CCS Adapter for North America

[bradtem]

EVgo is there to help NRG pay a fine.

Speak more. I know it was spun out of NRG but don't know these details. I know they now feel they need to make a profit selling electricity, as do some others. Volta says they are making money selling ads instead of electricity. (Though they are not greatly incentivized to repair bad stations and most of the stations charge at a nearly useless rate, but I am told they are putting in fast stations.) Flo and Blink have aspirations of selling kwh but Flo mostly acts like Chargepoint (which does not sell kwh, only services and stations)

 

[jsight]

EVgo is there to help NRG pay a fine.

I don't think they've been a part of NRG for a few years now. They are now a publicly traded company as well and are expanding reasonably quickly.

TBH, this space is starting to get more interesting financially. I thought Greenlots was dying, but I now see that Shell is making a renewed effort at expansion. Their new locations in South Carolina actually make a lot of sense to me too.

 

[e of pi]

Oh, the CCS/CdM networks are not nearly as reliable as the Tesla supercharging network. Some of that is Tesla doing a good job. Some of it is the fact that Tesla stations have large numbers of chargers, so even if one breaks it's not a big deal. Some is the fact that almost no fast charging network is a business, not even Tesla's, and so the incentives to keep them maintained are wrong. Tesla chargers are there to sell cars, not electricity. EA is there to help VW pay a fine. EVGo is trying to be a business and a few others.

Believe it or not, EA has notably better station availability via Plugshare than EVGo. Ranking only >70 kW capable stations with a Plugscore, EVGo has an network average plugscore of 7.8, Electrify Canada has an 8.4, Electrify America has a 9.0 (rounded from 8.997) and "all the rest" of the CCS stations meeting that criteria average 8.7. I don't have Plugscore data for the Tesla network because Plugshare's interface makes it a pain to count that many stations and I don't typically need to care about it, but I would expect it's also in the 9+ range. Still, a station with an 8 or especially a 9 and 4 or more stalls typically means you will be able to get a good charge when you arrive, and the network average for Electrify America means they're average station is well within that range--which matches my personal experience using them.

 

[bradtem]

Believe it or not, EA has notably better station availability via Plugshare than EVGo. Ranking only >70 kW capable stations with a Plugscore, EVGo has an network average plugscore of 7.8, Electrify Canada has an 8.4, Electrify America has a 9.0 (rounded from 8.997) and "all the rest" of the CCS stations meeting that criteria average 8.7. I don't have Plugscore data for the Tesla network because Plugshare's interface makes it a pain to count that many stations and I don't typically need to care about it, but I would expect it's also in the 9+ range. Still, a station with an 8 or especially a 9 and 4 or more stalls typically means you will be able to get a good charge when you arrive, and the network average for Electrify America means they're average station is well within that range--which matches my personal experience using them.

I have only used EA twice, and had issues with billing both times. Indeed, one of the things that makes Tesla the leader is that billing is frictionless and works 99.99% of the time. A large fraction of the problems I see on plugshare with DC fast stations are billing problems, not station problems. Some of the billing problems are customer mistakes, which is why Tesla is so good, it is not possible for the customer to make a mistake.
Maybe plug and charge will help fix that, 10 years after Tesla showed how to do it.

I really think that the charging stations should have taken the approach of "fail operational" -- design so that in the first year or two, that it gives you free power if there is a billing issue, and they eat it, but at least the user is happy. Maybe stick in a camera to record the licence plate to find out if anybody is trying to abuse it, but never let a customer leave unable to charge unless the hardware has failed. Though many failures are also communication problems not just billing.
Even Tesla missed this one and left a customer stranded when there was an error (tesla caused) in his credit card records.

Sure, you will lose some money with fail operational -- but that's a good incentive to make the billing work well. And no, you don't just give free charging -- something has to be tried, like an app, credit card or card or phone call.

 

[jsight]

Believe it or not, EA has notably better station availability via Plugshare than EVGo. Ranking only >70 kW capable stations with a Plugscore, EVGo has an network average plugscore of 7.8, Electrify Canada has an 8.4, Electrify America has a 9.0 (rounded from 8.997) and "all the rest" of the CCS stations meeting that criteria average 8.7. I don't have Plugscore data for the Tesla network because Plugshare's interface makes it a pain to count that many stations and I don't typically need to care about it, but I would expect it's also in the 9+ range. Still, a station with an 8 or especially a 9 and 4 or more stalls typically means you will be able to get a good charge when you arrive, and the network average for Electrify America means they're average station is well within that range--which matches my personal experience using them.

Evgo has a pretty bad track record, but they also have a history of installing little single stall 50kw stations years ago. I'd expect their newer stations to be more reliable as they have a higher number of stalls and seem to be a bigger priority in general.

In South Carolina, they've actually sold some of their old sites to Blink, which is the one network that is consistently worse than evgo, lol.

 

[Mockingbird]

I have only used EA twice, and had issues with billing both times. Indeed, one of the things that makes Tesla the leader is that billing is frictionless and works 99.99% of the time. A large fraction of the problems I see on plugshare with DC fast stations are billing problems, not station problems. Some of the billing problems are customer mistakes, which is why Tesla is so good, it is not possible for the customer to make a mistake.
Maybe plug and charge will help fix that, 10 years after Tesla showed how to do it.

I really think that the charging stations should have taken the approach of "fail operational" -- design so that in the first year or two, that it gives you free power if there is a billing issue, and they eat it, but at least the user is happy. Maybe stick in a camera to record the licence plate to find out if anybody is trying to abuse it, but never let a customer leave unable to charge unless the hardware has failed. Though many failures are also communication problems not just billing.
Even Tesla missed this one and left a customer stranded when there was an error (tesla caused) in his credit card records.

Sure, you will lose some money with fail operational -- but that's a good incentive to make the billing work well. And no, you don't just give free charging -- something has to be tried, like an app, credit card or card or phone call.

Electrify America gives free charging if the chargers lose contact with its backend (i.e. network issues), but not for declined payments.

 

[FalconFour]

CCS connector does not lock to the Setec adapter, and you can just pull it out directly, which creates an arc risk.

Eh. The latch is there. You have to push the button to pull it out. Can't just yank on it. Car locks the Tesla side, and the CCS handle locks the CCS side. Only way out is to press the button.

There is also a claim it ignores the pilot signal, so does not cut off the power even when the latch is activated.

Proximity "pilot" isn't normally referred to as "pilot" (normally control pilot is), but ... even then, yeah, it may be the case that the adapter is a POS and ignores the prox signal (because it's not essential for "making it work", it tends to get ignored).

However, even with that, it still needs to provide +5v, which as it turns out, is actually required for the EVSE (station) to be listening to as well - so the station can sense if the latch is pressed. Which means: you guessed it, it'll cut power when it sees the button being pressed to pull the handle out.

None of this is to say it's a good solution - but just quelling the hellfire preaching that "oh no it could arc and kill your babies if someone trips over it". If someone trips over it, they ought to die of an ass-beating for being stupid and tripping over an obvious, extremely visible, and short, cable before dying of an arc or whatever But more likely, a cat breathing in the general direction of the charging car would break it off first... because those CCS cables are WAY too thick for that tiny widdle Tesla stub plug to hold up!

There should be a locking pin to keep the latch from being pushable.
Of course the Setec adapter also ought to be listening to the prox signal for the button to be pressed, and pass that along to the car as an "emergency stop".

 

[masotime]

I have only used EA twice, and had issues with billing both times.

I’ve only had issues with EA once, in Bridgeport, CA. Otherwise, no issues having used them 33 times in total. I’m on the Pass+ program, and added my EA card to my iPhone’s NFC Wallet - makes charging pretty straightforward. I have more issues wrangling the clumsy Setec CCS adapter - I’m really looking forward to the official Tesla one.

 

[stopcrazypp]

Eh. The latch is there. You have to push the button to pull it out. Can't just yank on it. Car locks the Tesla side, and the CCS handle locks the CCS side. Only way out is to press the button.
View attachment 782381


Proximity "pilot" isn't normally referred to as "pilot" (normally control pilot is), but ... even then, yeah, it may be the case that the adapter is a POS and ignores the prox signal (because it's not essential for "making it work", it tends to get ignored).

However, even with that, it still needs to provide +5v, which as it turns out, is actually required for the EVSE (station) to be listening to as well - so the station can sense if the latch is pressed. Which means: you guessed it, it'll cut power when it sees the button being pressed to pull the handle out.

But I think the problem isn't just cutting power from the station, it's cutting power from the HV battery also. If the car doesn't get the signal to shut down, the pins may remain energized and that can cause arcing.
I remember a thread below on the Bolt forums talking about CCS chargers where arcing has demonstratively happened due to the fact the latches did not stay locked until the connection was fully deenergized (that the latch was pressed before pulling the plug apparently did not prevent arcing, the OP said the latch tangs were not broken, the issue was they don't stay locked to the car before car is deenergized):
Some Electrify America CCS chargers have defective high...
Apparently there is a need for current ramp down (the above thread discusses this). So that means either the latch staying locked until connection was fully deenergized, or the user deliberately waiting a few seconds after pressing the release before actually disconnecting. I remember @Ingineer mentioning a current ramp down in talking about the PCS (DC-to-DC) in one of his videos, that the current on the HV portion is ramped down to close to zero before contactors activate. I imagine given @rhuber got advice from him before making that adapter, he knows what he is talking about in terms of arc risk with the Setec adapter. Probably a similar issue comes to play in the CCS connection if you immediately unplug after pressing the latch (if either chimes in maybe they can comment if it does).

None of this is to say it's a good solution - but just quelling the hellfire preaching that "oh no it could arc and kill your babies if someone trips over it". If someone trips over it, they ought to die of an ass-beating for being stupid and tripping over an obvious, extremely visible, and short, cable before dying of an arc or whatever But more likely, a cat breathing in the general direction of the charging car would break it off first... because those CCS cables are WAY too thick for that tiny widdle Tesla stub plug to hold up!

There should be a locking pin to keep the latch from being pushable.
Of course the Setec adapter also ought to be listening to the prox signal for the button to be pressed, and pass that along to the car as an "emergency stop".

As above, it doesn't require tripping over a cable, just simply pulling the cable immediately after pressing the latch (which plenty of people can do if they are in a hurry).
The Tesla adapter does have a way to keep the CCS connection locked until the connection is deenergized (that little pin sticking out that everyone early on speculated if it was a photoshop error or if it actually has a use). Setec has no such mechanism (nor a way to pass through any disconnect messages apparently that can give an early warning of sorts).

 

[STS-134]

But I think the problem isn't just cutting power from the station, it's cutting power from the HV battery also. If the car doesn't get the signal to shut down, the pins may remain energized and that can cause arcing.
I remember a thread below on the Bolt forums talking about CCS chargers where arcing has demonstratively happened due to the fact the latches did not stay locked until the connection was fully deenergized (that the latch was pressed before pulling the plug apparently did not prevent arcing, the OP said the latch tangs were not broken, the issue was they don't stay locked to the car before car is deenergized):
Some Electrify America CCS chargers have defective high...
Apparently there is a need for current ramp down (the above thread discusses this). So that means either the latch staying locked until connection was fully deenergized, or the user deliberately waiting a few seconds after pressing the release before actually disconnecting. I remember @Ingineer mentioning a current ramp down in talking about the PCS (DC-to-DC) in one of his videos, that the current on the HV portion is ramped down to close to zero before contactors activate. I imagine given @rhuber got advice from him before making that adapter, he knows what he is talking about in terms of arc risk with the Setec adapter. Probably a similar issue comes to play in the CCS connection if you immediately unplug after pressing the latch (if either chimes in maybe they can comment if it does).

As above, it doesn't require tripping over a cable, just simply pulling the cable immediately after pressing the latch (which plenty of people can do if they are in a hurry).
The Tesla adapter does have a way to keep the CCS connection locked until the connection is deenergized (that little pin sticking out that everyone early on speculated if it was a photoshop error or if it actually has a use). Setec has no such mechanism (nor a way to pass through any disconnect messages apparently that can give an early warning of sorts).

This is yet another reason the US should ditch CCS1 and go to CCS2 ASAP. At least there's no real damage done if J1772 is disconnected before it can be de-energized (in fact I've done this before, by pushing on the latch and pulling the connector out as quickly as possible; the pins smelled like ozone for a while after that). But on J3068/CCS2, there's no way to do this, since the locking mechanism is part of the vehicle.

 

[Jeff N]

But I think the problem isn't just cutting power from the station, it's cutting power from the HV battery also. If the car doesn't get the signal to shut down, the pins may remain energized and that can cause arcing.
I remember a thread below on the Bolt forums talking about CCS chargers where arcing has demonstratively happened due to the fact the latches did not stay locked until the connection was fully deenergized (that the latch was pressed before pulling the plug apparently did not prevent arcing, the OP said the latch tangs were not broken, the issue was they don't stay locked to the car before car is deenergized):
Some Electrify America CCS chargers have defective high...
Apparently there is a need for current ramp down (the above thread discusses this). So that means either the latch staying locked until connection was fully deenergized, or the user deliberately waiting a few seconds after pressing the release before actually disconnecting. I remember @Ingineer mentioning a current ramp down in talking about the PCS (DC-to-DC) in one of his videos, that the current on the HV portion is ramped down to close to zero before contactors activate. I imagine given @rhuber got advice from him before making that adapter, he knows what he is talking about in terms of arc risk with the Setec adapter. Probably a similar issue comes to play in the CCS connection if you immediately unplug after pressing the latch (if either chimes in maybe they can comment if it does).

As above, it doesn't require tripping over a cable, just simply pulling the cable immediately after pressing the latch (which plenty of people can do if they are in a hurry).
The Tesla adapter does have a way to keep the CCS connection locked until the connection is deenergized (that little pin sticking out that everyone early on speculated if it was a photoshop error or if it actually has a use). Setec has no such mechanism (nor a way to pass through any disconnect messages apparently that can give an early warning of sorts).

Okay, I read that thread on the Bolt forum and do not assign high credibility to it. The person making this claim seemed to be making several unwarranted assumptions and was a bit hysterical.

There is a vehicle-controlled latch on the car itself which locks the plug into the car during DC charging. The car will not begin charging if it senses that this latch was unsuccessful in locking the plug to the car. This is an infamous cause of problems with Bolt EVs charging at EA and other chargers that use liquid-cooled cables. The heavier and less flexible cables cause the plug to sag enough in the car’s socket that the car is unable to successfully latch the plug to the car as required for charging. This causes a charge session initiation error. The solution is to support the cable near the car for a few seconds during session establishment so the latch can lock the plug onto the car.

So, the car shouldn’t be unlatching the plug from the car until the contactors to the HV battery are open (disconnected) and the current is thus obviously zero. Thus if there really was an arc when this person rapidly unplugged their car from CCS charging it would have been a failure by the car and not the EA charger.

Given the totality of the claims tossed around, I’m not giving that report much credence until I hear about other similar independent reports.

I don’t have personal experience with either the EVHUB or the Setec adapter but I don’t see either one apparently implementing the equivalent of the CCS vehicle-side latch while the official Tesla CCS adapter apparently does by using that pin that compresses against the vehicle port when inserting the adapter. That would seem to be a violation of the intent of the CCS specification. The whole point of the CCS vehicle-side latch is that the plug-side latch is insufficient for fully preventing arcing issues during DC charging if someone tries to abruptly unplug during a charging session.

 

[Earl]

US should ditch CCS1 and go to CCS2 ASAP.

Please NO! A battle between 3 incompatible standards is bad enough. Let's not introduce a 4th completely incompatible one as well!

 

[Earl]

I don't think they've been a part of NRG for a few years now.

I don't know that they were ever a part of NRG, any more than EA is a part of Volkswagen. Aren't/weren't those both independent companies funded by the penance of these companies.
I am happy that EVgo has seemed to be making a few good decisions including doing maintenance on the lousy Nissan CHAdeMO chargers and supporting Tesla. My only problem with them is that they only do urban chargers, nothing (or very little) on the open road.

 

[STS-134]

Please NO! A battle between 3 incompatible standards is bad enough. Let's not introduce a 4th completely incompatible one as well!

CCS2 already exists. Getting rid of (by phasing out) CCS1 would reduce the number of standards by one. But we'd still have 5 more (CCS2, CHAdeMO, Tesla, GB/T, ChaoJi). And I believe they're developing yet another one that can charge commercial vehicles at > 1MW.

 

[stopcrazypp]

This is yet another reason the US should ditch CCS1 and go to CCS2 ASAP. At least there's no real damage done if J1772 is disconnected before it can be de-energized (in fact I've done this before, by pushing on the latch and pulling the connector out as quickly as possible; the pins smelled like ozone for a while after that). But on J3068/CCS2, there's no way to do this, since the locking mechanism is part of the vehicle.

As mentioned, in a DC session, the car does lock the CCS1 connector to the car until it is deenergized. The Bolt thread I posted was describing a defective situation (arguable if the car was defective or connector though).

In an AC session, probably the PP signal is "good enough" given the max current is 80A. In DC, it can reach hundreds of amps though, so is more dangerous.

 

[bradtem]

CCS2 already exists. Getting rid of (by phasing out) CCS1 would reduce the number of standards by one. But we'd still have 5 more (CCS2, CHAdeMO, Tesla, GB/T, ChaoJi). And I believe they're developing yet another one that can charge commercial vehicles at > 1MW.

Actually, if not for egos, it seems pretty clear that the ideal universal standard would be the Tesla connector. It is the most common connector on cars in the USA and several other countries. It would be in Europe if laws had not pushed it away. China is its own thing.

The Tesla connector is superior in just about every way, though recently CCS has provided slightly higher wattage. But it's more compact, does all levels of charging and does high power too and is very easy to use. Well, I don't need to convince people at TMC about this.

The one thing I would improve is I would try to find a way to put in on the front or back of cars. Why? If it were there, a Tesla could plug itself in to a plug mounted on a fixed pole at the right height, with some slight spring to deal with the angle of insertion. Teslas could drive themselves in and out of supercharger spots and other charging stations. The best charging is where you literally don't have to do anything, your car is just always magically full unless you drive more than 250 miles in one day.

 

[stopcrazypp]

Okay, I read that thread on the Bolt forum and do not assign high credibility to it. The person making this claim seemed to be making several unwarranted assumptions and was a bit hysterical.

There is a vehicle-controlled latch on the car itself which locks the plug into the car during DC charging. The car will not begin charging if it senses that this latch was unsuccessful in locking the plug to the car. This is an infamous cause of problems with Bolt EVs charging at EA and other chargers that use liquid-cooled cables. The heavier and less flexible cables cause the plug to sag enough in the car’s socket that the car is unable to successfully latch the plug to the car as required for charging. This causes a charge session initiation error. The solution is to support the cable near the car for a few seconds during session establishment so the latch can lock the plug onto the car.

So, the car shouldn’t be unlatching the plug from the car until the contactors to the HV battery are open (disconnected) and the current is thus obviously zero. Thus if there really was an arc when this person rapidly unplugged their car from CCS charging it would have been a failure by the car and not the EA charger.

Given the totality of the claims tossed around, I’m not giving that report much credence until I hear about other similar independent reports.

I don’t have personal experience with either the EVHUB or the Setec adapter but I don’t see either one apparently implementing the equivalent of the CCS vehicle-side latch while the official Tesla CCS adapter apparently does by using that pin that compresses against the vehicle port when inserting the adapter. That would seem to be a violation of the intent of the CCS specification. The whole point of the CCS vehicle-side latch is that the plug-side latch is insufficient for fully preventing arcing issues during DC charging if someone tries to abruptly unplug during a charging session.

I don't know if it is the car or the EA connectors in that case, but in either case, the point I wanted to make is that the locking mechanism being discussed really does have practical use in preventing arcing, and that it's not just a myth that arcing can happen in the absence of it. And this is happening even with a Bolt which maxes out at about 150A (to be fair to Setec, by nature of the CHAdeMO emulation they are limited to similar speeds). A Tesla that can do hundreds of amps during CCS DC charging is at a even higher risk.

 

[Mockingbird]

This is yet another reason the US should ditch CCS1 and go to CCS2 ASAP. At least there's no real damage done if J1772 is disconnected before it can be de-energized (in fact I've done this before, by pushing on the latch and pulling the connector out as quickly as possible; the pins smelled like ozone for a while after that). But on J3068/CCS2, there's no way to do this, since the locking mechanism is part of the vehicle.

The ship has sailed.

If Tesla has shipped Tesla Model 3 with CCS2 in North America, things would have been different, but Tesla didn't.

 

[Mockingbird]

CCS2 already exists. Getting rid of (by phasing out) CCS1 would reduce the number of standards by one. But we'd still have 5 more (CCS2, CHAdeMO, Tesla, GB/T, ChaoJi). And I believe they're developing yet another one that can charge commercial vehicles at > 1MW.

CCS2 doesn't exist in US/CAN and neither do GB/T and ChaoJi.

 

[FalconFour]

isn't just cutting power from the station, it's cutting power from the HV battery also

Doesn't matter - arcing happens due to electrical "inertia" or by voltage jumping a gap by proximity... and since the station and battery are at the same voltage, there's no arcing that can happen if the current is zero. That's the whole principle of unplugging J1772. The relay doesn't always open until after unplugging (though Tesla implements it that way, where it does command EVSE to open relay when it senses prox switch press, not all do - particularly the Leaf).

So voltage != arcing. All the station has to do is stop sending current when it senses the latch button press, and boop, you can unplug. Now, whether you can touch the terminals immediately after disconnecting, hmm. Highly inadvisable. But arcing? Non-issue imho.

BTW, the station doesn't NEED to ramp current. It can jump straight to zero in a handful of milliseconds, the time it takes your muscles to move from pressing to pulling the latch. Don't underestimate the speed of microcontrollers Have you noticed that you can unplug Tesla's Supercharger connector near-instantly when you push the button to stop it? It stops, opens the relays, determines it's safe, AND UNLOCKS, all in the time you can barely register that the button has been pressed hard enough to click the little button inside. Fraction of a second.

Same with CCS and the handle latch button.