Tesla Model 3 Charger

[paaa]

Thanks for the info, I have one on the way from the US now1

[Jack Bauer]

Just an update on this I am currently investigatiing integrating pcs control into the ZombieVerter vcu project. Will update later in the week.

[Greenbeast]

"One board to rule them all"

[amzoo]

Does anyone here have the dimensions and weight of a model 3 PCS? I have an offer of one but need to work out if I can get it shipped from the US.

Took the raw dimensions of mine for future reference. Out of a 2018 Model 3 Long Range AWD LHD US

Weight 11.03 kg (24.3 lbs)
Bounding Box 30 x 80 x 10cm (11.8 x 31.5 x 3.9in)

Keep in mind there are two coolant ports at opposite corners that route vertically.

[amzoo]

First couple posts on this forum. I have a build using a full long range pack (2018 Model 3 US LHD AWD). Currently using the Model S HV components and an OrionBMS.

Jack, how can I help the development of the Model 3 PCS or the BMS? Semi-retired mechatronics engineer, so a lot of free time right now. I've been keeping up to date with both forum threads but wanted to ask what is directly needed at the moment.

[Jack Bauer]

The PCS(charger) is kinda something I have to do at this stage. BMS controller needs a full pack to test. works on my little bench pack but so far I've not been able to test with a full 4 modules.

[amzoo]

I don't have any spare Batman chips to harvest, but making a stop by my favorite dismantler this week to see if he'll sell me the boards by themselves. I'd be starting from scratch on the fpga board, so would be slowing you down more than anything there.

[Jack Bauer]

Well, Finally back to the pcs. Got it fully wired in to the E46 with the controller today. Now the fun part:)

[Jack Bauer]

So some good news on the pcs. DCDC stage is up and running in the E46. Work on the charge side progresses well. Killing off errors and getting closer...

[Jack Bauer]

Well this was fun. Did some software work on the eu pcs in my nice warm indoors and decided today to test it out on the us version in the E46. No can responses. Got the cover off and did some investigating. Yep you guessed it. ISO7741 blown, again. Luckily I had a spare so swapped it out on the car (tour de force) and can comms re established. Not that I would ever dare to critisise the big T but just what exactly was wrong with a can transciever Elon???

But its oookay because the good news is we can now set the target voltage for the DCDC and enable/disable at will. Tested working on both us and eu version. From what I can see I'm one error away from firing up the charger...

[Jack Bauer]

And now have the precharge from 12v function working:)

[fredd90]

Great progress, what is the reason why the digital isolator broke? Is it fed via external current and got a little sad when you fed it with 12v instead of 5v? =)

Have you seen any major differences on the eu / us version, apart from the feed of course

[Jack Bauer]

No idea why the isolator failed this time. Please see my earlier videos for indepth look at the pcs internals.

[Jack Bauer]

Major piece of PCS progress. Charger function now up and running:) Working the kinks out of the can messages and I still can't tell my left shift from my right shift as demonstrated by the 12v reading! Due to my evse only supplying 32amps on single phase I can't push the US version in the E46 to full 10kw just yet but I'll rig a temp 50amp supply just to prove it can be done.

Right now its using the prototype V1 pcs controller. As the firmware uses quite a bit of the work done by Matt and myself on the Model 3 charge port I'm not opensourcing at this time until we reach an agreement. Likewise I'm not sure on what direction to take on the hardware : standalone controller , integrate into m3 charge port / ccs controller , integrate into zombieverter vcu , integrate into gen3 controller etc.

Anyway, I will be doing a video soon showing the workings and describing the hardware side of things etc.

[Jack Bauer]

Bit more progress today. Auto start and stop charging working as well as dcdc activation while driving. Should have some hardware/software available for purchase in the next few weeks.

Plan is :

Hardware : Gerbers, pdf schematic and pdf board layout fully opensource. Design files and sources available on Patreon 10 Euros a month tier only at this time. Will be fully open at a future date. Both partial boards and full kits will be in the evbmw webshop with software preloaded.

Software : Johannes will be updating his new Tesla Model S Gen2/3 charger firmware for use with the PCS based on my arduino scribbles. This will be closed source for now with both Johannes and Matt receiving a % of the sales of boards and kits and also direct sales of the software to those who build their own hardware.

Fun incident from today. Of course this being Damien software I screwed up. The pcs charger power command is in a nice unsigned 16 bit int. 0 to 10000watts. yep you guessed it, I rolled it over to 65535 watts. I dove for the charging cable like a champion goalkeeper.

[Jack Bauer]

[collin80]

Given what day it is today I was hoping to hear about lucky charms or how magically delicious something is but instead I watched a 27 minute Irish video with none of that. But, it was good stuff nonetheless. Congrats on getting the PCS functional!

[Jack Bauer]

Thanks Colin:)

[Jack Bauer]

Sooo we had a problem today on a test drive in the Grey goose. DCDC dropped out after a few miles of driving and when I got home the charger would not kick in. The mini Elon brain told me I was being irrational again and should just take a stress pill and get some sleep. Instead I discovered that while all my driveway testing had been done the pack had been mostly fully charged at around 390v. This was very close to the fixed value I was using as the hv voltage reported by the high voltage controller. Of course when I went driving the hv measured by the pcs dropped lower but the fixed value stayed....well fixed and Elon was not happy. Two hours of bit shifting later I hammered the vcu measured hv into the muskian can frame and all was well in the world. Bonus feature : 12v derived precharge now works:) yay!

[mdrobnak]

Funny that.

How are you using the 12V powered precharge?

-Matt

[johu]

So if there is no second voltage measurement, is it possible to just loop back what the charger reports?

[Jack Bauer]

So to use 12v precharge :

-contactors must report open
-provide hv target voltage from the hvp
-enable dcdc hardware
-ask for precharge
-hvdc reported from the hvp must rise with the precharge voltage
-once target is hit close the contactors and switch dcdc to standby or support function.

Or just use a resistor:)

Yeah regards using the pcs measured hv as the hvp value I'm not sure. Only potential problem I see is the pcs won't start reporting the measured hv via can until its awake. Will test.

[Jack Bauer]

So design of new STM32F103 based controller is now done and I went full opensource on the release :
https://github.com/damienmaguire/Tesla-Model-3-Charger

Also uplaoded a redacted version of my hacky test code for the prototype controller.

Once Johannes gives me his blessing I'll order some prototypes from JLC and test in the E46. It is fully pin compatible with the prototype V1 controller and will use the Openinverter based charger firmware that Johannes authored for the Model S Gen2/3 chargers with a few tweaks for the pcs CAN and control lines.

[Jack Bauer]

You're welcome by the way.

[nkiernan]

Fantastic work JB, exciting update

Since your advice to try build around the Model 3 kit for my build on the NE course last August, have been amazed with your progress on the PCS and BMS, and what you and Matt are doing on the CCS, hats off.

Model 3 SR battery arrived last weekend and has the shorter battery modules to fit easier.

For your BMS work you say you're restricted without a full battery pack. Happy to try help if borrowing these is any use for a while?