To cut a long story short, our 2015 VW E-Golf is equipped with a Combined Charging System (CCS) which allows the vehicle to be fast charged using DC current from Electric Vehicle Supply Equipment (EVSE)
Unfortunately, the E-Golf failed to fast charge via DC but would charge via AC
Note above how the conversion of AC to DC is handled by the vehicle “On Board Charger” (OBC) when charging via Mode 2 & 3; however, with Mode 4 (for vehicles equipped with CCS) the AC to DC conversion is handled by the EVSE (so by-passing the functionality of the OBC)
Below we have the typical warning message received from the EVSE when attempting to fast charge
So where do you start?
If we follow the process below, I can confirm that no DTC’s were present and live data surrounding DC charging did not reveal and “leading” data to suggest where the fault may reside. Bullet point 5 below however does resolve the issue in this case as there is a software update for the High Voltage Battery Charger (I.D J966)
In the real world, this would be the end of the story but I was intrigued to understand what was happening to the DC charging system before and after fix.
In combination with our EV Guided Tests (which are no substitute for training and certification) measuring the Control Pilot (CP) communication line and Proximity Pilot (PP) of our vehicle charge inlet with the CCS EVSE connected to the vehicle got the ball rolling
N.B ensure the relevant CAT rated differential probes are used when the vehicle is connected to the mains network
Note above the change in PP and CP signals as the CCS EVSE is connected to the vehicle, however, the lock pin position signal indicates no change of state confirming the EVSE connector is not locked to the charge inlet! More information on lock pin position signals can be found here https://www.picoauto.com/library/case-s ... in-charger
Additional points to note:
• The noise level on all signals whilst the EVSE is connected to the vehicle
• The interruption to the CP signal around 32.5 seconds. (EVSE making a second attempt to communicate with the vehicle)
• The positive duty cycle of the CP signal at 5%.
N.B. A positive duty cycle of 5% indicates to the OBC that “Digital Communication” will be used to control an Off Board DC charger (Keep this in mind)
Below we have the typical behavior of the CP circuit for this E-Golf when connected to a Mode 2 charger (Often referred to as the “Granny Charger”)
Items to note above
• PP voltage differs from CCS EVSE PP voltage as the resistor used between PP & Protective Earth (PE) differs to indicate current capability of Mode 2 charge plug/lead. Refer to our EV Guided Test, Electric Vehicle > Charger-vehicle tests > Charger -vehicle proximity line (type 2) “Further Guidance”
• Reduced noise level across all signals (even when charging commences)
• Change of state of lock pin position voltage confirms EVSE connector is locked to vehicle
• The positive duty cycle of the CP signal at 13% indicates a charge current of approx. 10 A
• Channel D indicates 0 A as this is placed around the DC charging cables from the vehicle charge inlet (This is normal as we are not using DC charging in the example above)
After some research it came to light that when DC charging Bidirectional communication is required but how?
When AC charging, the CP line voltage level and PWM are used as a means of one-way communication between the EVSE and OBC, how then can we have Bidirectional communication?
The answer is Power Line Communication or PLC for short and you can read more about this awesome protocol here https://openecu.com/power-line-communication-plc/
Basically, the CP line also becomes the means to transmit data packets as well as a PWM signal. Think of this as data packets riding on a PWM signal. We have touched on this subject here viewtopic.php?p=101043#p101043 and Ben reinforced the discovery here topic22772.html. In these scenarios the PWM signal is momentarily changed to form a data packet, whereas PLC is something very different indeed!
Below we have PLC in action riding on our CP PWM signal
Back then to our CCS EVSE capture, notice the cyclic noise present on all signals below:
Could there actually be “data” within this noise? Below we zoom into the CP signal and note the sporadic nature of the positive duty cycle thanks to the noise level above.
Above we have added a DCC decoder to the signal in an attempt to display “something” interpretable but as you can see it appears garbled
Now look at the uniformity to the noise level riding on the CP signal (duty cycle) after the software update was carried out to J966
Note that the DCC Decoder is not the relevant decoder for this form of PLC, it is used to purely demonstrate that “structure” exists to the acquired waveform
Below we have confirmation that DC charging is functioning as we can see the high DC current being delivered from the CCS EVSE to the vehicle battery
Note the following in the above capture:
• The 5% positive duty (Digital Communication” will be used to control an Off Board DC charger)
• Lock pin position signal indicates lock pin has engaged and EVSE plug is locked to the vehicle charge inlet
• Peak charging current from the CCS EVSE 126 A
• Increased “general” noise level in all signals indicates charging has commenced
Below I have taken another capture at a high sample rate in order to captured the PLC structured noise during the commencement of DC charging
I think we can agree this looks remarkably like the theoretical image of PLC in "Picture 5A above
I have to say, having driven this vehicle for numerous miles and used a variety of CCS EVSE’s, even after the re-flash of J966 there are times where I have to disconnect and reconnect the EVSE plug due to communication failure!
I find that EVSE charge lanes (where there are multiple chargers on one site) are the biggest offenders (possibly due to increased “noise levels”) whereas stand-alone EVSE’s (often found on pub car parks) connect with no issues at all.
My gut feeling is that this vehicle is an early adopter of PLC which does seem vulnerable and fickle when it comes to electrical noise; which is not a good thing when "noise" is the transmission medium that is carrying the data required to commence charging!
The Tesla approach does feel a little more robust as Ben has discussed here topic22544.html
I hope this helps, take care……Steve
If you haven't heard of it before, you might take a glance at SAE J2497, also called PLC4Trucks. The heavy duty trucking industry has been using PLC for communication between tractor and trailer ABS systems for quite a while. It seems to be very reliable.
I didn't realize that a similar PLC system was used between EV Battery Controllers and the charging stations. Learn something new every day!
I have just added SAE J2497 to the "To Do" list and no doubt Ben will be all over this given his outstanding J1939 work
The only issue we all have is trying to remember how things work.
I am not sure about our users, when you are in the thick of diagnosis or research, at the time of completion you feel like you know it and now have it secured away forever.
Roll on 2 weeks after numerous other jobs and that "expert" knowledge is no more than a distant memory containing "flashbacks" at best
This is why the forum is so powerful at documenting diagnostic journeys and theories which we have compiled here for future use topic15231.html
I know time is always against us but I would urge anyone who has made discovery or uses a technique that can help others to document it somewhere (ideally here) or at least write it down such that it can be recalled "just-in-time" when it really counts at the coal face