A lot of debugging has been going on the Battery Charging Prediction. Next to the usual bug fixes, the number of minutes between each line is now dependent on the charging power of the chargepoint (i.e. 45 minutes per line for 3.7 kW, down to 6 minutes for 43 kW). Also, a column is added showing the expected real power intake, so you can not only made an educated guess about when you will have charged up to the range you need, but also when the battery starts throttling the charging power.

In the source code It has already moved from the experimental section to the technical section and we expect this to make the next release. We’d be very interested in your feedback. Please not that while the module might work for a Fluence or Kangoo Z.E., the predictions will probably be way off. We’ve only tested this on a Q210 so far.

I admit traffic lights are the places where, when driving the Zoe, the challenge is burning. Boys toys. But after watching Alloam’s video about trying to get a 100/100 Eco score I have changed my strategy and now try to get really good mileage, just for the fun of it. Here are my tips for those who want to try:

  • Be extremely gentle on the accelerator. When not hindering other traffic, try to keep your power below 8 kW
  • Anticipate like crazy. If there’s any indication of slowing down, aim for 0 kW if possible. Remember, try coasting on 0 kW when possible, then regeneration, then braking
  • Keep your speed down when possible. If you’re not in a hurry, just go with the lorry flow. Keeping 86 – 92 km/h makes a HUGE difference

Opinions of using the Cruise Control, especially in hilly terrain, varies. For flat terrain I always use it.

Screenshot_2016-06-02-11-23-17Where does CanZE come in? Well, for me, I keep an eye on the kWh/100km line in the middle graph of the Consumption screen. I noticed that after a while a second feedback mechanism emerged: the bottom graph shows a black line for the SOC in percentages, and a red line for the estimated range in kms (on a scale from 0 – 150 km). I really try to keep the red line above the black one, giving me a long-time average of over 150 km per full charge. As you can see, 170 km’s is doable.






Screenshot_2016-06-13-01-20-14And it really pays off. Look what I saw yesterday evening!


In the Bleeding Edge source code on Github, we’ve implemented a battery charging prediction model. It is accessible through the Experimental section. Here’s what it does.

When you start charging the battery, CanZE fetches the State of Charge, battery temperature, chargepoint power and  the range estimate from the car. The first three are fed into a mathematical model of the battery. The model is then run for 100 minutes. Each 10 minutes the State of Charge is displayed, along with a range prediction. This will allow you to estimate how much time you have before the required range to get to your destination is reached.

The model is of course not perfect, and by it’s nature, errors accumulate. Feedback is appreciated. We intent to put it in the next release.

Edit: The screen shows always 10 lines. It’s silly to pretend accuracy is better than that. Depending on the possible charging power detected, the number of minutes between two lines varies between 10 and 50 minutes. We’ve ironed out a couple of nasty bugs, but it’s ready for the next release now.

Sometimes chargepoints are wrongly commissioned, i.e. set for 63 amps, but fused for 40, which works mighty fine for all EVs, until a 43 kW capable Zoe comes along and trips the breaker.

In another case a driver in Belgium owning a new R240 and who has a single phase 32 amps chargepoint installed, the car consistently reports 20 amps max. He called the installer who insisted all was A-OK. So, the driver blamed the car. Renault couldn’t tell him if 32 A single phase is supported on the R240. However, they were nice enough to lend him a Q210 to try (and the Q210 supports the 32 A single phase configuration). Whoops, again 20 amps only.

CanZE of course reports this all, but to exclude any issue in the car, I build a little “amp checker” which reports the maximum current per lead the chargepoint communicates to the car. It’s a tiny, tiny little bugger. If you’re into some electronics fiddling and not afraid to try, here’s the instructable.

SAE J1772 EV charger checker

All EV’s are bloody zippy starters (that is why they have such carefully crafted it’s tail lights, those need to show off 😉 ). I did some theoretical calculations in the SpeakEV forum.

When accelerating the car maintains two limits:

  • the torque that the motor and drive train can handle. For the Zoe this is 220 Nm on the motor, 2000 Nm on the wheels. This translates to a force on the car of 7400 N.
  • the power the controllers and motor can handle. This is limited to 72 kW (with some peaking to 78, but we’ll stick to 72).

At first, the torque limit is maintained, giving the 1468 kg car, a linear acceleration of 5 m/s2. 7400 N acting on 1468 kg gives you that. Mind you, this is half of the earth’s gravity acceleration!

When reaching 9.8 m/s (35 km/h), so after just under 2 seconds, the car hits it’s power limit and torque (and therefore acceleration) starts to bleed off. To reach 100 km/h (27.8 m/s) under full power requires another 6.9 seconds. This can be calculated using the formula:

speed = Sqrt (2 * power * t / mass)

determining the t’s for 9.8 m/s (0.98) and 27.8 m/s (7.88), and taking the difference between those t’s.

This all totals to 8.9 seconds to get from standstill to 100 km/h, which I think is not earth shattering, but pretty impressive for a 1468 kg small family car. Most EV’s are a tad faster. These calculations ignore the weight of yours truly & friends, friction, but also said 78 kW.

A few weeks back I pulled to a traffic light, two lanes going left. Behind me was a guy, I think in an old 206,  who had determined he wanted to be “that” guy, and he tried to cut immediately in the leftmost lane. As I went there (because I had to go left at the next traffic light again), he made a crazy manoeuvre to get next to me in the right lane. I was with a friend and I told him to hold tight. The 206 came pretty close to blowing his engine, we heard it screaming behind us. The youngster ended up so frustrated that he jumped next queue over the bus lane. We had a good laugh over that one.

In the post about the braking system, we referred to the dealer mechanic introduction document, which has a bunch of very informative images. Here are two more about the climate system which I think are worth sharing. Pretty smart. From myrenaultzoe.com.

Dutch forum member OlafH dissected the pedestrian horn a bit and posted this picture.


The horn is a bit hard to get to, one has to remove the front bumper to get to it. I tried to analyse it a bit. Here is a summary of the most important components. Prepare to be amazed.

SPC5602: A generic processor, though aimed at the car industry, made by NXP. It’s current id is MPC5602P. It has 64KB data flash, 20KB RAM, 256KB code flash, and a PMW generator. MPC5602P

TPA3111Q1: A 10 Watt D-class (PWM) mono audio amplifier by Texas instruments. No real surprises on this one. TPA3111D1-Q1

UJA1076A: CAN transceiver, a nifty little chip, made by NXP. It can also acts as a power supply and watchdog. UJA1076A

25P16vpa:  16Mbit (2MB) flash memory, made by ST Microelectronics. I bet the sound files are stored here.  Enough for 2 minutes mp3 encoded data or 12 seconds of raw sound files. 25P16vpa

1334A: Stereo DAC. Hell, this horn is CD quality, they didn’t even use the PMW output of the main processor!!!! Also NXP. 1334A

7342: Dual hexfet. Basically an on-off switch, by International Rectifier, now Infineon: 7342

The horn resells for about 180 Euros. It is hooked up to the Electro CANbus, power, and a few more wires. While I haven’t touched it myself yet, I bet the sound selector push-button is wired directly to it.

By the way. Another name for a Pedestrian Horn is Vehicle Sound for Pedestrians or VSP, though this is officially a Nissan name and developed system.

Edit: Matthew posted a link to an in depth description of ZOE’s VSP in the comments. I’ve copied the file here on the CanZE blog should the link go dead. Thank you Mathew.

Edit: Thanks you Rudi42 for the additional picture of the location, and of course the custom project.

Renault has been selling the official Zoe granny charger for a while. The official name is AFAIK the “Flexi charger”. It retails for a bit over 700 euro’s. Charging current is fixed to 10 amps. Slovenian forum member “pirpy” dissected it (partly) and noticed there is a control wire to the Schuko plug to change the maximum power setting to 13 amps. It is done by using a “special socket”, that is coded to supply that current using….. a strong magnet!!

See this forum entry and a few posts down. “pirpy” decided to use this knowledge to recode the flexi charger to 13A permanently.


In all fairness user “Crf” in the UK SpeakEV forum pointed me to this and I fully admit I thought he was pulling off a pretty decent April 1st joke, but it is really true!

Edit: as expected the Flexi charger has two relays so it can switch polarity. That was impossible to do in my granny charger, simply because there was no room for it.