In the Dutch Renault forum, the issue came up that the maximum regenerative power allowed is higher than what CanZE reports as the Max battery charge / regen kW power. Based on a few experiments of my own only, it seems the Zoe will allow double the reported power for regeneration, capped at 40 kW DC of course. After all, it is supposed to be a short burst and the amount of energy, by definition, has already been taken from the battery.

I would appreciate your findings. Just open CanZE on the driving screen and whenever the Max battery charge / regen kW is substantially below 20 kW, try to do a serious regenerative brake (from a high speed works best, i.e. a motorway exit) and see how far the DC Power kW (or if it refreshes too slow, the blue kW in the dash) shows please. Thank you!

Edit: based on my own testing it seems there is indeed a short over-power possible. The rough formula is max_charge_power * 1.65 + 3.5. More testing by Borut suggests this seems sustainable for either roughly 40 seconds, or until the total battery voltage reaches 390 volts (4.06 volts per cell), see the comments.

It should be possible to calculate the max regen power by taking the max_regen_torque (the blue bar in the driving and consumption screen) and multiply that by the wheel speed (in rad/sec). I might implement this in a test screen.

The Zoe has a crawling mode that cannot be disabled (which is unfortunate I think, but I know others differ). It needs a bit of braking to counter this crawling when at a traffic light. When the brake pedal is pressed only a bit, the motor will still push and you’re spoiling a couple of hundred watts in the motor. So, when at a red light, either press the brake a bit firm, or switch the gear to N.

A (former?) Twizy and current Zoe driver in Austria called “AbRiNgOi” had a Twizplay laying around. This is an open source CANbus driven small display, based on Atmel micro controller. Anyone who has ever played with Arduinos knows what I am talking about. The specific controller used is an automotive version of an ATmega with a build in CAN controller.

He reprogrammed the Twizplay using the CANbus information that we gathered and that is available in the source code of CanZE on github. We are very pleased and proud that our hard work is spinning off toward other projects, in the true spirit of Open Source. Link here.

DSC02335

TwizPlay is originally programmed in BASCOM (a non-Free BASIC compiler for the Atmel and 8051 processors), but “AbRiNgOi” decided to do this the proper, but hard way and redo all using Atmel Studio 7 and C++, bringing it much, much closer to the Arduino community.

 

The R240 is the Zoe with the new, more efficient, air-cooled motor that Renault developed in-house. There is some consensus that the battery of the R240 is exactly the same as the Q210, but given the different characteristics of the motor, and the fact that the motor coils are used by the charger (very smart design!) it is capped to charge at a rate of 22 kW maximum. A UK based Zoe driver noticed the “Max battery charge” nicely went up way above 22 kW and asked how that could be.

polnjenje-zoe-r240-43kwToday, user Crf supplied us with this CanZE graph when charging his R240 at a 43 kW charger, and I believe it confirms what we were thinking. The Zoe simply caps the incoming pilot and fakes it’s value towards the other systems involved. Notice how it reports 32A and 22 kW available power. Those numbers are both off and way too neatly rounded.

The battery itself (the LBC’s) reports it’s willingness to take the full 40 kW DC power load, and as the SOC goes up it starts it’s capping at roughly 11 kWh. Again, the little jumps we believe is caused by the temperature rising. 2 kWh per degree Celsius is ballpark right. The very rapid capping near the end is kinda interesting.

So yes, it seems like the R240 has the same battery as the Q210. The BCB is tweaked a bit to simply ignore anything above 22 kW and the rest of the car simply behaves as if a 22 kW charger is attached.

So, loosing that little shelf for the dongle was a bit annoying right? Here is what I did. First, get yourself an ultra-flat OBD2 extension cable, such as this one www.ebay.com.au/itm/181765630282. You also need two RJ-45 connectors, an RJ-45 connection block and a crimp tool for the connectors.

Now that male plug is still not flat enough to fit under the shelf, so carefully shave off it’s top like this. Also, cut the cable 10 cm from the female end and crimp RJ-45 connectors on both ends, fitting the colors of the strands in the same sequence in both RJ-45 connectors. Be wise and check if the dongle still works with the two ends connected through the block before continuing.

20160107_182217

Next, widen the hole so a standard RJ-45 connector can be pushed through.

20160108_150805

Push a piece of stiff electical wire through to either side of the console. It works best if you pry the plastic of the console a bit away from the carpet.

20160108_150835

Use the electrical wire to pull the long part of the extension cable (with the male connector) through from the top, RJ-45 end first. Because of the locking clip of the RJ-45, this will be a once-only operation, unless you are prepared to widen that hole substantially more. Or simply cut the wire at the RJ-45 end when you need to remove it. Connect the two parts using the connection block. Finally, push the dongle into the female end of the now rejoined extension cable.

20160108_151112

Connect the male end to the car’s SAE 1962 connector and carefully wrap the cable so it won’t push up the shelf. Check again if everything works.

20160108_150957

Reseat the shelf.

20160108_151146

I tucked the cable under the console. I like this setup, as I can switch it off. Of course you can put it elsewhere or even hide it entirely if you don’t care about switching it off (you should care!).

20160108_151252

Done!

PS: Later on (no picture) I moved the dongle a lot more to the aft side. That got it out of the way of careless feet, and because of the shape of the floor it is pushed flush against the console instead of sticking out like you can see in the picture. Finally, it makes the LED’s far easier to see. Much better.

Our strong advice is not to leave the dongle active in the car while you’re away. In essence, it is a not-secure direct access device to the main CANbus of the car, with a default Bluetooth password. I have been able to wake up all the ECU’s from a sleeping Zoe using just the dongle. Things like opening doors are probably pretty hard to do but I wouldn’t consider it wise to believe it is impossible. That is one more reason we like the Konnwei – Maxiscans: they have this little on/off button on the top (unfortunately, they do not seem to go to sleep after half an hour of CANbus silence as the documentation suggests, but alas). At least it is easy to switch it off.

Another side of the question is power usage. I am bringing this up as my Zoe was completely and utterly dead this morning with a flat 12 volt battery. So, I decided to check how power hungry my dongle is: less than 1 mA when off, 40 mA when on with a sleeping CANbus, so that should drain the battery of not more than 1 Ah per day and should not be an issue.

Bottom line: be smart and switch it off. If not, and she’s still there where you left her, no harm done.

PS: It was really odd as the car was connected to the charger (had not charged for one reason or another), but the traction battery was still at least half full. My charging logging email from Renault suggested that she had her last charge in the night from December 30th to 31st. That could be right, as there was virtually no driving last week but last night definitely is strange. I have a suspicion about what happened but before I am sharing that I need to get rid of the error and ensure the 12 volt battery is not a dud.

Update: I had her up and running again after charging the 12 volt battery enough (a few hours on 3A) to wake her up and start charging. Never ever do this with the car connected. Remove the minus lead and, as the car can decide to try to charge the battery even with the contact off, make sure the lead is isolated. My suspicion is that the charger plug was not seated correctly and it tried locking it for a couple of hours with most of the ECU’s awakened by the dashboard computer. After charging there was a message to check the electrical system, but that went away by itself.

image22One of the new features in version 1.09 is the possibility to collect a time based graph of the charging process. It shows the development of the battery’s energy level, but also the maximum power level the charger is prepared to supply. If the charger is load balancing will show up in this graph. Furthermore, it is very clear how the car, while the SOC goes up, caps the maximum power that can go into the batteries, until it reaches the actual power level of the charger.

In the example here you can see the charger being prepared to supply 31 A over three phases (20 kW), and the car being prepared to take 43kW, but that last number going down as the SOC is going up. It starts capping at roughly 8 kWh in the battery, suggesting a battery temperature of about 12 C and then goes down until it reaches the charger’s power level of 20 kW at around 16 kWh SOC. If you look carefully, you can see the actual DC power following the capping in the last few minutes. The charge was ended there as charging speed bled off and putting in that last 5kWh would have taken excessive time.

The “blips” in the maximum power level we believe is caused by the temperature of the battery pack going up. It is measured with a resolution of 1 degree Celsius, which would explain the choppy line.

You can find these graphs under the Technical menu.