False Positives With FW 5.8.4 Charge Current Reduction?

[tezco]

I had 198V at a chargepoint station yesterday with full amperage (30A) yesterday with no degradation from the car for three hours in San Jose.

I think they've got algorithm problems. My suspicion is it has fixed voltage limits. There's clear evidence for this. There's a particular CS-90 that I've used, which has the lowest voltage I've seen anywhere. I've been told it now always backs down 25%. My office charger is on the low side, and it backs down about half the time. I recently charged on a CS-90 with > 235 volts and it never backed down at all. So it's pretty clear there's a fixed threshold. Bad design.

The grid is a major source of voltage variation. I've observed it at my office - significant changes in a fairly short time period, and it clearly wasn't our load doing it because we don't really have that much (no car plugged in at the time).

You can't have fixed limits - it simply doesn't work. What it SHOULD do is record the nominal open circuit voltage when it first connects. It should then calculate a threshold voltage below that point, where it is going to trigger the back-off.

The software may actually be doing that. I suspect the software probably does detect the nominal voltage when you first plug in; otherwise, how could you charge at 208 or 120 V? If the voltage is low, such at Lloyd's location, but doesn't fluctuate, then you're good to go.

If your voltage drop is high when the car first starts to charge due to high resistance in the circuit, you're already closer to the lower limit that the software calculates. ie, if the cut-off voltage drop is 5% of initial voltage and you start with a 4% drop, it might only take a small sag on the grid to drop you under the limit. Out here on the farm I typically see a 3% swing in voltage every day at a sampling rate of once a minute. I suspect the swing would be even higher with a very rapid sampling rate that would catch the sags and spikes of large inductive loads switching on and off. Just wait until airconditioning season sets in!

 

[cinergi]

I don't think it's fixed because I successfully charge from a 240v supply at 218 (starting from low 230's).

 

[Doug_G]

The software may actually be doing that. I suspect the software probably does detect the nominal voltage when you first plug in; otherwise, how could you charge at 208 or 120 V? If the voltage is low, such at Lloyd's location, but doesn't fluctuate, then you're good to go.

Well, for a starters it probably looks and decides if the voltage is in the 120V range or 220V range. From that point on we don't know what algorithm they use. They could be looking at rapid spikes up and down, but they could also have absolute thresholds. We simply don't know. That said, I've seen it cut back 70A chargers with low voltage < 200V, but not 70A chargers with higher voltage.

IMHO if it sees > 40A permitted by the EVSE it shouldn't use the new more sensitive algorithm, because it's guaranteed that you're not using a UMC.

 

[Cottonwood]

Here is a good sample of a HPWC charge at 80 Amps. The Tesla load caused 2 Volt deltas at the panel and 4 Volt deltas at the car; half of the resistance was upstream of the panel! Also note the bigger swings from what the grid was doing on its own...

This is a charge from 80% to 90%, getting ready for a range charge in the morning...

 

[J in MN]

Here's another data point towards understanding the algorithm: My third event, setup to demonstrate charge current increase, in this case from 28 A to 30 A.

As before, it trips on a voltage rise, but notice how it survives the first rise of 3.2 V @ 06:10:21, but then trips on the ~1.4 V rise at 06:12:21. My sampling rate is not good enough to define this better.

So, it seems pretty clear that it is looking for dV/dt > x, where x > 3 V/s. Maybe x > 60 V/s (i.e. 1 V per cycle).

 

[Cosmacelf]

Have you sent these graphs to tesla? Having the algorithm increase amperage draw is nuts. I sent them an email about it mid week but haven't heard back other than we are looking into it.

 

[neroden]

Tesla's engineering department is a black hole. Information goes in, but information doesn't come out.

This is not good. There are things Tesla has engineered well, and things where they've been awful. A little two-way back-and-forth feedback could probably have prevented most of the latter group of problems, but even the service departments seem to feel like the engineering department is an informational black hole.

 

[Vger]

FWIW, I talked with my service advisor about this whole issue last week. He said they are getting BURIED in phone calls and emails about it. It is the #1 complaint from customers in their shop. They are equally frustrated that HQ is not telling THEM what the algorithm is, so they are just as much in the dark as us!!

 

[cinergi]

Still send it to ownership if you haven't. I've also sent my contacts information about the current-increase bug since it's pretty serious.

 

[J in MN]

First one went to ownership and service center, the rest only to the service center.

My service center seems to have good lines of communication with engineering / "inside service", and they provide me with updates when they get them, something ownership does not always do.

 

[InTheShadows]

First one went to ownership and service center, the rest only to the service center.

My service center seems to have good lines of communication with engineering / "inside service", and they provide me with updates when they get them, something ownership does not always do.

I'm surprised to hear that. I have sent three emails to ownership and have heard back every time.

 

[Cosmacelf]

FWIW, I talked with my service advisor about this whole issue last week. He said they are getting BURIED in phone calls and emails about it. It is the #1 complaint from customers in their shop. They are equally frustrated that HQ is not telling THEM what the algorithm is, so they are just as much in the dark as us!!

Good to hear. Since this is actually a dangerous bug (and, frankly, prevents people from using home made adapters to access less than 40A electrical sources), it ought to be corrected ASAP.

 

[Cottonwood]

Good to hear. Since this is actually a dangerous bug (and, frankly, prevents people from using home made adapters to access less than 40A electrical sources), it ought to be corrected ASAP.

In addition, I have encountered 14-50's with 40A breakers and 14-30's with 20A breakers, had the breaker pop, found the breaker, reset it, set the current limit to 80% of the breaker, and then all worked fine. You need to be able to dial back the current and not have the car raise it above what the circuit will allow.

 

[mknox]

In addition, I have encountered 14-50's with 40A breakers...

Very common here in Ontario, Canada. Most homes are wired with NEMA 14-50's in the kitchen for stoves, but are wired and fused at 40 amps.

 

[neroden]

First one went to ownership and service center, the rest only to the service center.

My service center seems to have good lines of communication with engineering / "inside service", and they provide me with updates when they get them, something ownership does not always do.

I guess we need to contact *your* service center! :tongue:

 

[AndreyATC]

I was ok yesterday, charging at 40A with NEMA 15-40, but today, just after 2 hours of charging it went down to 30A with "check charging cable" message

 

[Rainbow]

I have had a couple times since the update where my charging has reacted to apparent fluctuations in power, but here is where it is a problem. It does not reduce the draw 25% from where you set it, but 25% from maximum. My garage is only equipped with a 30A breaker. The total to the garage is only 50A and I don't want to have to pull a new feed from the house. Because of this at home I charge at 24A. When the car detects whatever it is looking for it actually ups my setting to 30A instead of reducing it 25% to 18A.

I sure hope Tesla does something about this last revision soon, especially if the fault can cause an increase to 30A - unacceptable.

Wow, that is definitely not right! Just the kind of bug that gets introduced by a rushed f/w update!

He tells that car to draw only 24A. But the software sees a voltage fluctuation and then INCREASES the charging current to 30A. If the software though it was bad at 24A, increasing it to 30A is really stupid and, in this case, a violation of a safety code (80% of breaker rating). No breaker will trip, but wires and breakers will get hotter than they should.

This BUG (there is no other way to describe it) also makes ALL HOMEMADE ADAPTERS DANGEROUS TO USE. Homemade adapters for a variety of plugs that Tesla does not support usually use Tesla's NEMA 14-50 adapter. So you might plug into a 20A 240V receptacle, dial it down to 16A and then the software bumps it up to 30A. Seriously bad behavior.

This is dangerous situation. Does anyone have an idea when this problem will be resolved?

 

[Cosmacelf]

I just noticed that my daily midnight charging at my house has been reset by the car from 40A to 30A. I don't remember getting a message, but doesn't mean I can't have missed it when hurrying somewhere. Annoying. There is nothing wrong with my garage setup.

 

[Forty Creek]

I just noticed that my daily midnight charging at my house has been reset by the car from 40A to 30A. I don't remember getting a message, but doesn't mean I can't have missed it when hurrying somewhere. Annoying. There is nothing wrong with my garage setup.

Similar issue with me using my HPWC. Properly wired and inspected. Hoping for a less sensitive software 'fix' to come along soon.

 

[mnx]

I'm not sure if this is 5.8.4 related or not but I was at Great Wolf Lodge in Niagara Falls this past weekend, and the charge kept aborting on the 120V outlet they had there (which has been used by others to charge EV's in the past)

I tried backing down the amperage. I had to dial it down all the way to 6A for my car to not abort charging.

I suspect it was 5.8.4 related since the display said charging rate reduced before it would abort (even though it didn't actually reduce anything).