Any progress on future AC charging protocols?

widodh · Mar 7, 2015

The current J1772 protocol isn't really suited for EV charging imho.

While it currently does the job, it doesn't handle anything like delayed charging start or anything.

For example, you have a 100A connection (1-phase or 3-phase doesn't matter in this case) and you want to connect 10 chargers to it.

The current way it's done is that each car has to get at least 6A or the car doesn't charge. The J1772 protocol doesn't allow anything like: "Hey, welcome! You're connected now, but I'll let you know when you can start charging!"

Grid connections are a problem and usually also very expensive to upgrade. For condo's, apartments, parking garages, airports it would be great if they could install a large amount of chargers which share a small grid connection.

Using load balancing and such they can charge cars and even prioritize different vehicles.

Is anybody aware of any progress being made on this side?

arg · Mar 13, 2015

The standards-based way to do this (both J1772 and IEC61851) is that you put out the 5% pilot that means "switch to digital communications" and then exchange XML messages over the carrier modulated on the pilot pins.

However, the only use of this to date seems to be for CCS DC charging (even though it's equally defined to be used with the standard shape J1772/type2 connectors, and for AC as well as DC).

There seems to be no sign of Tesla having fitted the appropriate hardware.

widodh · Mar 13, 2015

arg said:
The standards-based way to do this (both J1772 and IEC61851) is that you put out the 5% pilot that means "switch to digital communications" and then exchange XML messages over the carrier modulated on the pilot pins.

However, the only use of this to date seems to be for CCS DC charging (even though it's equally defined to be used with the standard shape J1772/type2 connectors, and for AC as well as DC).

There seems to be no sign of Tesla having fitted the appropriate hardware.

Interesting!

Are those XML messages standardized in any way?

Or for AC charging there is simply no standard yet? It would be very, very useful to have this however. Smart Grids could really benefit from this.

arg · Mar 13, 2015

widodh said:
Interesting!

Are those XML messages standardized in any way?

Or for AC charging there is simply no standard yet? It would be very, very useful to have this however. Smart Grids could really benefit from this.

I think so, though I rather lost interest when it became clear that Tesla aren't using the PLC communication, and I can't remember how far I got in chasing the trail of references from standard to standard.

There's some interesting stuff here: http://openv2g.sourceforge.net/

sefs · Mar 13, 2015

You could do this right now with J1772. The EVSE can decrease/increase the amapacity it advertises during the charging session. Networked EVSE's could monitor the main feeder utilization and vary the advertised amapacity to not exceed a set limit. The EVSE could increase the pilot signal duty cycle until it saw the car wasn't taking any more current. This would be that car's max current draw. Now, it could start ramping up another EVSE. You could even have 2, 3, 4, whatever charging cords on the same circuit. This would save a bunch of money on wiring dedicated circuits. If someone turns on an oven or something, the network sees the high draw on the feeders, and orders the EVSE's to decrease their pilot signals accordingly. This is totally possible, I just don't think many places would want to pay for it right now.

The other end of the spectrum is load shedding of the other devices in favor of the EVSE's. You can do this with contactors and a load shedding device. PSP Products makes some nice load shedding equipment. I used their contactors for my generator installation.

Saghost · Mar 13, 2015

arg said:
The standards-based way to do this (both J1772 and IEC61851) is that you put out the 5% pilot that means "switch to digital communications" and then exchange XML messages over the carrier modulated on the pilot pins.However, the only use of this to date seems to be for CCS DC charging (even though it's equally defined to be used with the standard shape J1772/type2 connectors, and for AC as well as DC).There seems to be no sign of Tesla having fitted the appropriate hardware.

Tesla does exactly the same thing - but differently. The Model S complies with J1772 signaling in all respects AFAIK (only the connector is different.) However, J1772 is purely an AC standard - it doesn't cover the digital side.Tesla Superchargers use the same 5% cycle to go digital - and then handshake with the car using CANBus protocols.

arg · Mar 13, 2015

Saghost said:
Tesla does exactly the same thing - but differently. The Model S complies with J1772 signaling in all respects AFAIK (only the connector is different.) However, J1772 is purely an AC standard - it doesn't cover the digital side.Tesla Superchargers use the same 5% cycle to go digital - and then handshake with the car using CANBus protocols.

Actually, J1772 is not purely an AC standard - J1772 (2012 edition) has lots of text in it describing DC charging - both via the contacts in the standard connector ('DC level 1') and via the extra contacts in the combo connector. It doesn't itself contain the modulation scheme or the syntax of the XML messages, but it does reference other standards which define those and lists which XML messages you need to use in which sequence to perform DC charging (and their relationship to things like the mechanical lock on the connector, isolation monitoring etc.).

Some of the messages are in J2847, but you need a bookshelf full of documents to get the whole picture.

So it's not true to say that the Model S complies with J1772 signalling in all respects: the 5% pilot is defined by J1772 as going to a particular digital signalling scheme, but the Model S goes to a different one.

Saghost · Mar 13, 2015

arg said:
Actually, J1772 is not purely an AC standard - J1772 (2012 edition) has lots of text in it describing DC charging - both via the contacts in the standard connector ('DC level 1') and via the extra contacts in the combo connector. It doesn't itself contain the modulation scheme or the syntax of the XML messages, but it does reference other standards which define those and lists which XML messages you need to use in which sequence to perform DC charging (and their relationship to things like the mechanical lock on the connector, isolation monitoring etc.).Some of the messages are in J2847, but you need a bookshelf full of documents to get the whole picture.So it's not true to say that the Model S complies with J1772 signalling in all respects: the 5% pilot is defined by J1772 as going to a particular digital signalling scheme, but the Model S goes to a different one.

Huh. I hadn't read that before, and am not sufficiently involved to have read the full spec itself. Did they add the DCFC pieces to it in the 2012 revision?

widodh · Mar 13, 2015

sefs said:
You could do this right now with J1772. The EVSE can decrease/increase the amapacity it advertises during the charging session. Networked EVSE's could monitor the main feeder utilization and vary the advertised amapacity to not exceed a set limit. The EVSE could increase the pilot signal duty cycle until it saw the car wasn't taking any more current. This would be that car's max current draw. Now, it could start ramping up another EVSE. You could even have 2, 3, 4, whatever charging cords on the same circuit. This would save a bunch of money on wiring dedicated circuits. If someone turns on an oven or something, the network sees the high draw on the feeders, and orders the EVSE's to decrease their pilot signals accordingly. This is totally possible, I just don't think many places would want to pay for it right now.

The other end of the spectrum is load shedding of the other devices in favor of the EVSE's. You can do this with contactors and a load shedding device. PSP Products makes some nice load shedding equipment. I used their contactors for my generator installation.

Yes, I'm aware of that. But you can't schedule charging.

Say that your EVSE knows that the power becomes cheap at a certain point:
- Welcome vehicle, please standby.....
- Wait 4 hours
- Yes, you can charge now!

That's not covered currently with the J1772 protocol used for AC charging.

sefs · Mar 13, 2015

My EVSE has this built right in. Clippercreek CS-90. There are inputs to inhibit charging until the input goes off or restrict it to 6A.

Clippercreek Manual: https://www.clippercreek.com/pdf/CS-100%20User%20Manual%2075A%20DLP%20091124%20v01.pdf
Page 16 "Load Management Inputs"

arg · Mar 13, 2015

widodh said:
Yes, I'm aware of that. But you can't schedule charging.

Say that your EVSE knows that the power becomes cheap at a certain point:
- Welcome vehicle, please standby.....
- Wait 4 hours
- Yes, you can charge now!

That's not covered currently with the J1772 protocol used for AC charging.

That scenario can be kludged by just not offering the pilot until later.

But the thing that's really lacking to do this properly is a means to find out how much charge the vehicle wants.

Even today with simple time-of-day tarrifs, you might need to start charging before the cheap rate period starts to guarantee the charge completes in time to depart the next morning - but usually the car won't be empty so you want to wait for the cheap rate period. Or your other scenario of a hotel carpark full of cars that need charging by the next morning: when the charging starts, you probably have more demand than your supply can cope with, so you either have to limit them to a very low rate or charge some cars before others, yet by the end of the charging period most of the cars will be charged and you have spare capacity going to waste; yet you may have one car that's not charged because it came in empty and didn't get full power until too late. Even in an all-Tesla (ie. long range EV) environment, most cars won't arrive dead on empty - but unless you know which ones need the charge you can't manage things for maximum efficiency.

And in the future we are bound to have more variable rates (more than 2 time bands, or fully dynamic rates) - can you afford to wait until 03:00 when the super cheap rate is going to start, or do you need to start charging at 12:00 when it gets somewhat cheap?

widodh · Mar 14, 2015

arg said:
That scenario can be kludged by just not offering the pilot until later.

But the thing that's really lacking to do this properly is a means to find out how much charge the vehicle wants.

Even today with simple time-of-day tarrifs, you might need to start charging before the cheap rate period starts to guarantee the charge completes in time to depart the next morning - but usually the car won't be empty so you want to wait for the cheap rate period. Or your other scenario of a hotel carpark full of cars that need charging by the next morning: when the charging starts, you probably have more demand than your supply can cope with, so you either have to limit them to a very low rate or charge some cars before others, yet by the end of the charging period most of the cars will be charged and you have spare capacity going to waste; yet you may have one car that's not charged because it came in empty and didn't get full power until too late. Even in an all-Tesla (ie. long range EV) environment, most cars won't arrive dead on empty - but unless you know which ones need the charge you can't manage things for maximum efficiency.

And in the future we are bound to have more variable rates (more than 2 time bands, or fully dynamic rates) - can you afford to wait until 03:00 when the super cheap rate is going to start, or do you need to start charging at 12:00 when it gets somewhat cheap?

Some cars will fail if they don't get a pilot right away. They throw an error to the screen and such.

If a car connects to a AC charger it would state it's SoC or how many kWh it wants, next to that it could also tell how many Amps it can draw so we can calculate how long a charge would take.

So yes, changes are required, but as long as the car tells the AC charging point more about it's state the charging point can adjust accordingly.

I don't see any cheap ~10kW CHAdeMO or CCS chargers popping up at peoples home any time soon.

muleferg · Mar 15, 2015

Nissan Leaf

AT HOME AFTER DRIVING Charging the Li-ion battery When you return home, connect the vehicle to the charging station installed at your home using the normal charge connector. Charge the vehicle or set the charging timer function in the navigation system to have the vehicle charge at a specific time. See “Charging timer” in the “CH. Charging” section. 1. When the power switch is turned off, the ON/OFF settings of the charging timer and the Climate Ctrl. Timer functions are dis- played on the dot matrix liquid crystal display. See “Dot matrix liquid crystal dis- play” in the “2. Instruments and controls” section. 2. Open the charge port lid and charge port cap. See “Charge port lid” in the “3. Pre- driving checks and adjustments” section. 3. Connect the charge connector to the vehicle. 4. When a charging timer is turned on, char- ging starts at the set time. When a charging timer is not turned on, charging starts immediately

Search

Any progress on future AC charging protocols?

widodh

Model S 100D and Y LR

arg

Active Member

widodh

Model S 100D and Y LR

arg

Active Member

sefs

Member

Saghost

Well-Known Member

arg

Active Member

Saghost

Well-Known Member

widodh

Model S 100D and Y LR

sefs

Member

arg

Active Member

widodh

Model S 100D and Y LR

muleferg

Active Member

Similar threads