Car won’t charge on HPWCs/UMC, says check Wall Power

[boaterva]

Posting this here, even though Model X, since it’s really not X specific.

Just after loading 2018.20, this happened. Tesla doesn’t think it’s firmware dependent.....

I have two load-sharing HPWCs, and a backup 14-50 (see sig). On the first drive after loading 2018.20, and returning home, plugged in the car and got: Ready to Charge, check Wall Power. Never had a problem, of course, in eight months. Car usually and has always charged to 80, down to 77, and up. (Once in a while, a rogue firmware had it go higher, but a reboot fixed that, or it happened when the servers were down; none of that is relevant here.)

So, I tried the other HPWC, and the UMC using the 14-50. Reset the HPWCs with the button. Cycled the 100 Amp breaker is the subpanel and the main panel, nothing. Rebooted the MCU, of course (almost didn’t add that, it was so obvious that we do that for everything!)

Called Tesla, and got a remarkably speedy answer for a holiday weekend when California is awake. Wonder if they are beefing up support? After she checked the logs, and some discussion, all she could say is that I should try charging somewhere else to show the house power is fine.

Nearest Supercharger isn’t that close (guess I could go to the new Fredericksburg one!) or even see if the ev-go Chademo in Reston is open to test.

But my real question here is, wouldn’t the HPWCs show ‘bad power’? If it was so bad that the car couldn’t accept it?

And/or the UMC?

She is putting in a mobile ticket to have a Ranger stop by to check the charge port. So, woohoo, looks like NoVa is in Ranger territory for some things (or they will call me, say, so sorry, come to Tysons! )

And actually she just called back to check that the UMC didn’t do any special light blinking when it was plugged in, and we did it again, and it’s one steady green light, as it should be (which makes me think the power is indeed fine).

Where is the ‘power checking hardware’? With the chargeport, so that indeed could be the issue?

 

[omarsultan]

I just had a similar issue with a 5-year old gen 1 HPWC. Line supply was fine, lights were green, but car complained about checking wall power. Ranger came by and it ended up being the board in HPWC.

 

[boaterva]

I just had a similar issue with a 5-year old gen 1 HPWC. Line supply was fine, lights were green, but car complained about checking wall power. Ranger came by and it ended up being the board in HPWC.

Interesting. But here all three don’t work. And the UMC has never been plugged in except when I tested it when I got the car and now. So, hard for that one to have been blown up.

 

[hacer]

...
Called Tesla, and got a remarkably speedy answer for a holiday weekend when California is awake. Wonder if they are beefing up support? After she checked the logs, and some discussion, all she could say is that I should try charging somewhere else to show the house power is fine.

Nearest Supercharger isn’t that close (guess I could go to the new Fredericksburg one!) or even see if the ev-go Chademo in Reston is open to test.
...

It is likely that (1) your car has a problem and (2) it might only affect AC charging so it is likely that Supercharging and Chademo will work fine because they are DC charging.

 

[boaterva]

It is likely that (1) your car has a problem and (2) it might only affect AC charging so it is likely that Supercharging and Chademo will work fine because they are DC charging.

That was their thought. Going to see if we can try one of those tomorrow, probably chademo now. Strange how it breaks after a ten minute trip to the store. I asked the chademo question of Tesla since I didn’t recall if it was AC or DC at the moment but the support people didn’t know offhand. Thanks for the confirmation!

 

[mrElbe]

If Supercharging and ChaDeMo works and all other AC device inputs get the complaint, then that points to a problem with the on board charger.

 

[boaterva]

If Supercharging and ChaDeMo works and all other AC device inputs get the complaint, then that points to a problem with the on board charger.

That’s what I thought in the first place but logs/whatever diags showed nothing. And the car points to bad inlet power, which isn’t likely based on the testing I’ve done. You’d think the car’s/charger’s diags could self-test more than that but it can always fail, of course.

I’m betting the chademo test will work tomorrow (if I can get to do that at EV-go) and then I’ll pass that on to them when I get contacted for the Ranger apppintment.

 

[voidptr]

I had something similar happen about a year ago to mine. It would charge fine at the service center and public chargers (208v) but fail on my UMC (at 240v). Initially they said it was over voltage at my house according to the car logs, and it took about a week of diagnostics, swapping UMCs, and randomly sampling the voltage where my UMC plugs in before they took the car in and tested it overnight at the service manager's house, and then a few days back and forth with the engineering team in Fremont before they replaced the onboard charger.

They never told me the exact cause, but it sounded like a voltage sensor inside the charger failed in rather weird way they had never seen before where it read 10 - 15v higher than it should have, which caused the charger to go into over voltage protection mode. Yours sounds like it might be the same problem.

 

[boaterva]

Interesting.... I doubt the whole neighborhood is overvoltage. And it happened in the space of 20 minutes and hasn’t gone down. Let me get my volt meter out... I thought the devices measured that, too.

 

[boaterva]

So, this is interesting. First, does anyone know what the data field is in TeslaFi that shows input/inlet voltage? Is there one? Charger voltage doesn't seem to work. Would like to track what voltage the car is getting.

I measured the current voltage (lol) right now (0630 EDT) and it was 245 VAC. Hm, I sez, that shouldn't be bad, only about 2% overvoltage, should work fine. So, I plugged her in, heard the HPWC contactor clunk, and it all works. Arg...... Good thing I checked it again.

Wonder what the voltage was yesterday?!

Thanks much for the tip, and now I know to actually check the source levels with a voltmeter if the car complains.

I'm not even going to start on 'Tesla should have said something', there's no way they are going to even suggest people should measure 240 VAC themselves. I'm an EE and have been doing this stuff for decades. Other people could find themselves blown across the room if they do it wrong. But it would be nice if the HPWC has over/undervoltage alarms. I wonder if yesterday it was actually low, being a 90 degree day here?

Now to see how it does this afternoon and to discuss when I get a Ranger call. Glad the car is actually fine.

(I wonder if this would be in better shape when I get my solar and Powerwalls? Will it balance over/undervoltage?)

 

[boaterva]

Car charged right up and seems fine. So, if it tells you the voltage while it’s charging, WTF can’t it report the voltage on an under or overvoltage alarm? So you know it’s not the car’s fault and can do something about the house’s power or at least charge elsewhere.

Even TeslaFi only seems to be able to get the voltage while charging ‘properly’.

 

[animorph]

I've seen my car charge at 250V or 251V, though with reduced current to stay within a power limit. I think the voltage limit is around 255V, mentioned in a thread maybe a month or two ago.

 

[boaterva]

Good to know, since got home just now, and got the same issue. Voltage is around 248. And if I plug in the HPWC and wait a bit, besides the ‘charger’ error, I get a flag in the IC saying ‘no wall power’ which makes you think it’s the HPWC. But not all three. Arg.... perhaps it is the false high measurement as mentioned above.

I’m leaving the ticket in (haven’t called them back or heard from them yet) and will see what they think the cause is.

 

[boaterva]

Update: the HPWC manual seems to say the max is 250VAC but I’m leaning toward the sensor board is bad in the car since none of the three sources work at the current ‘acceptable’ voltage.

We may still try a DC source tomorrow to just verify/cross that off. Of course, based on today’s experience, AC will work in the AM.

 

[boaterva]

Ranger is looking at this remotely, first response from his team was a question about the load balanced HPWCs, which can’t have any relevance, as they have been there since day 1. The ONLY change in the car that lines up is that 2018.20 was loaded, and I mentioned that. Didn’t try a DC source as so far AC has been reproducible.

Car charged to 80 this AM, After noon, for Ranger’s benefit, when it got there, I pulled the wand out, upped limit to 85, put it back, and it failed as it has the other days. Remote S reports 242 VAC, which I can’t tell is the last reading or the current voltage. When it is charging it does report a good 242/241 which may just be the higher voltage I reported drawn down by the 40Amp load (what I have the car limited to).

 

[boaterva]

And, sigh, now she fixed herself.

I don’t know if they did something or what, but at 2:24 PM charging started. Started, ran up to the new limit of 85 and stopped normally at 3:00 PM. Waiting for a call to see if they actually did anything or they don’t have a clue. Remote S and TeslaFi show a normal voltage of 242. About what it’s been about that all day, no real increased voltage to indicate ‘I don’t like the power because’.

I’m now wondering if this is really something with 2018.20 since it’s awful coincidental (but that’s could be post hoc, as usual) and it’s not in very wide distribution. With a testing universe of one, hard to say!

More news as I get it...

 

[boaterva]

Now the Diagnostician Tech thinks it's the HPWC power connections, even though I said again I tested it with the two load balanced WC's AND the UMC in the first place. Sometimes I wonder how much troubleshooting skills these people have... They wanted me to have the electrician come out and 'check the connections'. How would that do anything when (1) it's worked for eight months, and (2) it works at 40 Amp when it does works. If the connection was getting bad, I doubt it would hold 40 Amp current at all (that's the max I have it at, it's a 100 Amp circuit, set to 80 on the HPWCs), especially at 240 VAC.

So, the next step will be to test charging once in a while and call them back and have the mobile car (a Model S they use for this that has a powered UMC in it!) to test it with power separate from my house.

They have not seen this before, recently (as in, with 2018.20). But the population for that is so low, who knows.

 

[ccdisce]

Now the Diagnostician Tech thinks it's the HPWC power connections, even though I said again I tested it with the two load balanced WC's AND the UMC in the first place. Sometimes I wonder how much troubleshooting skills these people have... They wanted me to have the electrician come out and 'check the connections'. How would that do anything when (1) it's worked for eight months, and (2) it works at 40 Amp when it does works. If the connection was getting bad, I doubt it would hold 40 Amp current at all (that's the max I have it at, it's a 100 Amp circuit, set to 80 on the HPWCs), especially at 240 VAC.

So, the next step will be to test charging once in a while and call them back and have the mobile car (a Model S they use for this that has a powered UMC in it!) to test it with power separate from my house.

They have not seen this before, recently (as in, with 2018.20). But the population for that is so low, who knows.

I have to agree with the Tesla Tech that you should have an electrician check out the wiring, breaker, wiring panel so you can flame away in my direction if you wish.

A motto I have developed over the years is:
'when you have gone down all the rabbit holes with no fix in hand, you have to take the shotgun approach and go down all the rat holes rapidly blasting away at anything that moves, later working out the reason why the last shot killed the bugger, and then build a test case to verify that your fix will fix the perceived problem.

Maybe you have multiple problems running serialy or in parallel.

I have been a trouble shooting engineer since 1968 messing around in electrical/ electronic systems.
Recently I had to fix a charging issue when I tried to charge my 100d in the garage using the umc and 120V.
I had to rewire a plug inside the house and I built a Supply monitoring extension cable using a PZEM 004 as 'suggested' by the HPWC thread here on TMC.
I later updated it to work on the nema14-50 that I installed to handle the HPWC.

 

[boaterva]

Well, I agree if it works on their power, I'll call Havepower in, but the variability of this makes no sense where it goes from dead to charging at 40 Amps for hours with no issue and back and forth. (And no source of heat, either, to indicate a loose connection.)

The Tesla tech saying its the current HPWC when I tried all three units also makes no sense. What I can try is running the UMC on a separate 15 or 20 Amp circuit, which is completely separate from the 100 Amp the WCs and the UMC (off the 14-50) were on.

I know what you're saying (EE here, myself!), but they just don't seem to be doing this 100% logically. They also don't seem to have enough data from the car, unfortunately. All it tells them is 'bad external source' and stops, no details on what is bad about it (high? low? frequency?).

Thanks for the insight, and I also wish I had some sort of power monitor on the line. When I get solar panels and Powerwalls in, that may give me more insight. Or more things for people to point at.

 

[MP3Mike]

@boaterva I haven't been able to find it but there was a thread on here a while back about someone having similar problems of not being able to AC charge at home most of the time, but it worked elsewhere all the time. Tesla was eventually able to combine data from other cars in the fleet to determine that it was a neighborhood wide problem and get the power company to find and fix the issue. (Since the power company wouldn't listen to the individual owners complaining that something was broken since everything else was working just fine.)

If you don't have a lot of Teslas in your neighborhood it could take longer to identify.