Weekend(s) Project - DIY EVSE

[charlestonleafer]

Are you ready to sell them yet???

 

[chris1howell]

Bluetooth SPP working to read and update settings (including charge current). Now I just need to figure out how to write and app for Android...


2011-10-03_21-12-16_815 by chris1howell, on Flickr


2011-10-03_21-12-46_128 by chris1howell, on Flickr

 

[fooljoe]

Looks awesome Chris! So with the Bluetooth chip you'd be able to switch the current from, for example, 16a to 12a wirelessly? That'd be amazing. By the way, do you have any idea if it's possible with the Leaf to draw 16a on a 120v circuit? I know a lot of people have L1 charging available at work, and commercial circuits are often 20a, so being able to get that extra 33% would be nice if the Leaf allows it.

 

[chris1howell]

12a is the max at 120v under normal circumstances. Lincomatic discovered an issue where the car would draw higher under low voltage brownout conditions. I believe he saw 17A during a brownout on a very hot day.

 

[johnr]

This is the most awesomest EVSE and it just keep getting better! Amazing how tiny it can be and even have bonus features like bluetooth.

Keep up the good work!

 

[fooljoe]

12a is the max at 120v under normal circumstances. Lincomatic discovered an issue where the car would draw higher under low voltage brownout conditions. I believe he saw 17A during a brownout on a very hot day.

But is that max dictated by the Leaf or by the EVSE?

 

[chris1howell]

The EVSE sets the max current.

 

[Volusiano]

The EVSE sets the max current.

We know that the EVSE sets the max, but it doesn't necessarily mean that the 3.3KW charger will draw whatever the max current set by the EVSE, right? The question is whether the 3.3KW charger maxes out at 12A on 120V or can it go to 16A at 120V if the EVSE says it has 16A to give.

I think if the 3.3kw charger could draw 16A at 120V, Phil would have mentioned it as an option when he was polling people for inputs on whether he should make the Rev2 upgrade programmable between 12A or 16A. He only gave the 12A vs 16A option for the 240V operation. He never mention a 16A option for the 120V operation at all, ever. So I'm inclined to think that the 12A at 120V is a limitation from the 3.3kw charger.

 

[chris1howell]

That is correct the LEAF on board charger will only draw 12A max at 120V even if the EVSE offers more...

 

[Deleted member 3147]

My understanding is that there is a "bug" in the on-board charger. It will draw only 12A at 120V (1.44KW), even if offered more by the EVSE. However, if presented with LESS THAN 120V (but above some minimum), it will draw more than 12A (in trying to do 1.44KW), even going past the offer from the EVSE!

Remember, watts = volts x amps. If the voltage drops, you need to pull more amperage to make the same wattage. The on-board charger is trying to do 1.44KW, even at lower voltages.

 

[fotajoye]

Not sure it's a bug; I would bet it's designed to work on a common house whole 120vac circuit which is 15amps, not a dedicated commercial circuit that is normally 120VAC 20amps. BTW, you can go to evseupgrade.com and have your Nissan Brick modded to work on 240vac at full charge current and also maintain the current 120vac 12amp input.

 

[jwallace3]

Very nice Chris! I may hit you up with some ideas when my students start there design.

 

[Deleted member 3147]

Not sure it's a bug; I would bet it's designed to work on a common house whole 120vac circuit which is 15amps, not a dedicated commercial circuit that is normally 120VAC 20amps. BTW, you can go to evseupgrade.com and have your Nissan Brick modded to work on 240vac at full charge current and also maintain the current 120vac 12amp input.

The bottom line is that the car will, under low-voltage conditions, draw more current than the EVSE is saying that it can. That seems like a clear violation of the J1772 standard.

Also, if it is designed for a residential 15A circuit, drawing more than 12 amps is probably a bad idea. It really should never draw more than 12 amps.

Unfortunately, if your house has 110 volts instead of 120, due maybe to being out in the boonies so you have a log utility run, this would mean that you would charge ~10% slower. Maybe this is why Nissan decided to try to pull the full wattage.

 

[garygid]

The EVSE tells the car a Max-Current value that the car SHOULD obey.
Under certain conditions, it has been reported that the LEAF's charger violates this Max (maybe not significantly). Probably, better measurements are needed to verify.

The EVSE could enforce the Max-Current, but it is not required to do so.
It is not recommended or even suggested that the EVSE limit the current.
The assumption appears to be that the supply circuit's breaker will take care of that.

If the EVSE changes the Max-Current during charging,
the EV SHOULD respect the new limit.
However, I think it was reported that the LEAF does not.

It now is reporter that the LEAF DOES shut off charging (stops drawing
current, if any) when the nozzle's Release Button is pressed.

Finally, the Nissan EVSE does not do the J1772-required "diode" safety check.

A few possible violations.

Edit: changed to match recent tests by others.
I should have done my own verification testing before posting.

 

[QueenBee]

It also appears that the LEAF does not shut off charging
(if any) when the nozzle's Release Button is pressed.

I was surprised by this as well. Is that a needed feature though considering that when you pull it out the communication pin breaks first shutting things down?

 

[garygid]

If the AC-Power contacts in the J1772 Plug and LEAF's Socket are conducting current, it is BEST to NOT slide them or they might "pit" (develop hot spots).

It is much BETTER for the EV to sense the button-press and quickly turn OFF the current BEFORE the connectors start to slide. That was one of the three main features of the Proximity Circuit:

1. Sense the Plug and not allow the car to move (but some EVs MIGHT be able to roll away, unpowered).

2. Not allow charging to begin if the Plug is not inserted far enough to allow the Plug to "latch" to the Socket.

3. Sense impending Removal of the Plug, and quickly shut down charging, if any. Stopping the current flow before the contacts slide tends to reduce "pitting" of the "high current" contacts. (This feature is VERY important on the QC Port.)

 

[DarkStar]

Not sure it's a bug; I would bet it's designed to work on a common house whole 120vac circuit which is 15amps, not a dedicated commercial circuit that is normally 120VAC 20amps. BTW, you can go to evseupgrade.com and have your Nissan Brick modded to work on 240vac at full charge current and also maintain the current 120vac 12amp input.

The bottom line is that the car will, under low-voltage conditions, draw more current than the EVSE is saying that it can. That seems like a clear violation of the J1772 standard.

Also, if it is designed for a residential 15A circuit, drawing more than 12 amps is probably a bad idea. It really should never draw more than 12 amps.

Unfortunately, if your house has 110 volts instead of 120, due maybe to being out in the boonies so you have a log utility run, this would mean that you would charge ~10% slower. Maybe this is why Nissan decided to try to pull the full wattage.

Not as far as the evidence has suggested. If the EVSE is sending a 16 amp signal but 120 volts, the Leaf will draw a maximum of 12 amps. If the voltage drops, why wouldn't it be acceptable to go up-to 16 amps per the EVSE? Now, if you tell it 12 amps, it should stay at 12 amps regardless of the voltage.

 

[tps]

My understanding is that there is a "bug" in the on-board charger. It will draw only 12A at 120V (1.44KW), even if offered more by the EVSE. However, if presented with LESS THAN 120V (but above some minimum), it will draw more than 12A (in trying to do 1.44KW), even going past the offer from the EVSE!

Remember, watts = volts x amps. If the voltage drops, you need to pull more amperage to make the same wattage. The on-board charger is trying to do 1.44KW, even at lower voltages.

Sounds as if, for whatever reason, the charger uses an algorithm which pays attention to the pilot current offering when in L2 mode, but it ignores the pilot and charges at fixed power in L1 mode. While this is not strictly kosher, it probably won't result in significant deviation or blown breakers if the voltage stays within the tolerance the utility is supposed to provide (in the US, within 5% of nominal 120V, or 114 to 126V).

A potential real problem: My guess is that the 7A setting on the Leviton L1 doesn't reduce the LEAF's charging current...

 

[chris1howell]

What the car should do and what it actually does are different. Having complete access to the EVSE hardware and software has uncovered several bugs in the LEAF charging...

- The car bounces the relays several times on connect (sounds like a machine gun for about a second) if the EVSE is fast, a delay was added between the ready and charge state to the DIY EVSE to stop it...
- Diode check not included in the NISSAN EVSE
- Ventilation required state not included in the NISSAN EVSE

 

[davewill]

...
- As gary mentioned the button on the handle does not stop charging as recommended by J1772.
...

Are you sure about that? We discussed this, but there was no conclusion. You WOULD be in a good position to have verified this. It would be a significant fault on the car's part if it didn't.