Page 1 of 2

New Tesla charger board

Posted: Thu Apr 04, 2019 3:43 pm
by Jack Bauer
Just a little heads up on a new version of the Gen 2 and Gen 3 charger boards under development. They will feature the ability to switch the pilot signal between standard J1772 protocol and SWCAN. I was donated a Tesla wall connector that finally got installed last week. These can be set to "Tesla only" or legacy mode. A lot of destination chargers in hotels etc are "Tesla only" hence the SWCAN option.

Re: New Tesla charger board

Posted: Fri Apr 05, 2019 11:23 pm
by tom91
So they use SWCAN instead of the CP? That is interesting but then again quite a pain in the ass interms of being compatible.

These little boxes also want to know the vin of your car and name of your first born? ;)

Re: New Tesla charger board

Posted: Sat May 04, 2019 1:27 pm
by Jack Bauer
Last of the Gen2 logic boards before the upgrade just in.
New rev will include :

SW CAN for future compatibility with Tesla destination and supercharging and Tesla Chademo adapter. I'm ignoring any legal/moral issues here regards using Tesla charge points. Just making the hardware future proof.

Extra I/O for controlling HVJB.

Onboard Xilinx XC2C64A Coolrunner 2 CPLD for handling extra functions including a BIT (Built In Test) routine and shutdown / wake of the main 5v rail.

ESP32 WiFi header for future upgades. e.g. web interface.

DS1302 battery backed RTC (real time clock) for scheduling charging.

Onboard LEDs to indicate fault states.

Mini USB connector.

Expecting to release in around 6 weeks once tested. Fully backwards compatible with existing software.

Re: New Tesla charger board

Posted: Sat May 04, 2019 1:57 pm
by Kevin Sharpe
Jack Bauer wrote: Sat May 04, 2019 1:27 pmExtra I/O for controlling HVJB.
Can you bring out any spare I/O to an expansion connector? Might be useful for CCS or other interfaces in the future 8-)

Would it be possible to have the option of fitting through hole LED's as an option? Might be useful to have LED's that can poke through the top cover.

Re: New Tesla charger board

Posted: Wed May 08, 2019 11:36 am
by Kevin Sharpe
Looks like we need to get you access to an upgraded Model S/X so you can hack the FC CAN bus and support CHAdeMO and CCS rapid charging :D


Re: New Tesla charger board

Posted: Tue Jun 18, 2019 7:51 am
by Jack Bauer
Finally got some time to work on this. Here is a list of the changes from the V3:

Changed from Type B USB to Mini. This will allow the charger lid to close without modification.

Added extra stage of filtering to proximity signal. I have had one customer with a particular brand of EVSE that loves to blow the esd protection diodes in the SAM3 on the proximity line. We modded a V3 board to get around this so rolling this into the V4.

Single wire CAN capability added using the NCV7356. Now uses the exact same switching method between pwm control pilot and sw can as used in the Tesla charger / bms boards. This will allow use of Tesla destination charging, Chademo adapter,CCS adapter and possibly supercharging. All software dependent.

Added two high power DC low side contactor drivers for use with DC fast charge or as general outputs.


Added 3rd CAN channel based around the MCP2517 controller. Allows for high speed as well as FD Can. This extra can channel will allow the Tesla charger to act as a Chademo controller. Just add contactors and a hall current sensor.

Two extra filtered and protected 12v digital inputs. General use or chademo control.

One extra filtered and protected analog input. General use or hall current sensor for chademo control.

Expect to have the pcb mods finished by weekend and prototypes ordered.

Re: New Tesla charger board

Posted: Wed Jun 19, 2019 1:56 pm
by Jack Bauer
Design and pcb layout done. Schematic in PDF format up on Github :
https://github.com/damienmaguire/Tesla- ... ematic.pdf

Design files will be released as soon as I confirm pinouts on a few parts are correct. Particularly the USB socket:)

all comments / suggestions welcome.

Re: New Tesla charger board

Posted: Fri Jun 21, 2019 12:34 pm
by Jack Bauer
Naturally the library part for the USB connector had the pins assigned wrong. That would have been fun. PCBs ordered and design files released.
https://github.com/damienmaguire/Tesla- ... gicBoardV4

Re: New Tesla charger board

Posted: Thu Jul 11, 2019 10:00 pm
by Shaharov
Is there new software or still use Toms?

Re: New Tesla charger board

Posted: Fri Jul 12, 2019 9:12 am
by Jack Bauer
BOM now uploaded to Github for the V4 design.

The software is currently Tom's version with a few small mods I made for my own needs such as shutdown on reaching a setpoint voltage etc. Once I have some prototypes of the V4 then sw development can start. The V4 is designed for backwards compatability.

Re: New Tesla charger board

Posted: Fri Aug 09, 2019 7:29 am
by Jack Bauer
Prototypes have arrived.

Re: New Tesla charger board

Posted: Mon Sep 30, 2019 12:30 am
by Roadstercycle
Hi Jack, Looking to order a new board and wondering the time frame. Going out of town and don't want to miss delivery. Your sales web says 1 to 3 weeks depending on supplier. What are we looking at?

Re: New Tesla charger board

Posted: Mon Sep 30, 2019 7:15 am
by Jack Bauer
Would be nearer 3 weeks

Re: New Tesla charger board

Posted: Tue May 26, 2020 9:55 am
by Edison
Jack Bauer wrote: Thu Apr 04, 2019 3:43 pm Just a little heads up on a new version of the Gen 2 and Gen 3 charger boards under development. They will feature the ability to switch the pilot signal between standard J1772 protocol and SWCAN. I was donated a Tesla wall connector that finally got installed last week. These can be set to "Tesla only" or legacy mode. A lot of destination chargers in hotels etc are "Tesla only" hence the SWCAN option.
What was the progress on this?

Did anyone work out how to switch from J1772 to SWCAN?

Once in SWCAN mode, it looks like the car sends a single byte message every 100mS to control charging:

Code: Select all

 can0  30C   [1]  00
A payload of 00 disables charging. 01 enables it.

I also have decoded other messages where it sends the TWC serial number, model number, version. And where the car sends the VIN number, SoC etc.

However, what I haven't worked out yet is how to switch to SWCAN. Monitoring a Model 3, it appears comms start using J1772.

When the car is plugged in, the +ve voltage drops to 9V ('Vehicle detected'). After about 600mS the TWC starts generating the 1KHz PWM CP.

About 1.2 seconds later the +ve voltage drops to 6V ('Ready/charging').

Shortly after the comms switch to SWCAN (0 / 5V).

If I simulate the car myself (Diode, 2.74k // 1.3k) the TWC turns on, but stays in J1772 PWM mode.

The SWCAN circuits appears to have an input impedance of 3.5K, which in parallel with the above combination would put the +ve voltage at around 5V (or SWCAN levels). I tried switching this in, but no go.

Re: New Tesla charger board

Posted: Tue May 26, 2020 1:16 pm
by Jack Bauer
Progress was I reverse engineered the Tesla board cp input, implemented this on a board. Made the board open source :
https://github.com/damienmaguire/Tesla- ... gicBoardV4

Sold them for a while. Even offered free to anyone willing to do some software and of course no takers. So that board is no longer in production but you are welcome to get some made using the resources on github. Appreciate the share.

Re: New Tesla charger board

Posted: Wed May 27, 2020 9:38 am
by Edison
Thanks Damien.

I'll continue to see if I can work out how to enter SWCAN Mode.

Attached is some of the CAN IDs and packets I see between a TWC and a Model 3.

Re: New Tesla charger board

Posted: Tue Jun 02, 2020 10:37 am
by Edison
Quick progress update:
Over the weekend, I wrote some code to decode many of the SWCAN messages:
https://github.com/craigpeacock/Tesla-C ... rt-Decoder

Just need to work out now how to get a TWC to switch to SWCAN mode.

Re: New Tesla charger board

Posted: Tue Jun 02, 2020 10:55 am
by Jack Bauer
Thanks for the update. I may consider adding it to the new v5 board.

Re: New Tesla charger board

Posted: Tue Jun 02, 2020 3:50 pm
by Kevin Sharpe
Edison wrote: Tue Jun 02, 2020 10:37 am Just need to work out now how to get a TWC to switch to SWCAN mode.
iirc one of the DIP switches inside the TWC selects between J1772 and SWCAN mode at power up. What are your switch settings?

Edit: here's a reference to the second DIP switch setting. Also note comments re timing;

https://teslamotorsclub.com/tmc/threads ... st-3698196

Re: New Tesla charger board

Posted: Wed Jun 03, 2020 8:57 am
by Edison
Kevin,

From what I can understand, different firmware on the TWC behave in different ways. I'm running what I believe is the latest (for Gen 2) 4.5.3. In 4.5.3, the DIP switch doesn't appear to do anything. With an oscilloscope on the CP when a Tesla is connected, shows it talks in SWCAN mode regardless of what position the DIP switch is in.

I'll see if I can get a Oscilloscope trace and post it, but as per my post above:

Monitoring a Model 3, it appears comms start using J1772.

When the car is plugged in, the +ve voltage drops to 9V ('Vehicle detected'). After about 600mS the TWC starts generating the 1KHz PWM CP.

About 1.2 seconds later the +ve voltage drops to 6V ('Ready/charging').

Shortly after the comms switch to SWCAN (0 / 5V).

If anyone has a TWC running a firmware version earlier than 4.5.3, I would be interested to hear how the SWCAN comms works in relation to the DIP switch.

Re: New Tesla charger board

Posted: Wed Jun 03, 2020 10:40 am
by Kevin Sharpe
Edison wrote: Wed Jun 03, 2020 8:57 am From what I can understand, different firmware on the TWC behave in different ways.
That would make sense, I know a lot of commercial sites that deployed TWC were unhappy with the incompatibility with non Tesla cars.

Re: New Tesla charger board

Posted: Sat Jun 06, 2020 2:38 am
by Edison
Kevin Sharpe wrote: Wed Jun 03, 2020 10:40 am That would make sense, I know a lot of commercial sites that deployed TWC were unhappy with the incompatibility with non Tesla cars.
Yes, I think the intent is to make it compatible, but practice is a different thing. My father has a Hyundai Kona - It refuses to charge from my TWC running 4.5.3. However, it will charge from older TWC, sometimes you just need to wait 30 seconds before SWCAN times out, and it switches to J1772.

Re: New Tesla charger board

Posted: Sat Jun 06, 2020 2:58 am
by Edison
Edison wrote: Wed Jun 03, 2020 8:57 am I'll see if I can get a Oscilloscope trace and post it, but as per my post above:
As promised, this is what a charging session between a Model 3 and TWC 4.5.3 looks like:
6sec.png
6sec.png (6.35 KiB) Viewed 6685 times
  • Charger idle @ 12V.
  • Plug Inserted @ approx 700mS. Signal pulled down to 9V by car's 2.74k resistor ('Vehicle detected').
  • Approx 1.3s. TWC starts generating 1KHz CP signal.
  • Approx 2.6s comms switch to SWCAN
Zoomed_out_transition.png
Zoomed_out_transition.png (5.16 KiB) Viewed 6685 times
  • First 1KHz pulses have positive peaks at 9V ('Vehicle detected')
  • Car then switches in 1.3K. Peaks drop to 6V ('Ready/charging')
  • 11 pulses later, switches to SWCAN
Zoomed_in_transition.png
Zoomed_in_transition.png (4.56 KiB) Viewed 6685 times
  • It appears line is released (not driven by TWC) shortly after 11th pulse.
  • SWCAN starts shortly after. Its not yet clear which device starts talking first.
What is also interesting, is the TWC was configured for 8A during this session. The 1KHz PWM signal was 50uS or 5% duty cycle (Approx 3A).

I wonder if this is why some third party vehicles refuse to charge.

More to come....

Re: New Tesla charger board

Posted: Sat Jun 06, 2020 6:02 am
by Kevin Sharpe
Edison wrote: Sat Jun 06, 2020 2:58 am What is also interesting, is the TWC was configured for 8A during this session. The 1KHz PWM signal was 50uS or 5% duty cycle (Approx 3A).
5% duty cycle indicates digital communications. Here's an extract from IEC61851-1{ed2.0}
61851_A5_A6.png
Edison wrote: Sat Jun 06, 2020 2:58 am I wonder if this is why some third party vehicles refuse to charge.
It's possible that cars have bugs or that Tesla are doing something strange in the TWC. Certainly early Model S had issues when the EVSE commanded 6A.

Re: New Tesla charger board

Posted: Sat Jun 06, 2020 7:30 am
by Edison
Kevin Sharpe wrote: Sat Jun 06, 2020 6:02 am 5% duty cycle indicates digital communications. Here's an extract from IEC61851-1{ed2.0}
Thanks.

What does IEC61851 define as digital comms? I believe it is ISO 15118-3 HomePlug Green PHY for talking to DC fast chargers.