-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
With all due respect, I didn't ask for more features. I just want to keep the ones I paid for. I had 40A charging for over a year, and then Tesla took it away. Yet, Tesla still advertises you can charge at 40A with the UMC on a 50A circuit.
- - - Updated - - -
@FlasherZ
When you purchase a product, it should work as advertised. I have already established that "I don't have a problem with my electrical connections". This is backed up by the experts at the power company. They publish what their standards are (240V, +/-5%). Fluctuations within that range are considered to be acceptable. Fourteen months after my sale, Tesla has incorporated OTHER standards which they refuse to disclose. It's Tesla's standards that have created problems for some unknown percentage of Model S owners. You say 95% don't have a problem, but actually you have no idea. Your analogy regarding the Windows operating system would only apply if Microsoft sold you Windows, and then removed some of the capability that you purchased. Since you're an electrical engineer of some sort, you're just fine with tinkering and doing your own troubleshooting to make your car work as advertised. However, everyone is not an electrician, nor should they have to be to own a Model S that functions as advertised.
Regarding your comments about the transformer (down the street), or the capacitor bank (further away), these aren't even things that a homeowner owns or controls. This equipment belongs to the power company. If their experts say they are functioning properly, and none of your other residential appliances are having issues, then you're fighting a losing battle, unless you have an EE degree, lots of test equipment, and a lot of time on your hands (which isn't the case for the average Model S owner).
I find it interesting that after UMC adapters burned up, and new UMC adapters were shipped to all customers, there was a dramatic price cut on the heftier High Power Wall Connector. It's as if Tesla wants most customers to use the HPWC on a daily basis (in lieu of the UMC).
It seems odd for this capability to have been included as a HW/FW capability yet only have been activated after the garage fire I referenced above (when it almost certainly was).
Are you sure this is the specific event it specifically designed to trigger on for current reduction?
It was said by Tesla when they introduced the feature that the car looks for voltage fluctuations that would indicate wiring problems and safety hazards, in which arc faulting would sit. A wide variety of cases have been listed in the various threads here that tend to fall into two categories: 1) the "extension cord" voltage sag that occurs as the result of ramping up charging current and getting a higher voltage drop than expected - this happens nearly every time in those cases that are experiencing it, and 2) those that are more intermittent in nature. My own observations at the shop with the bad ballast were of this second type, where I can guarantee good infrastructure all the way back to a dedicated transformer, and sufficient stable voltage with no other loads at the particular shop that caused my car to back off. The repeated switching of a defective ~200W ballast - on a dedicated 120V circuit - caused the car to back off.
So the bottom line is that I can't specifically confirm that they have purchased licensing rights to an arc-faulting algorithm, but the effects of their algorithm are the same, and it aligns with that second class.
- - - Updated - - -
Yes:
http://www.nytimes.com/2014/01/15/automobiles/citing-charging-concerns-tesla-issues-a-recall.html?_r=0
The "over-the-air" update they referred to was the charging current reduction in 5.8.4. The fire report makes note of an extremely high temperature point inside the receptacle itself, indicative of a loose connection.
And it can, on a proper electrical system. I have definitive proof in my garage.
Tesla also "advertises" this:
...which means you're not guaranteed you'll get 40A charging current.
DavidM said:
I am willing to guarantee that if I attach a scope to your power lines, flip your main breaker off, then back on, that I will see voltage on your main busbars shoot up well over 300V for a sub-second period of time. Would you like to take that bet?
You're just not going to see it, because you're not looking for it. Your power company isn't going to see it, because you're looking at it averaged over time, and not the instantaneous effects.
It's clear that you're going to continue to hold your conspiracy theories about the UMC's being Tesla's dirty secret and me being some sort of agent here to suppress any liability related to it or something like that: I'm not being paid by Tesla, TMC, or anyone else to have the views that I do. I think Tesla's doing the right thing by protecting its customers, and it's primarily because my house nearly burned 15 years ago from the conditions (unrelated to Tesla) that the software feature protects against. Assuming you and Tesla have tried isolating the issue by using a loaner/replacement UMC, should you replace your UMC with a wall connector, my money would be on your car still backing off its charge current.
I think that you're an exception and not the norm, because there is only a vocal minority complaining about this on TMC and Tesla Motors forums, with tens of thousands of customers. I don't accept your assertion that most people experience it and just don't know it because they charge at night, because you can see it when you use the mobile app or if you're charging during the day. Those people who are exceptions can do one of three things - 1) accept that you're an exception and enjoy life with a beautiful piece of machinery; 2) fix the exception by finding the problem; or 3) fix the exception by going to an alternative that puts you back in the mainstream. I've told you where you can look, but you're ignoring it because a power company man in the van put his voltmeter on and said your voltage is between 228V and 252V.
My best to you in resolving your situation, should you choose that path. There's nothing else that I can offer that will change your mind, and -- to the others here -- I'll let the evidence stand for itself. I'm proud that my Model S looks after my safety.
Last edited:
I understand the grage fire is what prompted the FW change. I was the one who referred to that initially.
What I'm asking for is are you sure of what the car is triggering on to reduce current? Could it perhaps be voltage fluctuation that the car is too sensitive too or EMI that's confusing it.
Electric700
Active Member
This reminds me of a similar situation with the Chevrolet Volt, which was released in 2011. For the 2013 model year Volt, Chevrolet introduced a change where the charge current over 120 V would automatically default to 8 A, as opposed to staying at 12 A. Many people complained about the change because it required manual intervention to go back to 12 A. After Chevrolet explained that it was due to safety concerns, I think most people understood that it was for their own protection.
I really like how Tesla decided to do it with the Model S, where the current is lowered if the software detects potential issues with the line.
Without access to the algorithm, I can't tell you definitively. This is based upon observation and the evidence that I've seen here. We can draw some conclusions from the implementation, cases, and the frequency of its occurrence as reported here and in the TM forums.
If it were EMI-triggered, you would have a far greater occurrence of the cars backing down given the wide variety of environments - residential, commercial, industrial. My home, for example, is a case where EMI is a monster and if it were the case I would expect to see the car *hate* my home. I have a couple racks of 1RU electronic IT gear (with lots of wireless), I have a machine shop with plenty of motor loads, I use PowerLine Ethernet for connections to outbuildings. As a sign of a troubling EMI environment, my X10 & Insteon automation only works occasionally (and I haven't bothered to track down the EMI interference to apply filters, but I'm guessing it would be rather expensive), despite using coupler/repeaters. My electric service is an EMI hell.
EDIT to add: If it were just a long-average voltage fluctuation, you would see it backing down far more often than it does, too. I recall one person here reporting that the power company had placed its line analyzer on the bus bars and found no more than a +/- 2V variance, with the car backing down sometimes and not backing down others. That would point away from a super-sensitive voltage drop algorithm and more toward that sub-second arc-fault type detection I spoke of earlier. I experience busbar voltage between 245V - 250V based on how loaded my service is at any given time.
So, to your other question - is the car too sensitive? Well, that's not a method question, but rather a quantity question. I believe we have enough evidence to conclude it's based on voltage (Tesla has said "voltage fluctuations" a number of times, along with what the evidence has suggested). It is difficult to balance: if you reduce sensitivity, you will miss some potential wiring problems until they progress into larger problems. It may be that the charger can't easily tell the difference between a reactive voltage surge and fall created by another appliance, vs. measuring the car's own effect on that. There is the potential that a redesign might help with that. However, I don't think we should continue to go down this path much further, because while I'm relatively confident that the evidence lines up with the theory that it's looking for an arc-fault-like signature, I'm less confident in how Tesla measures it or solutions for it.
I'm willing to accept that I'm wrong, by the way, and that it's something we haven't even considered... but there's only a limited number of things that the car can "see" attached to a branch circuit -- and we've largely covered them.
Last edited:
Apple also. And any Microsoft based PC. As well as any iPad, tablet, iPhone, android... Apps that don't keep pace with the OS will eventually become inoperable, or so neutered that they are useless. Users who don't keep pace with OS changes will lose functionality, access to new apps and features, and - in some cases - warranty support.
Come to think of it, this also happens with ANY software updated device. My Nest thermostat automatically downloads new software. It processed a download last winter that had benefit of connecting with smoke detectors, but in process could no longer keep battery charged without a hot lead from furnace... wiring requirement that was not initially there at installation.
Come to think of it, this also happens with ANY software updated device. My Nest thermostat automatically downloads new software. It processed a download last winter that had benefit of connecting with smoke detectors, but in process could no longer keep battery charged without a hot lead from furnace... wiring requirement that was not initially there at installation.
In an arc fault situation, maintaining the hazardous/undesired arc in an arc fault situation would generally be voltage dependent across the air gap, no? If it is a series arc fault (a break in one conductor) then a load would be required to make it happen, but I think perhaps my lower level electrical knowledge is failing me on grasping if a 25% reduction in *current* (thus *increasing* voltage) is sufficient to stop an arc fault situation in the first place. I'm not sure that reduction in charge current is an adequate safeguard against an arc fault anyway, which is why I'm confused about the nature of the algorithm. (Disclaimer: speculation, I could be wrong!) Every arc fault protection system I've seen to-date operates by completely interrupting the circuit.
stopcrazypp
Well-Known Member
I don't believe it prevents arcs directly, but rather reduces the risk of fire by lowering the amount of energy and heat through the conductors.
Two thoughts (now firmly in speculation territory):
First, I don't even know if they're really trying to prevent arc faults, per se - they may just use an arc fault signature as an early warning indicator of a "glowing conductor" situation. When a connection is loose, there are times you'll experience an arc and then the conductor welds itself to the terminal (can especially be true on multi-strand wire). Once that happens, the arc fault gives way to a higher-resistance "glowing" connection. This is different than an arc-fault situation on an appliance branch circuit, where typically solid wire or solid appliance plug blades will be melting, increasing the arc - that's where you definitely want current interruption.
Second, you're right that lowering the current will increase voltage and therefore increase the potential gap that could be covered with an arc; however, I suspect they're looking to protect the wiring more (and by the time an arc jumped the gap, I'd argue the voltage increase from a 25% decrease in current is relatively meaningless anyway). I think the more important element is preserving the wire -- in an arc welder, for example, higher current creates a higher electrode melt rate.
As a result, if they reduce the current by 25%, they get two things: under a case that migrates to a "glowing conductor" scenario, there's less heat through a higher-resistance junction (roughly half); and in the case of an arc, you get less of a wiring melt (although I don't even know if they're considering that as the primary issue to tackle).
Again, speculation on my part but I think that's where they're heading.
Last edited:
If the MS originally had a non-peak reading voltmeter on the screen, could TM have changed it to a peak-reading vm simply thru a software upgrade? Wouldn't that be necessary to support FlasherZ's idea that it now can sense arc faults as opposed to just voltage variations within +/- 10% or so?
If I were qwk my next step would be to pull out my old crt scope and see what noise might be riding on top of the 240vac. But apparently his problem has abated for the time being. But DavidM continues with an identical issue - (big shameless hint!)
--
If I were qwk my next step would be to pull out my old crt scope and see what noise might be riding on top of the 240vac. But apparently his problem has abated for the time being. But DavidM continues with an identical issue - (big shameless hint!)
--
SeminoleFSU
Voluntaryist
In his defense, I had the same problem. Charged at 40amps just fine before the FW upgrade came along that dialed back the amps. When I charged at 40amps, the UMCs were getting hotter than they should. My first one was so hot I felt it was unsafe and had it swapped. Second one was cooler, but still hotter than I wanted to feel it getting for longevity. Some time later the first FW came out that dialed things back, and I've just let it do it's thing. I've noticed since then that the UMC is noticeably less hot when left to charge at the dialed back current 30amps. For me the difference in time didn't matter much because I can recoup all the miles I drive in my commute to work at 30 just fine... but it is also a little annoying to me that the UMC feels hotter than it should (especially in the summer) when charging at the full 40. This is a UMC that is in a garage, in the shade too.
That's the crux of one of my questions: it seems the arc fault detection would need specific hardware support, thus it would have had to have been there all along in order for a firmware upgrade to take advantage of it.
Seems odd...
Yes.
The read-out on the display is just that - a display of a point in time. It becomes an "averaging voltmeter" that measures Vrms when Tesla chooses to display that, and with the resolution that Tesla chooses to offer.
AC doesn't have static voltage. At any given instantaneous point in time, you'll see various voltages across the mains - from 0V to 350V (normal operation), perhaps even higher than that (reactive currents). Tesla won't display that - they'll display the Vrms, the voltage that most people understand, as it would just confuse everyone without a firm understanding of AC. If the charger has an ability to read a voltage, then its ability to detect arc-fault signatures is limited to the time-resolution of its voltage sensing and the ability of its microcontroller to devote the time to doing so.
If I were qwk my next step would be to pull out my old crt scope and see what noise might be riding on top of the 240vac. But apparently his problem has abated for the time being. But DavidM continues with an identical issue - (big shameless hint!)
--
It's not about what noise is normally riding over it, but catching the event that makes it downshift -- but even then, that won't tell you what causes it. It's pretty much a given that you're seeing a voltage fluctuation of some-sort, and it's a guaranteed (in my book) the scope would show that. It might give you a clue as to whether it's a reactive current that is driving a spike in voltage vs. a drain on voltage. It would be faster to demand the service center engage engineering to help you out - they know more about the situation (and perhaps they have some event logs that detail the type of failure detected).
- - - Updated - - -
I don't want it to seem like I'm saying it isn't / won't be the UMC. I'm making an assumption that Tesla Service has already given him a loaner / replacement UMC, and we haven't determined that it's not even the charger. The UMC's clearly had a problem, but if the downshifting occurs after replacement of the UMC or on a replacement of the UMC, it follows that it's unlikely to be the UMC as a cause.
- - - Updated - - -
It's just voltage sensing, limited to the resolution of your voltage sensing loop. Let's use an average 20V inverter DC arc welder as an example, because it's an easy way to visualize it. Prior to establishing the arc, if you were to measure the voltage, you'd find an open-circuit voltage of somewhere around 80V. As the arc starts, the air gap becomes conductive and the voltage drops dramatically (to 20V) as the circuit is established and current flows. Now, that's a simplistic case and the amplitudes mentioned assume a series arc-fault that the Tesla charging circuit was involved in. However, even if there's an arc fault that occurs in another branch circuit, the effects of it can be seen across the entire system and could be sensed by the Tesla.
Since we're all pretty much topping off every night, or every other night, we're almost never "drawing 40A for 8 hours". One and a half hours at 40A would be more accurate. But yes, it's 30A or 40A continuous if the car's firmware permits. Even a 5 ton heat pump struggling to heat a house when it's below freezing outside won't continuously draw 40A for hours. Maybe just 30A. But it would do that all night and give it it's best shot at maintaining indoor temperature. Believe me, if it shut itself down, or if it's 40 degrees outside and it can't maintain 70 degrees inside, you'll be on the phone to HVAC contractor. 99% probability that when they leave your house, you're back in business for the foreseeable future.
This is certainly true for me - and actually I've specifically set the amperage down to 15-20A in order to not stress charging components.
I wish this was done automatically though - there are times when I need 40A (pretty rare), but most of the time I'd rather tell the system "make sure I'm at X% charge by the time it's Y o'clock" and let it do its thing to minimize strain on components.
-- Greg
Similar threads
