you seeing UDC at the correct voltage? have you set udcsw and udcmin to sane values and calibrated the throttle? Is there a direction selected and T15 on? I feel the need for a setup video ...

[et0]

Lexus GS450h as far as I know.
Inverter number G9200-30030 A06K06Y023
Can't see a number on the transmission.

Yes, UDC is showing 65V after starting, so is INVudc. udcsw is set to 45 as is udcmin.
Yes I have calibrated the throttle (set min just below the lowest value and max just above the highest value) and can see the pot input moving correctly.
As I said, Forward is selected (and showing on the GUI) and T15 is on. Start has been triggered and pre-charge happened OK.
All this can be seen on the screenshot I think?

Regarding setup instructions, I will be more than happy to write those if you can spare a minute to debug this setup with me.

[et0]

Due to an "upgrade" on Olimex wifi modules I can no longer supply them as open networks. By default all vcu kits will have SSID : inverter PASSWORD : inverter123

If someone could please update the wiki with this would be much appreciated.

Done.

[Jack Bauer]

It's been a bit since I messed about with the lexus stuff but sounds like you are good to go. Data exchange is happening with the inverter. 65v is more then enough to spin up the motors. What is invstat parameter? When you press throttle is potnom and torque registering a change?