[prensel]

Hi all,


I'm having a few problems with the CP (ControlPilot) signal generated by a few commercial-, public and private charging points or EVSE's.


After visiting almost 10 different EVSE's the last few days i noticed different behaviour at almost every point.



Here's my setup:
- Brusa NLG511 charger, CP wired directly to the Mennekes Type2 plug/EVSE.
- adapter cable where the CP signal is manually triggered (like the right side of this scheme) inside the Mennekes type2 plug.


https://en.wikipedia.org/wiki/SAE_J1772#/media/File:J1772_signaling_circuit.gif

Using just the adaptercable (without anything attached) and flipping the manual switch toggles the CP resistance between 2.74k and 880R causing to either halt/pause the EVSE or start supplying current. So i can actually start the charging process and pause/stop if needed using the switch.



A lot of the EVSE's kick me of after a few minutes when requesting more current then allowed to pull which seems fair because there is no negotiation about the allowed current using the CP 'channel'.
Other just simply deliver the max allowed current per fase as long as the switch is in 'charge' mode.



However when using the Brusa the EVSE's never delivers power but keeps on waiting for someone (the Brusa?) to flip the internal switch...


A few years back the switched variant wasnt acutally needed because one could use the 880R trick to start the EVSE to do its job. But now it seems the EVSE's got smarter and expect more CP stuff then just seeing some R on the cars end...


I assume the Brusa doesnt support these smarter EVSE's of changing the R and trigger the EVSE's to start delivering and just gets 'stuck' in the waiting fase.



So how could i combine these two variants ? I'd like to handle the Brusa the allowed current using the CP where at the same time I need to add the switch the start the charge session.



Paul

 

[prensel]

I'm still struggling with Control Pilot issues.

As far as i know the only 'valid' CP protocol is defined in J1772.
The European IEC 61851 equivalent is supposed to be using the same standard.

For some reason some public charging stations /EVSE's (in The Netherlands) dont seem to work with that protocol. This means either a different protocol or some undefined tweak or addition on the J1772 protocol.

After calling with several different public charging station service providers they cant give me the answer as why some EVSE are working and others are not.

At this point i'm trying to build some sort of diagnostic CP device using an Arduino to measure the different CP voltages and PWM signal on different EVSE's according to the J1772 protocol.

Can anyone tell me if there are other protocols to look at ?

 

[Kevin Sharpe]

For some reason some public charging stations /EVSE's (in The Netherlands) dont seem to work with that protocol.

I doubt this is a problem with the EVSE because we would undoubtably have an outcry from all the EV drivers in The Netherlands.

My suggestion is that you buy an AVC2 and rig up a simple test load that you can plug into the EVSE that are giving you problems. This will eliminate the Brusa from the equation and give you confidence that the EVSE are working. The AVC2 is 39.99 (USD) and you can probably sell it on ebay once you're finished with it

 

[prensel]

I doubt this is a problem with the EVSE because we would undoubtably have an outcry from all the EV drivers in The Netherlands.

Not perse true, i believe about 99% of the EV's driving here in the Netherlands are (software updatable) COTS like Tesla, Leaf, Zoe and a few others having (updateable) chargers that do respect the CP PWM signalling or possibly react well on any/the altered or 'misbehaving' CP protocol. I have read somewhere that even Brusa has firmware for different CP protocols but cant confirm that (or even know whats different). According specs the Brusa NLG5 series supports J1772 and i have tried two different chargers with the same (very unpredictable) results between EVSE's

I have also noticed a very clear difference between for example Alfen EVSE and ECOtap EVSE's. And even in the Alfen range there is a difference between two EVSE's within a mile/km located from each other.

My suggestion is that you buy an AVC2 and rig up a simple test load that you can plug into the EVSE that are giving you problems. This will eliminate the Brusa from the equation and give you confidence that the EVSE are working. The AVC2 is 39.99 (USD) and you can probably sell it on ebay once you're finished with it

As far as I can see this AVC2 is nothing more then a 2 resistors, a diode and a switch and should work perfectly well according the J1772 standards when changing the different stages as defined in the J1772. Fact is that also this device (like my dumb mode3 converter test device/cable) doesnt respect the CP PWM signalling meaning when the EVSE commands to limit current to 10A the connected charger doesnt have a clue about that request which will end up in being kicked off buy the EVSE after a few seconds when requesting more then 10A.

Since V2G is becoming more into the public area the CP PWM signalling is going to be used more and more and probably going to be a PITA for us DIY'ers without a decent working CP protocol and charger.

So for know my only 'solution' would be to rig up a test device with logging capabilities and visit a bunch of EVSE's to see what the differences are. Based on that outcome i probably need to build a MITM (ManInTheMiddle) EVSE coupler which will talk to the EVSE on one side and talk to any charger with AC current limiting capabilities on the other side. There are some whitepapers on the net about these devices so i assume more people have noticed this earlier.

 

[Kevin Sharpe]

Not perse true

I designed my first EVSE Controller in 2010 and when we sold the company last year more than 10,000 were in use in multiple vendors Charging Stations. After the initial bugs were ironed out the firmware was never updated, the EVSE protocol is simple and robust.

Fact is that also this device (like my dumb mode3 converter test device/cable) doesnt respect the CP PWM signalling meaning when the EVSE commands to limit current to 10A the connected charger doesnt have a clue about that request which will end up in being kicked off buy the EVSE after a few seconds when requesting more then 10A.

Obviously, any device drawing more current than indicated by CP PWM will be 'kicked off' by design. You cannot exceed the safe working limits of the EVSE.

The reason I suggested using the AVC2 was so that you would have a known benchmark to test third party EVSEs against. Obviously you could invest in a EVSE protocol analyser but they are not cheap... iirc the last one we purchased cost $10K

How about building an Open EVSE board (here) and test your charger works as expected? You can then modify the Open EVSE software to test for edge cases to see if your charger protocol breaks down at some point.

 

[prensel]

How about building an Open EVSE board (here) and test your charger works as expected? You can then modify the Open EVSE software to test for edge cases to see if your charger protocol breaks down at some point.

I have the OpenEVSE based SmartEVSE (https://www.smartevse.nl/) and the AnalogEVSE (http://analogevse.xyz/AnalogEVSE-en.html) for testing locally.

I'm looking at combining this
http://dozuki-guide-pdfs.s3.amazonaws.com/pdf/botsbits/guide_171_en.pdf

with this:

http://www.instructables.com/id/SAE-J1772-EV-Charger-Checker/

To create a MITM device able to either function as a simple diagnostic device or to act as an J1772 'translator' to any non-J1772 charger.

Using the Arduino to control the two parts and log all data over serial/usb output when needed.

 

[Kevin Sharpe]

I have the OpenEVSE based SmartEVSE

We did a lot of testing and bug reports on OpenEVSE code in the early days... assuming your SmartEVSE product has not deviated then you should have a solid reference for your tests.

I've used almost every type of EVSE in the UK (and lots in Europe) with a whole bunch of different EVs. In the early days of our EVSE development we tested everything we could find. I don't remember having problems with EVSE/EV protocol for many years, just occasionally we see problems with grounding because of poor EVSE earthing during installation.

I don't recall hearing many problem reports for the AZD or other EV conversions that use the brusa chargers

Do you want to post some details and photos of your wiring? Posting your brusa setup would be helpful. A few videos of the EVSE failures might also give some clues.

What version of the standards are you using as a design reference?

 

[Kevin Sharpe]

Wolf has done a lot of work with the NLG511... he might have some thoughts

 

[prensel]

I don't recall hearing many problem reports for the AZD or other EV conversions that use the brusa chargers

Do you want to post some details and photos of your wiring? Posting your brusa setup would be helpful. A few videos of the EVSE failures might also give some clues.

What version of the standards are you using as a design reference?

Maybe its because i'm using the NLG511 and not the NLG513 as is used in the AZD builds ?

My charger wiring is as follows:
- GND on pin 1
- permanent power +12V on pin 2
- switched +12V on pin 3 (PON) when Mennekes plug is insterted in the car

- CP pilot wire from charger output port (pin 4) wired directly on Mennekes CP pin.
- resistor from GND to PP pin in Mennekes plug.

Normal behaviour:
1: When the plug inserted in the car inlet the BMS closes the contactors
2: enables the +12V on the PON pin.
3: This causes the charger to start and wait for the AC to come on.
4: When plugging the EVSE end of the cable the EVSE signals green that a plug is attached.
5:After presenting the RFID card the EVSE switches to wait-state
6: EVSE then switches to charge state and closes the AC contactors.
7: AC current is now coming into the car
8: The charger is starting with the profile and charging.

Misbehaving EVSE's:
- EVSE toggles between 5 and 6 continuously
or
- during 8 the EVSE goes into error mode and closes AC contactor, going into error-state
or
- during 8 EVSE goes into wait mode and stays there, no AC flow into the car

Most of the time these errors happen when theres already another EV plugged into the 2nd/other outlet of the EVSE or when the 2nd EV is plugged in during my session.
So my assumption is that it is caused by the EVSE wanting to limit the available current.
I have a power limit switch connected to the Brusa so i can limit max AC current to 10A. This morning when an EVSE was causing forementioned troubles I used this option and the EVSE kept on working.
Without it (requesting full power or CP PWM power) the EVSE kicked me off after a few seconds.
Again this was on a two car EVSE with both outlets in use.