Tesla Universal EVSE Connected, but 12V battery drained

So I've found at least one other vehicle-side charge controller, the New Eagle VPIM v7, which has a separate "control pilot feedback" line, and its design might offer an explanation. Their feedback line is pulled high to force the EVSE off until the vehicle is ready to charge -- which is odd, but not disallowed by the standard. The problem is that while they also have a low-power mode, the unit wakes up based on the proximity pilot line -- so while it's "forcing the charger off" it could waste power if the EVSE will actively sink current on that line instead of just not sourcing it.

I think this issue might be mitigated by adding an extra Schottky diode between the CP line in the power input port and the charge controller, which might be included in e.g. the Lectron adapter. I bet the voltage tolerances are wide enough it would handle that no problem, and it would keep current from leaking out to the EVSE.

 

Looks like Teslas use a proprietary extension to the J1772 signaling standard; they can switch to a mode where the Control Pilot line is used as a single-wire CAN bus at 0V and 5V, rather than J1772's 12V PWM signal. This mode requires that their EVSE actively pull down that line to ground, thereby wasting whatever 12V power the Clarity is willing to put on the line and draining the battery. I'd bet the adapters that solve the problem include an extra diode on CP line to force any connected Tesla EVSE to operate in plain old J1772 mode, no matter what crazy thing the car circuitry might be doing which might accidentally trick the charger into the proprietary CAN mode... and that just happens to solve the Clarity's issue too.

Upshot here is that Honda's the one not really following the standard properly. (There's supposed to be an input diode *before* the 12V pullup thing the charge control unit is doing, not *after*.)

 

I followed up with Tesla support, and they had no updates from engineering other than to use the mitigation of a third-party J1772 adapter in place of theirs.

 

I admire your persistence !

I'm pretty sure this would be an 'easy' problem for them to fix, but they apparently have 'better' things to do.
If a Tesla EVSE designer had access to a Clarity, they could probably solve this in 5 minutes.

It does make me wonder if there is just something a little different with a Clarity. If not, then you would think Tesla would be inundated with complaints from other users of their 'Universal' EVSE's.

 

Perhaps Tesla could just take a look at one of the adapters that works with a Clarity...

 

My wife does not ;-p

Agreed!

Also agreed!

 

Do you happen to have a multimeter with a diode test mode? If so, could you try testing the control pilot line on the third-party adapter to see if there's a diode in there?

 

I have my multimeter in another location. I can get access to it the weekend of the 10th. Then I'll share a picture and maybe you can help me see if it has the feature you're asking about!?

 

The pins I've circled in red here are the "Control Pilot" pins, and my working theory is that they're how the 12V battery charge is leaking out of the Clarity. (Note that the picture has a Tesla plug, not a Tesla receptacle; on your adapter, you'll have a Tesla receptacle so the pin will be on the other side. For further reference, both pins should be on the left side of their respective sockets when the adapter is plugged in.)

I think the Tesla adapter probably just connects those pins together with a wire, but the third-party adapter has a diode which only allows current to flow into the car and not back out again. If so, you could use the continuity mode on a multimeter (the one where it beeps when you touch the probes) on those pins in Tesla adapter, and it would beep, while on the third-party one it wouldn't. Furthermore, in diode test mode, the multimeter should show a voltage (probably 0.6-0.7 volts or so, possibly less) when you have the negative lead connected to the J1722 CP pin and the positive lead connected to the Tesla CP pin, whereas the meter should not show anything if the leads are the other way around.

If that's the case, I'll probably try to DIY a fix for my Clarity; no promises I'll actually succeed, but I'll be happy to share progess one way or the other.

 

Cool! I'll try it out in just over a week. Thanks for the tag team effort!

 

@MeNerdHair,
This will be an interesting experiment... It would be quite surprising if there could be adequate leakage through a signaling pin to cause this kind of drain on the Clarity's battery however. It seems more likely (if the diode theory pans out) that this difference is affecting the signaling protocol and the Clarity is remaining in an 'on' state when it should be shutting down. The observation is that the failure mode causes a periodic 'cycling' behavior, and it is this incorrect continuous cycling that may be causing the vehicle to drain itself rather than a parasitic sneak path.

 

That makes a lot of sense. I too was wondering how you could pull enough current over that pin to drain the battery but not break anything! I'd just assumed my still-factory 12V must suck really really badly.

 

My understanding is that the 12V Clarity battery drain is happening due to improper pilot signaling between Honda and the TUWC. This causes the vehicle to never enter its sleep state. The TUWC starts clicking the contactor with approximately 1 minute frequency after the Clarity battery is fully charged. The TUWC misinterprets Clarity's signal as requesting a charge and then Clarity disables charging because the battery is already full, and the cycle repeats.

 

What does the adapter known to solve the problem do to alter the improper pilot signaling? I don't believe the adapter is "smart."

 

If you're asking about a Chinese Tesla to J1772 Adapter, I think there might be a resistor on the pilot signal that tricks it into properly recognizing a request from a Clarity. As for why Tesla support isn't solving this issue, it's likely because they're a large corporation, and support staff may not prioritize it until it becomes a bigger problem for Tesla.
I've spent probably close to 40 hours troubleshooting with Tesla tech support, and to be polite, they all seem clueless.
Elon likely has no idea what's happening with TUWC. He's busy solving big problems and generating big ideas.

 

I'm tracking the TUWC Pilot signal. Here's the screenshot while using a Chinese adapter. I can remove the adapter tonight and record what happens with the original Tesla adapter

Screenshot 2024-04-25 1257041 by alexsahka posted Apr 25, 2024 at 1:18 PM

 

Here's a screenshot with the TUWC Pilot signal using the original Tesla adapter. There are a lot of spikes causing Clarity to not sleep. See the blue lines.

Screenshot 2024-04-25 210005 by alexsahka posted Apr 25, 2024 at 11:05 PM

 

Thank you!
@MeNerdHair Do you still need me to test the Lectron adapter with my multimeter?

 

If someone needs the schematics for the adapter I'm using.

 

@alexsahka,

You have done an excellent job of examining what is reality happening with the 'good' and 'bad' setup. The problem is clearly evident and it is totally ridiculous that this information is not adequate for Tesla to get to the bottom of this issue. Their support team is obviously inept and they are not passing the relevant information to their engineering team like they should.

Your Chinese adapter circuit is consistent with the J1772 standard. (See post #65 earlier in this thread...)

I am confused about a couple things.
First, the Pilot signal is bipolar (should range from +12 to -12V).
Your signal plots (both) are 0-12V. Are you only able to measure positive voltages with your setup?

What is interesting (and probably salient) is the fact that with the working Chinese adapter you are seeing 3 states (9V, 8V, and 6V). With the Tesla adapter, it is just two states (9V and 6V). Also, when unplugged, I would expect the pilot to be swinging the full +/- 12V as it does with the Chinese adapter.

What seems to be lacking here is the equivalent schematic of the TUWC's J1772 adapter. I don't have one, but from pictures I have seen online, it looks like there is a latch button on the Tesla adapter, in addition to the normal one on the Tesla NACS handle.

If I had one of these, I would ohm the Tesla J1772 adapter to see if it has the resistors / switch in the adapter (or does it just conect all pins through to the Tesla NACS)? Then depending on the outcome of this test, I would disconnect the TUWC from the 240V line and measure the CP pin on the NACS to see if the resistor and switch are there.