Supporting Roadster owners!!!

[3PHASE]

I'm trying to get an issue resolved with some older Schneider Electric charging stations at a location in WV (14 of them in fact). Have you ever experienced this? Hoping to get at least a few fixed/replaced so we can support a few Roadster owners who have never been able to successfully charge for an upcoming event in May.

From Schneider support...

This sounds like an issue with the CCID (Charge Circuit Interrupt Device) device. It is intended to detect a fault when the current differences between the phases is greater than 20mA. Prior to late 2012, the CCID setpoint was 5mA (early EV industry Standard). The 5mA setpoint proved to be too sensitive and was thus changed to 20mA. Commercial outdoor units do not have a part number change to indicate the 20mA vs 5mA CCID. Check the serial number of those devices, located on the nameplate on left side of unit. Date codes starting with 11 or 12 and only 5 digits long will have the 5mA CCID setpoint. If the EVSE has a 5mA setpoint, then it should be replaced for a 20mA CCID unit.

If you find the the units have a serial number starting with 11 or 12, we can work with APUS to see if they would like to move forward with correcting this issue.

 

[Curt]

Mark - I have NEVER been able to charge at that location (and I've tried every one of the 14 stations there).

But - you knew that already.

 

[MileHighMotoring]

I wonder if Henry Sharp @hcsharp (The CAN - Roadster to J1772 or Model S/X connector) could possibly create a device to go between the Roadster and the J1772 that would make both happy and allow the car to charge?

 

[gregd]

I wonder if Henry Sharp @hcsharp (The CAN - Roadster to J1772 or Model S/X connector) could possibly create a device to go between the Roadster and the J1772 that would make both happy and allow the car to charge?

Probably not possible, depending on your definition of "happy". Besides the challenge to maintain safety (I'm guessing a separate slightly-more-tolerant GFI on the car's side of the device, with a separate ground), it would somehow need to restore the current balance between the two hots as seen by the EVSE. I don't see how that would be possible, unless the leakage to ground is already equal from both hots.

And, where would that separate ground come from? We certainly don't want to end up with a situation like this:

http://i.imgur.com/rbAo0Vc.jpg

 

[3PHASE]

I'm making some progress. The facility, APUS, is going to take some pictures of a few of the station's serial #'s today and send them to me. I have no idea yet what the manufacturer of the EVSE will do, but I would probably be happy if they fixed or replaced 3 or 4 of them. I will let you know what happens. I could imagine there are more older installations of Schneider Electric charging stations like this around the country. It's just that they never get used by Roadsters. We have a good number (which still is low) of Roadsters in the DC/MD/VA region.

 

[hcsharp]

I'm trying to get an issue resolved with some older Schneider Electric charging stations at a location in WV (14 of them in fact). Have you ever experienced this? Hoping to get at least a few fixed/replaced so we can support a few Roadster owners who have never been able to successfully charge for an upcoming event in May.

From Schneider support...

This sounds like an issue with the CCID (Charge Circuit Interrupt Device) device. It is intended to detect a fault when the current differences between the phases is greater than 20mA. Prior to late 2012, the CCID setpoint was 5mA (early EV industry Standard). The 5mA setpoint proved to be too sensitive and was thus changed to 20mA. Commercial outdoor units do not have a part number change to indicate the 20mA vs 5mA CCID. Check the serial number of those devices, located on the nameplate on left side of unit. Date codes starting with 11 or 12 and only 5 digits long will have the 5mA CCID setpoint. If the EVSE has a 5mA setpoint, then it should be replaced for a 20mA CCID unit.

If you find the the units have a serial number starting with 11 or 12, we can work with APUS to see if they would like to move forward with correcting this issue.

Schneider support was correct except for one thing. The 5mA threshold was never the industry standard when they got into the business. When they started making those units every other EVSE maker was using 20mA and had been for a couple of years. Furthermore they were made aware of their mistake early on (and acknowledged this) but kept right on making the units with the wrong threshold.

Worse, they re-packaged those units under other brand names such as ChargePro which was popular in the greater DC area. The ChargePro salesman was very aggressive all around there.

 

[hcsharp]

I wonder if Henry Sharp @hcsharp (The CAN - Roadster to J1772 or Model S/X connector) could possibly create a device to go between the Roadster and the J1772 that would make both happy and allow the car to charge?

It's funny you ask because I had a CAN JR customer who was attempting to charge (unsuccessfully) outside an office where some Schneider engineers worked. One of the engineers was aware of the GFI issue with their chargers, but had never seen a CAN JR and came over to talk with the Roadster owner. He got excited when he saw it and said he could design electronics for the adapter that would overcome the GFI/CCID problem. Nothing ever came of it.

I did actually build a CAN JR with electronics that circumvented the problem. I only used it once. I never found a solution for the safety issues that gregd predicted, and it was unreliable on the chargers with the correct threshold.

 

[BartJ]

It's funny you ask because I had a CAN JR customer who was attempting to charge (unsuccessfully) outside an office where some Schneider engineers worked. One of the engineers was aware of the GFI issue with their chargers, but had never seen a CAN JR and came over to talk with the Roadster owner. He got excited when he saw it and said he could design electronics for the adapter that would overcome the GFI/CCID problem. Nothing ever came of it.

I did actually build a CAN JR with electronics that circumvented the problem. I only used it once. I never found a solution for the safety issues that gregd predicted, and it was unreliable on the chargers with the correct threshold.

I assume I had the same issue using the CAN JR on Siemens Chargers in Europe.
never got it to work on those.

 

[hcsharp]

I assume I had the same issue using the CAN JR on Siemens Chargers in Europe.
never got it to work on those.

The Siemens chargers made for home use have the same problem. I don't know about their commercial chargers. Unlike Schneider, Siemens has denied they did anything wrong and continues to take a "Everybody's marching out of step but us" attitude, refusing to address the problem. Ironically they used to show a Tesla Roadster 1.5 in their advertising materials and claimed they tested it.

Part of the problem is the SAE J1772 standard does not specify a GFI threshold. They only require that the EVSE have a CCID (charge current interrupting device) that will stop when a ground leak is detected. They don't specify how much current should cause a fault. Household GFI outlets trip at 5mA. That's probably where they're getting that threshold. UL allows 21mA for their testing but they don't punish you if it's less.