Hacking into Zoe's CANbus

Zubehör für den Renault ZOE
AntwortenAntworten Options Options Arrow

Re: Hacking into Zoe's CANbus

michaell
  • Beiträge: 1499
  • Registriert: Fr 28. Nov 2014, 16:26
  • Wohnort: 90765 Fürth
  • Hat sich bedankt: 1 Mal
  • Danke erhalten: 20 Mal
  • Website
read
Zoelibat hat geschrieben:Was perfekt wäre für eine zusätzliche Anzeige wäre der Platz unter dem Tachodisplay.
oder in die Frontscheibe einblenden lassen
so als Alternative
ZOE intens seit 11/2014, Peugeot seit 11/2020
KONA seit 12/2018, 7/2021 Ioniq 5 mit P45 Paket, Enyaq 80 01/2022
Info (at) Solar-lamprecht.de
Anzeige

Re: Hacking into Zoe's CANbus

USER_AVATAR
  • Zoelibat
  • Beiträge: 4337
  • Registriert: Mi 27. Aug 2014, 23:03
  • Wohnort: Zoe (Rückbank)
  • Hat sich bedankt: 268 Mal
  • Danke erhalten: 210 Mal
read
HUD hatte ich mir auch schon gedacht, aber so einfach wird das nicht sein.
Schon gar nicht weil immer die Armaturenbrett-Spiegelungen die Anzeigen überdecken. :-D

Was anderes: Warum zeigt CanZE bei einer 11kW Ladestation nur ca. 8,5 kW mögliche Ladeleistung an?
Habe ich sowohl an Typ2 als auch an CEE.
Beim NrgKick ist es ähnlich (dort fast 9kW).
Sollten es nicht 10-10,5 kW sein?

Bei 43kW sind es ca. 41kW mögliche Ladeleistung, bei 22kW sind es ca. 21kW. Das passt nach meinem Verständnis eigentlich. Nur bei 11kW Stationen sind es 8,5kW. Warum?

Re: Hacking into Zoe's CANbus

michaell
  • Beiträge: 1499
  • Registriert: Fr 28. Nov 2014, 16:26
  • Wohnort: 90765 Fürth
  • Hat sich bedankt: 1 Mal
  • Danke erhalten: 20 Mal
  • Website
read
bei mir hat es heute an einer 22 kW Ladestation auch nur 20,4 angezeigt
ZOE intens seit 11/2014, Peugeot seit 11/2020
KONA seit 12/2018, 7/2021 Ioniq 5 mit P45 Paket, Enyaq 80 01/2022
Info (at) Solar-lamprecht.de

Re: Hacking into Zoe's CANbus

lingley
read
Zoelibat hat geschrieben:Nur bei 11kW Stationen sind es 8,5kW. Warum?
11kW.jpg
cosphi.jpg

Re: Hacking into Zoe's CANbus

jmeijer
  • Beiträge: 264
  • Registriert: Mo 2. Sep 2013, 15:08
  • Hat sich bedankt: 30 Mal
  • Danke erhalten: 40 Mal
read
Probably:
1) because of the relatively high current, the voltage at the end of the line is closer to 220 than 230 (-5%);
2) because the car is taking slightly below 16A (as lingley shows above, another -5%);
3) because of a shifting cos phi (phase shift between voltage and current). It gets worse with lower power ratings, I will have to look up on my charger how much that is, but I can only do it for 1 phase). Edit: see below.
4) because the car might report it in final DC power. I would need to dig up some charging graphs to be sure about that.

1, 2 and 3 are fact, 4 is a guess that need to be confirmed / denied / quantified.

@lingley: I think you added the Power Factor (cos phi) later. And there you have it, a whopping 17% is gone there. This is not lost as in "up in smoke": it is not measured by your power meter. But it is eating into the charging speed.
Jeroen - ZOE Q210 Nov 2013 - PV powered - CanZE co-developer http://canze.fisch.lu

Re: Hacking into Zoe's CANbus

USER_AVATAR
  • AbRiNgOi
  • Beiträge: 15206
  • Registriert: Do 27. Jun 2013, 17:43
  • Wohnort: Guntramsdorf (bei Wien)
  • Hat sich bedankt: 560 Mal
  • Danke erhalten: 4653 Mal
read
Some status report.

I am still on the topic to bring the CanZE Software to twizplay. What is working now:
I changed to C++: display, InOut-ports, CAN (reading) is working again fine. This weekend I tried the transport protocol, so I started to implement a CAN transmit object (request object). I received now a Single frame on Transport protocol, means only 8 Byte on transport protocol is working. For more I need to implement a Flow control, there is still a problem, so I get the first 2 cell voltage, but than the Flow control should request the next 94 cell voltages... It is something easy like "0x30 0x00 0x00" or "0x30 0x01 0x14"... but twizplay is not requesting, or ZOE is not listening, or the bytes are wrong, or the timing is wrong,... I will find :)
Yesterday I stopped trying, not to discharge my three year old 12V Battery to much.. I tried the whole day without driving my ZOE...

so long.

(PS: why C++? because I like it more than Basic and you can program it directly with Atmel Studio 7.0 without extra payment for the C++ compiler and you can program the device directly AND debug....)
ZOE Live Q210 6/2013 * AHK legal Typisiert 18.07.2017 * 40kWh Batterie 12.03.2019
Aktuell: 149.000 km

Niemand ist bei mir auf der Ignor-Liste!

Re: Hacking into Zoe's CANbus

jmeijer
  • Beiträge: 264
  • Registriert: Mo 2. Sep 2013, 15:08
  • Hat sich bedankt: 30 Mal
  • Danke erhalten: 40 Mal
read
Cool!

On TP: indeed if the first frame you receive starts with 0x1, you should send a flow control frame. The format is:

0x30 0xaa 0xbb

where aa is the number of frames to receive, 0x00 for "all remaining", and bb is the number of milliseconds to wait between the frames, should be <= 127

I assume you want all the remaining frames to be send, so something like 0x30 0x00 0x14 is correct; wait 20 ms between frames. If you send bb as 0x00 you will get full speed, which maybe more than the ATMega can handle.

If aa > 0 the meaning of bb is slightly different. It should be 0xf1 - 0xf9, meaning 100 - 900 microseconds, but I never use aa > 0.

Here is a good description: https://en.wikipedia.org/wiki/ISO_15765-2
Jeroen - ZOE Q210 Nov 2013 - PV powered - CanZE co-developer http://canze.fisch.lu

Re: Hacking into Zoe's CANbus

USER_AVATAR
  • AbRiNgOi
  • Beiträge: 15206
  • Registriert: Do 27. Jun 2013, 17:43
  • Wohnort: Guntramsdorf (bei Wien)
  • Hat sich bedankt: 560 Mal
  • Danke erhalten: 4653 Mal
read
Yes, I got a 0x10 Message, that means "First Frame".
DSC_0145.JPG
For testing I send the request frame only after pressing the button 1 on Twizplay and the counter goes up, then after the first frame automatically the Flow control should come (no testing if it is a not response by now...). But the mirror of the send frame still stays at zero, and the counter stays 1. Setting a while-loop on TXOK is never ending... So I think the Flow Control is not send at all. That is what I do not understand, because on my Desk it sends normal (no never ending while on TXOK), in the car it doesn't...

PS: everyone who can read hex sees, my Battery is full charged... 4,102V on cell one and two.

PPS: does anybody know, what is the cell voltage at 0% SOC???
ZOE Live Q210 6/2013 * AHK legal Typisiert 18.07.2017 * 40kWh Batterie 12.03.2019
Aktuell: 149.000 km

Niemand ist bei mir auf der Ignor-Liste!

Re: Hacking into Zoe's CANbus

jmeijer
  • Beiträge: 264
  • Registriert: Mo 2. Sep 2013, 15:08
  • Hat sich bedankt: 30 Mal
  • Danke erhalten: 40 Mal
read
Nitpick (I know I am horribe): 0x1 means First, the next nibble is already the size, so your message size is 0x07e. But don't worry, I don't think I have seen messages larger than 255 bytes, so in practice, 0x10 is correct.

I have never programmed the Atmel CANbus controller on AVR's, and on the Due I used a library (build by Collin Kidder, great work). I do know that one of the CanZE programmers had a problem sending frames on a Teensy (whick is more a Due than an AVR, but still) that he didn't wait for either the transmitter buffer being ready before sending, or after sending. I can't remember exactly, but maybe it is a hint you can use.

Also, in ISO-TP timing is very critical and nothing can last more than 100 ms or the ECU's will time out.

Maybe this helps a bit.
Jeroen - ZOE Q210 Nov 2013 - PV powered - CanZE co-developer http://canze.fisch.lu

Re: Hacking into Zoe's CANbus

USER_AVATAR
  • AbRiNgOi
  • Beiträge: 15206
  • Registriert: Do 27. Jun 2013, 17:43
  • Wohnort: Guntramsdorf (bei Wien)
  • Hat sich bedankt: 560 Mal
  • Danke erhalten: 4653 Mal
read
it worked at least!

I changed, and changed. First I checkt that my code waits on "transmitter buffer ready", that was so allraedy, than I tryed your 0x00 in the middle, than I saw that the original code I used is useing a Interrupt, so I started one interrupt after my channal for transmit, still not working. (I dont like interrupts, the landing is in main(), that is not the right object, and so I am programming c and not c++...
At least I tryed to send 8 Bytes, like that: 0x30 0x00 0x14 0x55 0x55 0x55 0x55 0x55, and it works !!! :lol: !!!
It works and runns upt to 0x22 0x0E ..., than I tryed my 0x30 0x01 0x14 again, and yes it works also, ending with 0x21 0x01...
so I can control my flow by setting each flow control all 10ms. The Interrupt will stay where it is, better not waiting on TXOK, doing somthin other with the CPU-Power than waiting...0xEBB is not very high prior CAN id, so it could be a long time until sending will start. Thank you for thinking with me, and so long! IO test is ready now, lets start the screens and the handling. I'll try to get a relais to switch of when ZOE in idle mode, and switch on with a tip.
ZOE Live Q210 6/2013 * AHK legal Typisiert 18.07.2017 * 40kWh Batterie 12.03.2019
Aktuell: 149.000 km

Niemand ist bei mir auf der Ignor-Liste!
Anzeige
AntwortenAntworten

Zurück zu „ZOE - Zubehör“

Gehe zu Profile
  • Vergleichbare Themen
    Antworten
    Zugriffe
    Letzter Beitrag