QC adapter for Rav4EV - the Project

With a fairly good experimental understanding of the QC
as used on the LEAF, we just need a good understanding
of the DC charging (if any) available on the Rav4EV.

Is the Rav4EV similar to the Tesla, or is it lacking any DC
charging capability at all? Tony will know.

Since Tesla advertises a Chademo adapter for the Tesla's
DC port, that would normally be fed by their Super Charger,
designing such an adapter would just require knowing more
about the car's DC control interface.

What does the Rav4EV have, other than a L1 / L2 input
(120v / 240v AC input)?

Whisper...
Someone we know should be getting a Tesla "soon".
I will have to inquire if they have ordered the "Chademo
Adapter" along with their vehicle.

The RAV4 EV has only a J1772 port connected to the standard Tesla 10 kW charger, so it can charge up to 40A at 240V.

To add CHAdeMO, I would expect you would have to add the port, a contactor connected to the battery pack, and tap into the Tesla BMS.

The Tesla supplied 40+ kWh battery pack is supposed to be 8 modules of 5 kWh each. These modules are in parallel, each supplying the full voltage, so they each have some independence. In the Tesla model S they somehow work together to control the charging profile when the car is getting a DC charge.

Tony has reported that the RAV4 coolant loop management is simplified compared to the Model S. Perhaps there are other simplifications in the BMS of the Toyota.

garygid said:

... Is the Rav4EV similar to the Tesla, or is it lacking any DC
charging capability at all? Tony will know.

Since Tesla advertises a Chademo adapter for the Tesla's
DC port, that would normally be fed by their Super Charger...

Click to expand...

Well, this is exciting. I have talked about this with GregH, since this will take a bit of CAN bus massaging.

There is nothing on the Rav4 EV except the the J1772 port and a single Tesla 40 amp charger. The Tesla CHAdeMO adaptor is fed by the CHAdeMO charger, not a Supercharger. It wouldn't have anything to do with this project.

In a nutshell, we would need to charge a 386 volt battery. I suspect that we would need an entire mostly independent car-side CHAdeMO control. Are you up for that? How much could we take directly from the LEAF CHAdeMO CAN bus?

Doing the car-side of the QC is the easy part,
however, understanding how to interact with the
existing car charging control and BMS is the
considerably more difficult part.

However, the BMS and charger are probably/possibly
quite/somewhat similar to the same components in the Tesla,
so learning how the Super-Charger interacts with the
car might help us to understand the internal charge
control and the Interaction with the BMS.

Somewhat like the Bursa in the LEAF, we might have to
trick the AC charger into charging with just a minimum
acceptable trickle, and then shove some extra
DC into the battery until the BMS starts limiting the
charging rate.

Have the CAN bus ports on the Rav4EV been identified?
Have they been logged during a L2 (or L1) charging session?

Indeed this would be a very interesting project. Not sure how exactly I can help but if there is some way, I'd be glad to.

I don't have all the cool tools for can bus logging, alas.

When I find out more about the car's CAN buses,
then we can specify what is needed to do the logging.

Most often, we use the AVR-CAN development board,
the same as we have in the GID-Meter. Add some
suitable cables and connectors, and you can Log.

You use the USB port on a Windows PC to connect
to my CAN-Do program to capture and examine
the data. Try it with one of the posted logs.

When I get some reasonable Tesla or Rav4EV logs,
I will post them so that you can join in with trying
to figure out the meaning of the Logged data,
with nothing other than a PC needed.

It is my understanding that the Tesla supercharger is using the HomePlug Green PHY protocol, which is the same as the SAE DC "frankenplug". Tesla is going to make an adapter so that they can convert PHY to CAN in order to control a ChaDeMo charger. Since the RAV4EV does not have any of the supercharger hardware, that project would probably be easier (from a data capture point of view) since it would only require CAN.

http://www.homeplug.org/tech/homeplug_gp" onclick="window.open(this.href);return false;

What would be needed to generate some logs? I'm not aware of anything like the GID-meter out there for the Rav yet, but I'd be interested to do some testing.

The heart of the GID-Meter is a CAN reader and logger.

One needs to identify the car's CAN buses, the pin numbers,
which is the CAN-High signal, and which is the CAN-Low signal.
Generally this is best done with an oscilloscope, or similar device.

Identify the bit rate for the CAN bus, usually 500,000 bits per second.
Identify the Message ID type, where 11 bits is typical.

With this information, we should be able to Log the CAN data.

The Tesla S has one OBD type connector near the steering wheel,
and the pins 6 and 14 are the "standard" CAN bus, as I recall.

There is a second OBD-type connector "hidden" in a compartment
in the cargo area, behind a panel on the driver-side wall.
We had the impression that pins 6 and 14 of that non-standard
function connector were also a CAN bus. However, an initial
attempt at logging was somewhat unsuccessful, I am told.

One might expect a similar situation with the Rav4EV.

At the moment, I am unaware of a nearby Tesla or Rav4EV,
with an owner willing to allow some careful measurements.

Volunteers most welcome.

yes, of course the car has CAN, and you should be able to get lots of information about the battery and such, but the charger control communication for the supercharger is not CAN. The on-board charger is probably controlled by CAN, and you should be able to get some good information from monitoring that.

If we get a volunteer with a vehicle, we first need to
find a meeting place. After we have our breakfast Gathering
in Cerritos, frequently we can use Dori's office in Cypress,
near the QC at Mitsubishi Headquarters.

Or, we could met at my home in Laguna Hills,
or I / we could come to your location, especially
if I would be allowed to charge my LEAF there.

Second, the "OBD" connector near the steering column
might not be the only CAN bus. We need to find the
equivalent of the EV-CAN bus, if it exists. In the Tesla
it is in the cargo area, on the left side.

Then, connecting a short extension so that we do not
have to get inverted or otherwise contorted, we carefully
measure the signals on the various pins, preferably with
an oscilloscope (I have one).

From these observations, we try to guess the type of
signals that we are seeing. Usually, if we see what
looks like a CAN signal pair, we can make an adapter,
and log some data.

Once we have logging working successfully, then gathering
data while charging is the best first start. Logging while
driving is next.

If there are two (or multiple) CAN buses we will try to
arrange to log both buses simultaneously, to better
correlate the data.

Sometimes it might take a couple of sessions to get
everything discovered and working.

Of course, logging the "up front" CAN bus is usually
fairly easy, since it might match some standards
that actually only apply to cars with emissions.

Our next Gathering is in Cerritos on Nov 9th,
typically from 8 to 11, at the Hometown Buffet
just 2 blocks East of 605, and a half-block North
of South Street (first exit south of freeway 91).

The Tesla 85 kWh pack charging at a maximum of 90 kW from a Supercharger is a charging rate just a little over 1 C. This suggests that the RAV4 42 kWh pack could safely charge at a maximum rate of 45 kW with the same charging profile and tapering as the Tesla supercharge. Tesla has recently qualified their cars to charge at a starting rate of 120 kW, which is available on all their newer Superchargers unless there is another Tesla charging on the other half of the same charging station pair. 120 kW is shared between two hoses. This means 1C is conservative.

The 40 kWh pack was originally available for the Tesla, but it was canceled due to low demand just before production of this size was scheduled to begin. Tesla clearly stated that the 40 kWh pack would never have Supercharger access. It would appear this was primarily a marketing decision, because the charge rate could be adjusted to the same C rate.

It would seem that RAV4 pack would be well-matched to a 48 kW Quickcharger. Although not elegant, you could capture the charge profile by manually monitoring a Tesla Supercharging session. Since the pack is actively cooled during both driving and charging, the session timing should be relatively stable.
This profile could then be programmed into the added DC charging controller.

I realize you still need to see bus traffic to/from the BMS during a charge. How the AC charger and BMS communicate is unknown. The modularity of the battery pack possibly complicates things.

Ok, let's set up a date to make a plan. As far as the "programmed charge schedule", sure, just copy whatever Tesla does on the 60kWh car, minus a wee bit to be conservative. But, I don't really want to spend any time on the actual data on our first meeting, but the method to get there.

First and foremost is that communication between CHAdeMO and the Tesla BMS. Everything else will fall into place. I like the idea of piggy backing on the AC onboard charger while it's operating to see how much extra current we can shove in there without a fault. We don't need CHAdeMO for that; just a big charger that can be controlled at 386 volts DC.

I propose Tuesday, since I'm driving up to AeroVironment in Monrovia, and we can also have Jeremy there and hopefully GregH.

Besides the communication issue, the question of safe access to the HV DC seems relevant. The LEAF has a manual emergency disconnect plug under the floor carpet in front of the rear seats. The Model S has very tight integration between its battery pack and its inverter. Peak drive currents can exceed 800 amps, so cable runs are kept very short.

Tesla has an interesting emergency response guide that gives some good information on the HV DC system. Much, but not all of this should carry over to the RAV4 EV. As Tony has described, the battery pack orientation in the RAV4 is reversed from the Model S to accommodate the RAV4's front-wheel drive.
http://www.teslamotors.com/sites/default/files/downloads/en_EU/emergency_response_guide.pdf

As described in this document, emergency disconnect in the Model S is achieved by cutting a low voltage cable. This apparently opens one or more contractors. If these contactors are integrated into the battery pack, they are probably also present in the RAV4 EV.

The HV DC cable running to the 10 kW AC charger would be one access point, but it is most likely not sized for 44 kW (about 120 A).

tbleakne said:

The HV DC cable running to the 10 kW AC charger would be one access point, but it is most likely not sized for 44 kW (about 120 A).

Click to expand...

I'm confident that we can input the CHAdeMO energy with an appropriate sized conductor at the termination point at the front of the battery. I'm confident that the relays up "upstream" from this point. So, at the most basic level, we have to be able to modulate:

1) High voltage relay (there likely needs to be a dedicated CHAdeMO relay, also)
2) Battery cooling system
3) Fully functional BMS

And this is all CAN bus work.

These Tesla battery packs have been described
as having 8 parallel "stacks" of cells in series.

I realize that one "cell" might actually be a number cells
in parallel (the LEAF has 2 in parallel), a "p-cell" perhaps.
Then, a string of p-cells in series (perhaps 100 or 200)
would provide the necessary voltage to make a "stack".

If the Tesla design has several of these "stacks" in parallel,
does each stack have its own BMS, measuring the voltage
of each p-cell?

If so, it would have 8 BMS functions, measuring the voltages
on about 800 to 1600 p-cells.

Or, do the p-cells really contain 8 times as many cells, and
the Battery Pack only measures (about) 100 to 200 cell voltages?

Sure, either way, it can be made to look like a single series
of super-p-cells, but I had assumed that there lots of cells
in parallel, and only one series string of super-p-cells.

How do the battery pack onfigurations of the Tesla R, S,
and Rav4EV compare with each other?

Heck, I did not even know that the Tesla is rear wheel drive.

Now, you can guess that it is not me buying a Tesla.

The Roadster is a completely different battery pack and BMS system.

I doubt that we will have much interaction with the BMS activity, or that we even need to know much about the layout of the battery. For the purpose of DC charging, it's one big 386 volt battery. For the BMS, it's thousands of cells. We just need to let the BMS do it's job, and make it operational during a charge to protect the cells, just like it does with the onboard charger.

I doubt that there are triggers in the BMS that will limit the maximum current input, but we'll find out.

The high voltage leakage protections that the LEAF has should all be adopted to the RAV4. Actually, we should just take all the hardware from a wrecked LEAF.

What exactly are the pieces of information we need from the CAN to supply the "CHAdeMO bus" that would have to be built?

The only thing that we have to look at is the behavior of the Rav4EV during charging.

The most important data to look for is something like the BMS telling
the internal charger to taper off the full current charging, and then
when to stop charging.

Also, the charger would look at the L2 Control Pilot to determine how
much AC current is available, and compute the available power.

Subtracting loses and overhead (like cooling) the charger would
essentially decide upon the initial current to deliver to the battery,
and perhaps ask the BMS if that is too much initial current.
When the battery is very near full, the initial current should be quite low.

Then, there should be some agreement about when to close the
charging-access contactor.

If we can understand the Charger-BMS conversation, we might be
able to avoid waking up the charger control, and just become a
faux-charger, having our own conversation with the BMS.
But, this understanding is probably not easily won.

Where is the 2nd CAN port to be found in the car?

garygid said:

Where is the 2nd CAN port to be found in the car?

Click to expand...

One under the steering wheel, like most cars. The other is located in the left rear of the car on the "gateway" and communication module.