The Open CHAdeMO Charger Topic

[JeremyW]

It may actually turn that the lowest cost way for me to offer a residential or small business "medium" rate charger, is to buy the Nissan unit and upgrade (downgrade) it!!! :shock:

Indeed! Electronics is only really cheap if you produce a million of something. :lol:

Jeremy

 

[Ingineer]

Alright, lets talk CAN.

There are 5 different ID's on the CAN bus. Each message has an ID and 8 bytes of data. Three are the car, two are the charger. The car ID's are 100, 101, and 102. The charger ID's are 108 and 109.

(image)

Note the static bytes may be fault or other status flags. There's still more logs and more investigating to do.

Jeremy

There's an NDA preventing me from giving specific info, but don't forget about the initialization and test sequence. (HiPot) This is mentioned in the public CHAdeMO documentation.

-Phil

 

[nr56]

If the new chademo communication protocol is something like the old protocol, then a typical communication session should go something like this:

1.Car sends “Vehicle Stand By for Charging” message. [0x20]
2.Charger responds with “Information about Charger” message containing “charging mode”, “max output voltage”, “max output current”, “max output power”. [0x31]
3.Car sends “Constant current request” message containing “voltage upper limit value” and “required current value” [0x26]
4.Charger responds with “Constant current status” message containing echo of the “voltage upper limit” and “required current value” [0x37]
5.If no errors or interruptions repeat 3 and 4 while charging.
6.Car sends “Charge finished message”. [0x29]
7.Charger sends “Charging stopped message” [0x38]
8.Charging stops.

The numbers in the square brackets are the hex codes for the respective command.
This is one of the straight forward paths – of course there are messages for when something goes wrong and time limits for how long a response should take.
All messages are max 8 bytes:
[FF] [number of bytes] [command] [data x 4] [control byte]
Depending on the command byte, up to 4 data bytes may be expected.
This is just speculation – it is possible that the new chademo protocol is significantly different from the old one.

Edit:
Voltage data is 2 bytes (8bit ones), 1V/bit
Current data is 2 bytes (8bit ones), 0.1A/bit
Power data is 1 byte (8bit), 1kVA/bit

 

[nr56]

Just found this:
Ground fault relay for electric vehicle (Based on CHAdeMO Standard)

http://www.alibaba.com/product-tp/117560782/Ground_fault_relay_for_electric_vehicle.html

 

[JeremyW]

There are 5 different ID's on the CAN bus. Each message has an ID and 8 bytes of data. Three are the car, two are the charger. The car ID's are 100, 101, and 102. The charger ID's are 108 and 109.

I removed the old can chart because it had more unknowns then knows at this point. I'll be posting a new one eventually. For now, I'll just tell you what we know:

Car to charger:
102 - Charge amperage is byte 3. Byte 6 is SOC, not in GIDs, but expressed as .1kWh per bit.

Charger to Car:
109 - Byte 3 is again charge amperage.

The structure does generally follow what nr56 said. However it seems "information about charger" and "information about car" seem to be thrown back and forth throughout the session, every 100ms.

We could use more people sniffing out this bus... that is, if you have access to a CHAdeMO charger... :? I'm surprised by the lack of enthusiasm in this thread. Perhaps it is because a CHAdeMO charger is significantly less "DIYable" then j1772. Still, I figured more people would be curious.

 

[nr56]

I removed the old can chart because it had more unknowns then knows at this point. I'll be posting a new one eventually.

I fail to understand what/who/how does that help?

I'm surprised by the lack of enthusiasm in this thread. Perhaps it is because a CHAdeMO charger is significantly less "DIYable" then j1772. Still, I figured more people would be curious.

Maybe you will see more enthusiasm if you publish the raw logs?
More people would be looking at them and we will have something to discuss instead of passively read what your conclusions are.

 

[JeremyW]

I can't release the raw logs as they are not originally "mine." They are from another leaf member. I can ask, but I think the answer would be no at this point. They also made me take down the second chart because it was "too much too quickly" and there are "more unknowns then knowns at this point." Very, very frustrating, I know. I want to get a leaf so I can do this without intervention, but financially it is just not in the cards for me right now. The level 3 stuff was what made me very interested in the Leaf and electric cars in general.

 

[TonyWilliams]

If somebody wants to log some QC data in Anaheim next Tuesday, drop me a line. I'll be personally inviting the "likely suspects".

Tony

 

[hill]

If somebody wants to log some QC data in Anaheim next Tuesday, drop me a line. I'll be personally inviting the "likely suspects".

Tony

Anaheim? ... or Cypress

 

[TonyWilliams]

If somebody wants to log some QC data in Anaheim next Tuesday, drop me a line. I'll be personally inviting the "likely suspects".

Tony

Anaheim? ... or Cypress

Not Cypress, and not an Eaton labeled machine owned by a Japanese auto manufacturer.

 

[garygid]

A few have requested access to a QC-session log, and
wish to understand the QC process better.

We want to encourage investigation, understanding,
and colaboration, but do NOT want contributions from
anybody who has (or might appear to have) access
to the Closed Chademo Standards. All the results here
should come from observation of the QC process as
it is implemented in various QC machines and EVs.

The QC process, as observed, can lead one to assumptions
that might be correct, but might be incomplete, or wrong.
Please be properly cautious in your use of any of this data.

So, in the interest of understanding the QC process better,
and knowing how QC sessions affect the LEAF, I have
posted a 4-CAN Log of one 80% QC session, at:

http://www.wwwsite.com/puzzles/cando/" onclick="window.open(this.href);return false;

Caution: This is just one QC session, of one LEAF on one QC machine.
These logs should be helpful in investigating the QC process, but
do not contain enough data to implement QC completely or even safely.
They are only an observation of a process for intellectual curiousity.

NOTE: Accidently, not all the CAN Messages were captured in the log.

 

[garygid]

I suggest changing this topic's title to ... Open QC Session ...
rather than using the CHAdeMO (a proprietary standard) term.

 

[TonyWilliams]

I suggest changing this topic's title to ... Open QC Session ...
rather than using the CHAdeMO (a proprietary standard) term.

That is very much in the public domain. You're not co-opting the name for something else. We say Nissan, and LEAF (trademark by Nissan).

 

[garygid]

I was trying to distinguish between possibly giving the impression of discussing the Chademo specifications (which we do not have access to), and giving more of an impression that we are (I am) observing the typical QC process to see what we can learn from that. But, that's just my suggestion, nothing more.

I am mostly observing how the LEAF handles the QC process. For example: How does the LEAF handle the charging profile in response to Temperature changes during a QC session?

The Chademo "standard" probably covers handling many operational and error conditions that I am very unlikely to observe in any Log (but, who knows for sure).

Although there are those that might attempt to construct a QC device from what they learn here, it is unlikely that their implementation of "QC" would be complete enough to safely meet the full Chademo specifications. But, they might succeed enough to charge their own LEAF without "bricking" it. Please use considerable caution in any such attempts.

We hope that, as we learn more, and get closer to understanding the details of the process, that Nissan (and Tepco, etc.) will see the tremendous strategic wisdom in making their Chademo standard fully public, sooner rather than later, to better battle against the emergence of an alternate US L3 standard.

 

[JeremyW]

I suggest changing this topic's title to ... Open QC Session ...
rather than using the CHAdeMO (a proprietary standard) term.

That is very much in the public domain. You're not co-opting the name for something else. We say Nissan, and LEAF (trademark by Nissan).

I'm going to agree with Tony on this one. Plus, for searching purposes its best to have the name in the title. I will add the word charger to the first post, so it is made clear we are talking about open source chargers, instead of open source standards.

I do thank you Gary for releasing your QC CAN log. I hope others can contribute logs. To facilitate this, I've started a google code project. You can check it out at http://code.google.com/p/open-chademo/ There's nothing on there right now, but I will be working on it over the next few days. I'm going to be ordering one of these AVR development boards from Olimex. I'm also looking into ways to do a non-invasive monitoring of this bus, for those leasing or renting (if the rentals have QC). I'm not sure if the QC bus is on the OBD-II port.

As a working Electrical Engineer, at some point soon in my professional career I may be working on a CHAdeMO charger or testing solution. At that point, I'm probably signing an NDA and I will not be able to directly help with figuring things out. What I'm doing now is to try and make it easy for everyone around the world to monitor their QC sessions so we can get as much data as possible.

Jeremy

 

[JeremyW]

Hey remember that Tebiki? Here it is in English: http://www.chademo.com/pdf/QCInstallationManual.pdf

This manual was translated from its original Japanese version based on Japanese Law and
Supply Provisions of Japanese Electricity Utility, to meet requests from many CHAdeMO
international members. Therefore, we ask readers to keep in mind that different law and utility
supply provision of other countries/states might make some content of this manual irrelevant.

Jeremy

 

[Ingineer]

if the charging fails for any reason while the contactor is closed, the car will refuse to charge from ether the DC port or the J1772 until the car is brought to a Nissan dealer and reset.

This is not entirely true, if charging just stops, no problem. If charging is stopped due to an isolation fault, or if the QC doesn't respond correctly, the car could indeed set a DTC. Some of the DTC's (but only critical ones) also set an "inhibit restart" which prevents the main contactor from closing again. (for safety) This code can be cleared w/o a dealer trip by simply disconnecting the 12v battery for a few minutes. In fact, all codes can be cleared in this way, with the exception of certain critical system codes, such as braking system, SRS (airbag), etc.

Sounds like this guy didn't know much about the Leaf!

-Phil

 

[garygid]

I am aware that many Codes can be Reset/Cleared by re-Booting the LEAF.

I also believe that there are some that cannot be reset by reBooting.

I have been warned (by some who should know) that there is some condition (that QC experimentation or testing on a LEAF can create) that will "brick" the LEAF, and the "condition" can ONLY be cleared with Consult-III (usually at a dealer).

Do you have any reason (maybe that you cannot mention explicitly) to believe that this warning is not true?

Thanks, Gary

 

[garygid]

The OBD connector in the LEAF is described as having 3 CAN busses connected: EV, AV, and CAR.

Using 3 AVR-CAN boards and a 4-port RS232-to-USB adapter, we can easily Log all/any of these three CAN busses, running CAN-Do (in a Windows PC).

Apparently the QC-CAN bus only runs from the QC Port (in the front hatch) to the LEAF's Charging-Controller (apparently located in the back hump"). So, access to the QC-CAN bus is not as "easy" as access to the other 3 CAN busses.

 

[Ingineer]

I am aware that many Codes can be Reset/Cleared by re-Booting the LEAF.

I also believe that there are some that cannot be reset by reBooting.

I have been warned (by some who should know) that there is some condition (that QC experimentation or testing on a LEAF can create) that will "brick" the LEAF, and the "condition" can ONLY be cleared with Consult-III (usually at a dealer).

Do you have any reason (maybe that you cannot mention explicitly) to believe that this warning is not true?

Thanks, Gary

Personally, I have done all kinds of "non-approved" things with the CHAdeMO port on the Leaf, and I've never set any codes that cannot be cleared by a battery reset. That doesn't mean that you can't "brick" though, but I can tell you for sure that halting the charge process will NOT cause this! Hoever, if you are messing with the high-voltage, there are a number of ways you can trigger an inhibit restart code. YMMV, and DO NOT TRY THIS AT HOME! :shock:

Also, you are correct, there is no test connection for the CHAdeMO CAN link. (Not technically a bus, at least until you tap it anyway! =)

-Phil