I have issues charging my Renault Zoe ph2.
Firmware is 2.03
If I Plug in the Zoe, charging is starting for about 3s. Then, the car reduces charging current to 0 whilst starting the Heat pump for Battery cooling.
Unfortunately the EVCS detects this as the car being charged and stops the charging.
My car dealer tells me that my charging Station is non Compliant and they can‘t do anything for me which is quite frustrating.
Hello,
I’ve exactly the same behaviour with 35°C outside and a battery temperature (get with CanZE) around 32 °C
I found a not user friendly workaround:
use the Renault 10/16 A charger during one or two minutes. Unplug it. And start the charge with the Victron box in the next minute
Is the minimum charging current set correctly for the Zoe?
If I remember correctly it needs at last 10A
1 Like
Yes it is, I’m aware of that. The reason is that if a cooling phase occur, the consumption of fan + air conditionning compressor is near 8A => in this case the energy left to charge the battery is very low. But the compressor is usually not on for a big time.
I tried in manual mode with 10, 16, and 32 A setting with same bad result.
I’ m asking about the information “ventilation required” which can be provided by a VE to the EVCS ( more info here.
It could be send by the Zoe, but in this case how is it manage by the EVCS? It could be a reason for error. It is just an idee because whene the charge failed, if we use standard 10/16A charge cable, the charge start by acooling phase of the battery.
Is there a log accessible somewhere in the EVCS to try to undestand why the charge failed ?
Thanks,
charging current is set to 16A. I am charging 1ph, afaik the 10A is the minimum current for 3ph for the Zoe.
From what I read, the Zoe reduces the charging current to 0 while starting the cooling system to not accidentally overload the charger.
I do not know if the Zoe signals this somehow on CP to the EVCS or if it just cuts a relay (certainly sounds like it under the hood) for a few seconds while still signaling “I want charge” to the ECVS and the EVCS mistakes this for a end of charge.
If I can provide protocols somehow please let me know.
PS: I was able to test with a charging brick. Indeed, the brick does not terminate the charge when current goes to 0 but shows “connected” and resumes delivering current to the Zoe after the startup of the cooling system is completed.
1 Like
Hello,
I got some data with an oscilloscope. I’m analysing them regarding compliance to the IEC 61851 standard. As soon as I will have confirmed interesting information I will come back.
Benoit
1 Like
Hello,
first some information.
The interface between the car and the wallbox is standardized by the IEC 61851-1 : 2019. The document is quite expensive if you want to get it, but I have an access through my company.
On the internet we can find for free the 2011 release which is sufficient to understand the subject.
When the weather is hot (from my experience when the battery temperature is above 26 °C) the behaviour of the Zoe with the wallbox is different from the one when the temperature is lower:
-
the Zoe request the charge during 6 to 7 s, request to stop it (ie request the box to open its relay) , launch fans and air conditioning system, and request again the charge (relay to close) around 10 s later. You can check that with a simple charger by monitoring the charge led.
-
with the Victron EVCS, the sequence stops (Zoe raise an error) just after the request to stop the charge. I get the sequence on an oscilloscope.
In yellow you have the Control Pilot signal, in blue the 220V availability after the relay (got with a wire rounded over the charge cable, not connected directly for safety).
If you don’t know the signalling level, you can read the standard or other doc on the internet.
On the picture you see only enveloppe of signals due to scanning speed ( 500 ms / division)
At T1 the Zoe request the wallbox to open the charge relay by raising the positive peak of the pwm signal from 6V to 9V
At T2 ( around 60 ms later) the wallbox stops the PWM signal : it is not request by standard and can be an issue for the Zoe which will need it few second later to request charging again
At T3 (around 2 s later), the wallbox open the relay, and the 220 v disappear. The problem is that it is a deviation from IEC 61851 standard, as the relay shall open at latest 100 ms after the Zoe has raised the PWM signal to 9V. If the Zoe check the compliance of that, it can also raise an error.
=> I think the problem we face is coming from one or both of this fact.
Benoit
1 Like
I’ve continued to analyse the problem. For that I’ve simulated the behavior of the car with the CP signal using small electronics described on this schema :
extracted from this site.
When S3 is closed, it is simulating plugin of the car. When S2 is closed, it is simulating charging request ie closure of the relay. When S2 is opened, it is simulating the opening request of the relay.
With this test system, I get following results with oscilloscope:
I remind: in yellow the PWM signal, in blue the 220 V get from the charge cable with a coil for safety.
We can notice that the delay for relay opening is below 100 ms, which is compliant with the IEC 61851 much shorter than the value get on request by the Zoe
We can notice also that the PWM signal disappear during around 75 ms, which is not forbidden by the IEC 61851 but has low added value and could be source of problems for some sensitive car.
So next question is : what is the difference for the EVCS between the test with the Zoe and the car with resistors and diode : with the Zoe, some current is drawn.
So I add a curent transformer on the phase to display the current drawn by the zoe when the problem occur. It is the purple signal below:
We can see the current drawn by the Zoe, in purple, decreasing to zero, 1s later (T1) the relay opening request send by the car ( raising of pwm to 9V) and 2s later (T3) the opening of the relay.
We can notice that the car is “fair” and first decrease the current drawn before request relay opening to avoid sparks in the relay and to not decrease its lifetime.
The last idea is : How to trick the EVCS to mask it the current drawn by the Zoe ? It’s quite easy, just open the box, unplug the phase L1 wire from the relay, remove it from the EVCS phase L1 current transformer and replug it on the relay. The impact is that the EVCS won’t see the current drawn, won’t be able to display the power provided to the car, won’t be able to count energy etc… But it is possible to do a charge test.
On the scope we get the following result :
We can notice that the relay open around 140 ms after request provided by the car, which is a little above the IEC constraint of 100 ms, but the measure is not very accurate.
We can not see it on the picture but the Zoe charging procedure don’t fail and continue. The Zoe request again relay closure 20 s later to really proceed with charging 
So the conclusion is that the EVCS delay relay opening for no real reason in this case (there is no risk of sparks into the relay as current has decrease to zero for 1s), in deviation with the IEC 61851 lead to raise an error in Zoe charger (which could think that the relay is stuck).
It is not a solution because EVCS has no interest without energy metering, this test allows to find the source of the issue. The solution could be in the firmware.
Thanks if you read me up to there.
1 Like
Excellent analysis. I guess the Victron team has enough information to raise their sleeves and get to work. Protocols should be followed as close as possible.
From here it looks like Victron should be paying you for your contribution.
Hi! I’m really impressed by the community technical level!
Did you try to toggle the “Contactor active when charged:” on the EVCS Settings/Charger? This should keep the relay closed, and not bother the Zoé.
I have this setting “disabled” (ie. the relay opens when the car is charged), which doesn’t seem to bother my Stellantis car, and is better for the relay, which buzzed when closed for too long.
Yes, I’ve tried this setting without any success, and it is not surprising.
Following the iec 61851the contactot shall be open in less than 100 ms on EV request (équivalent to end of charge signalling). If the the setting « Conctactor active when charged» is on, it is in déviation from IEC 61851, the Zoe think the contactor is stuck and raise an error.
Normally, if an EV need electrical power for heating or air conditionning, according IEC61851, it shall do via a request to the wallbox (by lowering the PWM positive peaks to 6V)
If the EV needs the contactor to stay active for that’ it doesn’t comply the IEC.
Hello @Benoit . Nice analytics.
There is some delay before turning the contactor OFF to disable it when power consumption goes to 0. It can be changed over the Modbus TCP (need to connect directly to EVCS, not to the GX).
The Modbus TCP registers map can be found here. Holding register 5108. By default it is 2s. Can you please change it to 0?
Hello @Nicolas,
Yes, I will try it. I thank you for the quick answer.
As the weather need to be around 30 °C to have a battery temp above 26 °C, it should be possible to have results at the end of the week. I’m in Toulouse, France.