My L1 ESVE findings.....

[garygid]

1. Yes, the Mod-L1 EVSE by Ingineer does not meet all standards, mostly because ...

2. The un-modified L1 EVSE as made by Panasonic and sold by Nissan does not meet all the J1772 requirements (no diode check).

3. For my info, what SAE "standard" requires hardwire of all EVSEs?

4. Trying to be helpful: NEC 625 specifically DOES allow for design of plug-in L2 (240v) EVSEs. It might be in Section 625.13 if I remember correctly.

5. Of course, the NEC codes are often modified when they are (partially) adopted by different states.

6. Also, local Building Department regulations and customs can vary substantially.

[omkar]

For those with kind words, there will be more to come.

Hope to read more!

[jwallace3]

This thread has become quite lengthy. Here's the results thus far:

Without depotting the board, I swapped my 120VAC 3 wire pigtail for a 3 wire 240 20 amp cable and connector. Originally I completely bypassed all off the on board relays and GFI CT and wired 240 directly to the J1772 connector and yes it worked. I DON'T RECOMMEND ANYONE TRY THIS AT HOME AND AM NOT RESPOSIBLE FOR ANYONE THAT DOES! This made the end of the connector always hot and if you licked it you could die. But it did work. Next, trying to maintain the safety features of the device I wired one hot and the neutral/ground to the input lugs just like the standard 120 connector and then jumpered the neutral lug to the ground lug going to the J1772 connector. This had to be done because my 240 connector is only a 3 wire, not 4. 4 is more ideal in the even that you have a loose neutral wire. This is why most modern dryer outlets are 4 wire, not 3. Taking my old 120 pigtail, I wired the loose end to a female 240VAC/20 amp connector. And of course this worked the way it was supposed to.

Using a Form C relay I wired the common contact to the J1772 connector, the normally closed contact to the neutral lug that originally went to the J1772 connector (this is after the relays on the board and GFI CT) and then I wired the normally open contact to the second 120 leg looping it through the GFI CT. The coil I wired from the input neutral lug to the second 120 leg, so if no second 120 leg is present for 240 charging, the common is connected to the normally closed contact which is neutral and the L1 charging happens. If the second 120 leg is present the relay closes and common is now connected to the normally open contact which connects the second 120 leg to the J1772 cable while looping through the CT. There is one downside to this and that is that there is a hot leg present on the J1772 connector as soon as it's plugged in. The solution to this would be to connect the neutral side of the relay coil to the ouput lug that originally went to the J1772 connector, only then would the relay close the normally open contact and supply 240 once the ESVE closed the relay on the board to supply hot and neutral. I didn't go this second route as of yet because there were some issues encountered.

Back to results, so with this setup L1 works fine, as soon as my installed relay energizes when 240 is appled, I get a solid red fault light. If I remove the second 240 leg from the CT, I get a blinking Fault led. To see if it was the GFI circuit or some other circuit thats potted in the board that I can't see, I split the problem. I snipped the jumpers which were after the relays and then run through the CT. Then using the output of the relays and bypassing the CT I tried using that route and L1 threw a fault as well as L2. So, my conclusion is that there is another circuit that does some integrity checking which is why I will depot and locate what the deal is so I can report if continuing down this path is possible without tearing your ESVE apart and making a complicated mod. Without seeing any code for the device as well it will make it difficult to "see" exactly what the ESVE is measuring and what the limits are. I will make progress, and it may be slow, but I will report on the findings as they come about.

[chris1howell]

As far as long term plans go, there should be a simpler solution for a mod that includes all of the safety feature, unfortunately I will have to depot to find out whats going on internally. I don't think the CT likes seeing more current than the 12 amps on 1 phase. This is my main problem. At this point I do plan to continue the mod effort, but only to let others know if there is an easy mod. For me it would be easier just to design my own ESVE and remove my J1772 pigtail from the panasonic module. Might even be cheaper to.

For a GFCI application the CT does not care if it is 16 or 12 Amps, it works by measuring the difference of "what's going in" vs. "what's going out"... In both cases, if you do not reroute any current, you should have 12 in 12 out ct reading 0 or 16 in 16 out ct reading 0.

[jwallace3]

Absolutely correct! Very similar to how we use CTs in a bus differential scheme, if the sum of the currents entering and exiting a node are not zero from our measuring points, then current is "leaking" somewhere. I'm not even doing the 16 amp thing yet, just trying to find a safe, lesser expensive 12 amp L2 alternative.

As far as long term plans go, there should be a simpler solution for a mod that includes all of the safety feature, unfortunately I will have to depot to find out whats going on internally. I don't think the CT likes seeing more current than the 12 amps on 1 phase. This is my main problem. At this point I do plan to continue the mod effort, but only to let others know if there is an easy mod. For me it would be easier just to design my own ESVE and remove my J1772 pigtail from the panasonic module. Might even be cheaper to.

For a GFCI application the CT does not care if it is 16 or 12 Amps, it works by measuring the difference of "what's going in" vs. "what's going out"... In both cases, if you do not reroute any current, you should have 12 in 12 out ct reading 0 or 16 in 16 out ct reading 0.

[chris1howell]

chris1howell wrote:
I believe the ground check uses the 3rd smaller relay (3rd picture far right) to pass a small current from L1 Hot to ground through a 5k or so resistor. If you listen carefully you can hear a clicking right after connection before the main relays engage. I do not think it is checking to see if Neutral and Ground are tied together only that there is a path to ground. Routing current through the 5k resistor would be enough to cause a ground fault so.... my theory is the EVSE powers the 3rd relay causing current to flow to Ground causing the GFCI to read a fault if so, ground and the GFCI CT coil is good. If ground is bad no current flows and the GFCI does not read a fault and the EVSE goes into Fault state.

tps wrote:
I thought about that, but wouldn't that cause an upstream GFCI outlet which was supplying the EVSE to trip? My calcuation is that 5K would cause 24 mA to flow, which would definitely cause the GFCI to trip out. I'm guessing the current which is injected comes from the secondary of the 20 volt transformer. With this isolated source, they could put current into the ground, splitting the return current equally on the line and neutral, thus allowing a flow of current to test ground continuity while maintaining current balance on the L and N to prevent tripping upstream GFCI.

Edit: Now I think of it, that wouldn't work. Since return current it the same phase in L and N, it would add rather than cancel to zero in the CT core, so it would cause the external GFCI to trip...

Maybe it is a resistance of 100k from hot to ground. That would not trip GFCI but the voltage generated by the CT could be measured to verify that current is passing to ground... Without taking it apart it’s just a guess....

[jwallace3]

I was thinking of something similar to that as well. I would really dig into the standard and probe and make guesses, but I'm at the point where I might as well depot so I can give the folks, who would like to know if this is possible, easy, and inexpensive, some results. If I find that the mod would approach being too difficult for the average person or the cost approaches $50 I will finish the mod, but will then shift gears on designing a portable L1/L2. I'm pretty sure I could get the electronics/components required, minus the J1772 connector, for about $50.

chris1howell wrote:
I believe the ground check uses the 3rd smaller relay (3rd picture far right) to pass a small current from L1 Hot to ground through a 5k or so resistor. If you listen carefully you can hear a clicking right after connection before the main relays engage. I do not think it is checking to see if Neutral and Ground are tied together only that there is a path to ground. Routing current through the 5k resistor would be enough to cause a ground fault so.... my theory is the EVSE powers the 3rd relay causing current to flow to Ground causing the GFCI to read a fault if so, ground and the GFCI CT coil is good. If ground is bad no current flows and the GFCI does not read a fault and the EVSE goes into Fault state.

tps wrote:
I thought about that, but wouldn't that cause an upstream GFCI outlet which was supplying the EVSE to trip? My calcuation is that 5K would cause 24 mA to flow, which would definitely cause the GFCI to trip out. I'm guessing the current which is injected comes from the secondary of the 20 volt transformer. With this isolated source, they could put current into the ground, splitting the return current equally on the line and neutral, thus allowing a flow of current to test ground continuity while maintaining current balance on the L and N to prevent tripping upstream GFCI.

Edit: Now I think of it, that wouldn't work. Since return current it the same phase in L and N, it would add rather than cancel to zero in the CT core, so it would cause the external GFCI to trip...

Maybe it is a resistance of 100k from hot to ground. That would not trip GFCI but the voltage generated by the CT could be measured to verify that current is passing to ground... Without taking it apart it’s just a guess....

[DaveEV]

I was thinking of something similar to that as well. I would really dig into the standard and probe and make guesses, but I'm at the point where I might as well depot so I can give the folks, who would like to know if this is possible, easy, and inexpensive, some results. If I find that the mod would approach being too difficult for the average person or the cost approaches $50 I will finish the mod, but will then shift gears on designing a portable L1/L2. I'm pretty sure I could get the electronics/components required, minus the J1772 connector, for about $50.

You've seen chris1howell's DIY EVSE project thread, right?

[jwallace3]

I have read some of it and remember seeing it a while back. It's all just for the heck of it. I know this thread got out of control quick, but I did mention early on I'm mentoring 2 EE seniors in their senior project, which is to design a portable L1/L2 ESVE so I will probably do it even if I get the mod to work, just so I am forced to go over the standard with a fine tooth comb and be able to help the students define their specifications and solidify their design.

I was thinking of something similar to that as well. I would really dig into the standard and probe and make guesses, but I'm at the point where I might as well depot so I can give the folks, who would like to know if this is possible, easy, and inexpensive, some results. If I find that the mod would approach being too difficult for the average person or the cost approaches $50 I will finish the mod, but will then shift gears on designing a portable L1/L2. I'm pretty sure I could get the electronics/components required, minus the J1772 connector, for about $50.

You've seen chris1howell's DIY EVSE project thread, right?

[DaveEV]

I have read some of it and remember seeing it a while back. It's all just for the heck of it. I know this thread got out of control quick, but I did mention early on I'm mentoring 2 EE seniors in their senior project, which is to design a portable L1/L2 ESVE so I will probably do it even if I get the mod to work, just so I am forced to go over the standard with a fine tooth comb and be able to help the students define their specifications and solidify their design.

Right - I was only pointing it out in case you hadn't seen it and knew that he already has a DIY EVSE control module in case you wanted to use that as a starting point for your own design or depending on the level of the students give them something to build from if going down the build from scratch route.