Charging voltage tolerances

[cinergi]

As a result of a problem at the house, I was able to take note of something that I thought would be interesting and worth documenting with a new thread as there's been discussion spread across various threads about what voltages the Model S charger is capable of using. For others posting, please note the circuit nominal voltage and whether this is single (split or delta/wye - U.S.) phase or 3 phase (EU).

The substation switchover at midnight last night at 6am this morning caused my voltages to temporarily read very high. I decided to let the car charge at 5 amps after the midnight event (the first one I noticed) so I could see what the 240 looked like instead of the 120 as I don't have any measuring equipment anywhere. The only reason I knew I had high voltage was my A/V power strips shut down at high voltage and have a voltage readout. And they shut down (which is why I noticed) and when they came back on they were reading 130-something. When it happen this morning it woke me up (TV and cable box clicked off) and opened the Tesla app as quickly as my groggy self could. 270 volts :scared: and still charging!
This is standard U.S. residential single/split phase (nominal 240).
(side node -- I'll be calling nstar and trying to figure out how to explain this to someone)

 

[tom66]

The car is rated 85 to 265VAC. So 270VAC is out of tolerance, but shouldn't cause any harm. If it were 300VAC+ I suspect the EVSE/Mobile connector contactors would disengage rapidly.

 

[Cottonwood]

The car is rated 85 to 265VAC. So 270VAC is out of tolerance, but shouldn't cause any harm. If it were 300VAC+ I suspect the EVSE/Mobile connector contactors would disengage rapidly.

The basic 10kW charger is also used in Superchargers. In Superchargers, the charger appears to be connected to 277 Volts nominal in the 120 kW Superchargers, Line to Neutral, (480V Delta, 277V Wye, 3-Phase connections). If it is 277V nominal, then the charger must be able to take a reasonable percentage over 277V for normal line variations.

 

[scaesare]

There have been a couple of discussions here about feeding the Model S 277V via the HPWC. That actually would make the chargers in the Model S capable of about 11kW each.

I suspect this to be true (despite the HPWC manual not specifying anything above 240), because the superchargers are known to have a stack of 12 chargers in them. If you work out the input voltage for the superchargers, each if the chargers in the stack is being fed 277V/40A.

We also have seen slightly higher than 120kW charge rates (I've seen 123kW, and others 124...), indicating that the chargers can produce more than 10kW power...

 

[DavidM]

With the 5.8.4 software change, my Model S routinely reduces amps from 40A to 30A, because it sees slightly high voltage (246 - 250V). I also get the dreaded "charging problems" message on the dashboard. No issues whatsoever in the 12 months and 12,000 miles before the software change. No messages, no melting, nothing. Now, on rare occasions, charging stops completely, before the session should normally finish. Does Tesla think that putting a fusible link in the adapter will give them the confidence to go back to the old tolerances, or are all Model S customers just stuck with the new tolerances forever - resulting in crippled charging capability for about 40% of all customers (according to the survey).

 

[FlasherZ]

With the 5.8.4 software change, my Model S routinely reduces amps from 40A to 30A, because it sees slightly high voltage (246 - 250V). I also get the dreaded "charging problems" message on the dashboard. No issues whatsoever in the 12 months and 12,000 miles before the software change. No messages, no melting, nothing. Now, on rare occasions, charging stops completely, before the session should normally finish. Does Tesla think that putting a fusible link in the adapter will give them the confidence to go back to the old tolerances, or are all Model S customers just stuck with the new tolerances forever - resulting in crippled charging capability for about 40% of all customers (according to the survey).

Your reduction is not caused by higher voltage. When I plug the car in, my car sees 249-250V and at 80A drops to 242-243V. In your case, the Tesla is sensing a fluctuation in the voltage and backing off. This could be due to a number of things - the first one they're concerned about is the high-resistance problem that generates heat, but there are other reasons (e.g., your service conductors could be too small and the car backs off when another large appliance starts up).

- - - Updated - - -

There have been a couple of discussions here about feeding the Model S 277V via the HPWC. That actually would make the chargers in the Model S capable of about 11kW each.

Keep in mind there are a few components at play here. While it is speculated that the SpC's use the same chargers as the cars, and they are capable of doing 277V (for SpC's), there are other elements at play... SpC's don't use HPWC or UMC electronics - the UMC and HPWC have intelligent electronics with a switching power supply, and a contactor. I doubt that either of them use parts rated for 277V.

That said, perhaps their tolerances allow them to work, but the lifespan may be shortened if run at 277V. Charging will be faster, though.

- - - Updated - - -

The substation switchover at midnight last night at 6am this morning caused my voltages to temporarily read very high.

Generally, power companies have protection against this, because this means that the distribution voltage was extremely high (16.2 kV vs. 14.4, or 8.1 kV vs. 7.2) and every house on that distribution line got the same voltage. That would be a lot of liability for the power company! I'm very surprised at this.

I have experienced the opposite direction, where my voltage sank to 220V due to a failed component in the substation. The power company fixed it the same day and thanked me for reporting it (it's still within tolerances, so no alarms went off). It's fairly easy to see a 15% reduction in your charge power.

 

[Cottonwood]

With the 5.8.4 software change, my Model S routinely reduces amps from 40A to 30A, because it sees slightly high voltage (246 - 250V). I also get the dreaded "charging problems" message on the dashboard. No issues whatsoever in the 12 months and 12,000 miles before the software change. No messages, no melting, nothing. Now, on rare occasions, charging stops completely, before the session should normally finish. Does Tesla think that putting a fusible link in the adapter will give them the confidence to go back to the old tolerances, or are all Model S customers just stuck with the new tolerances forever - resulting in crippled charging capability for about 40% of all customers (according to the survey).

Must be something else. With 5.8.4 and up, 250V that droops to 246 at 80A, and an HPWC I have had no problems with current reduction.

My normal sunny afternoon line Voltage is about 250V with 24 kW of solar panels feeding into the panel the same sub-panel as the HPWC. On nominal, evenings, I have about 247V, and when the ETS heaters are sucking their full 192 Amps (4 60A breakers on different sub-panels than the HPWC), the Voltage at the HPWC drops to 244V or so. All of these Voltages are measured by the MS with 0A; they are all 4V less with 80A. None of this has caused a reduction of charging current on my MS.

 

[wycolo]

Whatever happened to nominal 117/234 vac as a North American standard? Are you folks able to crank your solar systems up to a hotter setting perhaps?
--

 

[Cottonwood]

Whatever happened to nominal 117/234 vac as a North American standard? Are you folks able to crank your solar systems up to a hotter setting perhaps?
--

My co-op runs on a nominal 120V/240V split phase standard. I don't know what is with the line to my house and/or my transformer, but I have always run a little hot at 247V; that's only 3% high, but I would rather be high than low. The solar kicks that nominal up and the ETS sucks it down, more at their respective sub-panels, and less at the main feed.

I am far enough out in the woods that I have my own transformer. With the almost 50kW of ETS heat, 20kW of HPWC, steam shower, extra 14-50's in the garage, 24kW of solar PV, etc, my electrician talked my co-op into putting a 100 kVA transformer in for me, so besides the Voltage being a little on the high side, it is pretty stiff (low resistance) and does not vary much with relatively large currents.

 

[FlasherZ]

Whatever happened to nominal 117/234 vac as a North American standard? Are you folks able to crank your solar systems up to a hotter setting perhaps?
--

It's a complicated topic. The targeted nominal voltage is 240V on a split-phase household service, so most power companies target that in their settings. However, getting there can be a challenge. Most distribution transformers have multiple taps that allow them to take into account voltage drop over the distribution run, and they may in practice deliver a higher voltage than the engineering/design work figured.

In addition, those who take advantage of distributed generation (solar, wind, etc.) with grid-tie inverters can drive up the RMS voltage on the secondary side of the transformer.

ANSI C84.1-2011 specifies a nominal voltage of 120V above ground for each leg with a 5% tolerance. That gives a tolerance range of 114V-126V, or 228V-252V across both legs. So I'm still within the tolerance (my meter shows my home at 248V, Tesla shows 249-250V).

 

[Cottonwood]

...

In addition, those who take advantage of distributed generation (solar, wind, etc.) with grid-tie inverters can drive up the RMS voltage on the secondary side of the transformer.

ANSI C84.1-2011 specifies a nominal voltage of 120V above ground for each leg with a 5% tolerance. That gives a tolerance range of 114V-126V, or 228V-252V across both legs. So I'm still within the tolerance (my meter shows my home at 248V, Tesla shows 249-250V).

Thanks for the reference!

Glad that I am staying within tolerances, at least with respect to Voltages...

 

[DavidM]

Unfortunately, nobody knows what Tesla's new software tolerances are because it's a closely held secret. Makes it difficult to design a compliant solution using the Universal cable and adapter that came with the car. It is also not known if software tolerances are less strict in environments where the HPW is used (less risky hardware). I appreciate all the guesswork, but you can't practice your pole vault in the dark.

 

[wycolo]

Was the 117 vac referred to as RMS in the old days - is that where the '117' came from?

@DavidM - You could hook up a buck/boost transformer in 'buck' mode to reduce your 'hot' 250 by 8% (I think they do ~8%). Then your MS might be happy. You might find one at a local electrician, scrap dealer or even borrow one from a 60 cycle house.

I've always had my own pole pig since 1981 so I'm spoiled rotten. :smile: The most I've seen here is 241v and @30 amps the Teslas have dropped to 236 or 237 with 235 the lowest they've shown.
--

 

[FlasherZ]

Was the 117 vac referred to as RMS in the old days - is that where the '117' came from?

A/C voltage has always been expressed RMS because it's the most effective way to use it in calculations. Peak waveform voltage will be about 169V with V(rms) = 120V. To my knowledge, 117 was never an "official" number, although perhaps it was a target of a local power company.

@DavidM - You could hook up a buck/boost transformer in 'buck' mode to reduce your 'hot' 250 by 8% (I think they do ~8%). Then your MS might be happy. You might find one at a local electrician, scrap dealer or even borrow one from a 60 cycle house.

If it were the high voltage that caused the Tesla to throttle back, then I imagine I would be seeing my car throttle back - and it simply has not. Tesla has said it is fluctuations in voltage that causes the dial-back, not the voltage itself. As a result, I'd start looking for high-resistance points or overloaded points. I'd consider turning off a few major loads to see if I could isolate it. I might even start with only a single lighting circuit and the Tesla as the only loads on your service to see if it still throttles back. If it does, then I'd start involving the power company. If it doesn't, then I'd start turning loads on one by one to determine at what point you see the fluctuations start. I'd take a look at the size of the service conductors, and see what the power company suggests once I've determined that my appliances were operating correctly. (In my case, prior to my upgrade I had 200A service across 100 ft. on 2/0 AL conductors, which are listed as good to 135A in NEC charts... and it needed to be upgraded.)

I've always had my own pole pig since 1981 so I'm spoiled rotten. :smile: The most I've seen here is 241v and @30 amps the Teslas have dropped to 236 or 237 with 235 the lowest they've shown.

I live far enough out to have the same thing. I don't mind that mine is configured to run @ 248V, as most of my loads are switching loads (even most of my light bulbs). Only the oven and emergency heat strips are resistance heating loads.

 

[DavidM]

Thanks to "wycolo" and "FlasherZ" for your suggestions. I've got a 200A service panel, and several 240V appliances (2 high efficiency heat pumps, oven, clothes dryer, 50A circuit to Model S). My local electric utility company has paid me a visit and "blessed" both the voltages and minor fluctuations at the service panel. We've been in the house for 13 years, and none of the 240V appliances have been negatively affected by the electrical service. For the 12 months prior to the new 5.8.4 update to the Tesla thresholds, my Model S has been quite happy with the 50A circuit. I charged without issue at both 30A and 40A settings. After 5.8.4 that all changed. I won't be spending another dime without:
1. Receiving documents from Tesla with the specific new thresholds.
2. Communication from Tesla indicating whether or not the new thresholds are temporary or permanent.

It would be foolish to spend money on potential band-aid fixes, only to find out that Tesla winds up going back to the old tolerances after distributing a better quality charging adapter for the Universal Cable.

 

[FlasherZ]

My local electric utility company has paid me a visit and "blessed" both the voltages and minor fluctuations at the service panel. We've been in the house for 13 years, and none of the 240V appliances have been negatively affected by the electrical service.

The appliances you listed are all motor or resistance loads that are not sensitive to fluctuations, I wouldn't expect you'd have any issues with them. How large is your power company, and did they just touch a meter on the bus bar and sit there for 30 seconds, or did they attach a power quality meter and let it run for a day or so? In tough cases of power quality problems, the latter is an important step.

Unfortunately, a lot of power company technicians still live in the old world where loads to them are only resistance (light bulb or heat) and motors. To them, electronic loads are small and connect only to household receptacles. The notion of a large electronic load (40A/80A continuous) is pretty rare in most homes.

For the 12 months prior to the new 5.8.4 update to the Tesla thresholds, my Model S has been quite happy with the 50A circuit. I charged without issue at both 30A and 40A settings. After 5.8.4 that all changed.

I think it's fairly smart to have the voltage fluctuation protection, as it should prevent a number of fires due to high currents through high-resistance connections.

I won't be spending another dime without:
1. Receiving documents from Tesla with the specific new thresholds.
2. Communication from Tesla indicating whether or not the new thresholds are temporary or permanent.

It would be foolish to spend money on potential band-aid fixes, only to find out that Tesla winds up going back to the old tolerances after distributing a better quality charging adapter for the Universal Cable.

I think that's the wrong attitude to have, in my opinion. I don't believe Tesla will go back to the old tolerances (although perhaps they would relax them slightly), and the reasoning is simple - the new adapter's heat-fuse will only detect problems at the receptacle where the car is plugged in. If the car is backing off, it's because you have a source of resistance *somewhere* in your house. It may be the service conductors, it may be a malfunctioning or misbehaving appliance, or it could still be a high-resistance issue.

At a minimum, I'd follow some of the troubleshooting steps above - isolate your home so that only the Tesla and a lighting circuit are on, and see if it still drops. Then follow where that leads.

 

[DavidM]

@FlasherZ - My power company is Florida Power and Light (FPL). They're pretty big in this region. They came in and attached a special meter that monitors voltage and fluctuation. It was attached for 5 days. During that time, there were some fluctuations, but nothing out of their tolerances (230V - 252V). In the report, I could see the voltage fluctuate when a major appliance cycles on and off. We also have a "whole house" surge protector, which is attached to the main panel, and provides some protection against major spikes due to thunderstorms during our hurricane season.

I don't know why the new Tesla thresholds have to be a secret. But a surgeon can't perform a successful surgery if he/she doesn't know exactly what needs to be remedied. It is impossible to remove all voltage fluctuations. The hardware and software are designed to handle fluctuations within some margin (undisclosed). They have also presumably been designed to handle 40 Amps of heat for hours at a time. Although the adapter may be lacking. And the jury is out on the cable. The Tesla cable actually has a smaller conductor/insulator than the cable that came with my 3kW Honda generator. The Tesla universal cable is supposed to carry 3x the current at 9.6kW. The Tesla cable gets warm, the Honda cable doesn't.

 

[FlasherZ]

@FlasherZ - My power company is Florida Power and Light (FPL). They're pretty big in this region. They came in and attached a special meter that monitors voltage and fluctuation. It was attached for 5 days. During that time, there were some fluctuations, but nothing out of their tolerances (230V - 252V). In the report, I could see the voltage fluctuate when a major appliance cycles on and off. We also have a "whole house" surge protector, which is attached to the main panel, and provides some protection against major spikes due to thunderstorms during our hurricane season.

Good - this is the type of analyzer I hope they would use. If they connected it to the main panel and they saw a fluctuation when a major appliance turned on/off, I'd start looking at the magnitude of that fluctuation. It could very well be that you have smaller-than-usual service conductors combined with a longer path to the transformer combined with a failing start capacitor on one of your heat-pump units (that causes the instantaneous load to shoot up to 150A or even 200A for more than a second or so). Even if you knew the fluctuation parameters that the Tesla used, it wouldn't solve your problem. Yes, it would able you to argue with Tesla the thresholds, but at the end of the day, your home electrical system is seeing a voltage fluctuation that is outside the "norms" that Tesla has observed. You need to fix those fluctuations if you want 100% charging current. If all you want to do is argue with Tesla, then I suppose you could continue banging the drum about knowing the specific thresholds.

This combination of factors is why each case can be different, and why you need to rule out various cases. It could be a single motor appliance that has a bad start capacitor or bearings that are going bad; it could be that your service conductors are horribly undersized, like mine were before the upgrade. It could be a combination of them - let's say for a second that my home was 300 ft from my transformer (0.6 kft round trip), that I had 2/0 cable feeding it (0.128 Ohms/kft), and that my heat pump had a 200A start-up load for 1.5 seconds (because it had a capacitor that was starting to go bad). My voltage drop would be V=IR, so V = 200A * (.6 * .128) or 15.36V during that 1.5 seconds. If, during charging, the Tesla saw my voltage drop by 15.5 volts (what we typically saw as "dimming lights" with incandescents) for one and a half seconds, I'd hope it would back off for the sake of safety.

I don't know why the new Tesla thresholds have to be a secret. But a surgeon can't perform a successful surgery if he/she doesn't know exactly what needs to be remedied. It is impossible to remove all voltage fluctuations. The hardware and software are designed to handle fluctuations within some margin (undisclosed). They have also presumably been designed to handle 40 Amps of heat for hours at a time. Although the adapter may be lacking. And the jury is out on the cable. The Tesla cable actually has a smaller conductor/insulator than the cable that came with my 3kW Honda generator. The Tesla universal cable is supposed to carry 3x the current at 9.6kW. The Tesla cable gets warm, the Honda cable doesn't.

It's not as simple as "handling 40 amps of heat". Heat is resistance times the square of current. If you have a high resistance (due to a loose connection somewhere), the heat generated will be tremendous and can start a fire. Yes, a smaller cable will be warmer - and Tesla uses 2 parallel, smaller conductors from the plug to the UMC and to the car. That will generate some heat, but it's nothing like the heat generated from a loose connection.

Yes, voltage fluctuations will always happen, but it's not as simple as knowing what the thresholds are. It's not like you will tune some appliances to remove just enough to keep it from happening. I have found that Tesla's thresholds are pretty reasonable - maybe they're on the conservative side, but I'd rather they protect my home than be reckless about it. It's obvious that you have something that is triggering a fluctuation (or a loose connection that makes it non-deterministic), so do some troubleshooting to narrow it down then make some decisions. Turn off all other loads and see if the Tesla charging rate stabilizes. If it doesn't, then you'll have to look at your service conductors and/or transformer loading, and you may have to argue with the power company. If it does, then start introducing loads to see what causes the Tesla to back off. Start with biggest loads first... the first time you see it back off, isolate that load for a while and see if the rest of your loads allow it to be steady. Then you know where to start looking - whether it's an aggregate problem (service conductors) or an appliance problem.

Bottom line: even if you knew the thresholds, you'd still have to perform troubleshooting to find the cause of the fluctuation.

 

[DavidM]

@FlasherZ - "Bottom line: even if you knew the thresholds, you'd still have to perform troubleshooting to find the cause of the fluctuation."

If I knew the thresholds, then my troubleshooting money would be well spent, because I would know what to look for. The survey on this site indicated that 35% - 40% of Model S owners are experiencing reduced current charging, or error messages at the conclusion of a charging session. The products that Tesla sells need to work with a wide majority of consumers and home electrical environments. Some owners are even getting error messages and reduced current at public charging stations. Yet you seem sure that "Tesla's new thresholds are pretty reasonable". Would that imply that their previous thresholds were reckless? I wouldn't think so. I believe their initial thresholds were deemed to be reasonable, while the new thresholds reduce their liability and buy them time while the adapter is redesigned and distributed. It must be a bigger problem than originally thought, because the adapters were to start shipping in January. I'm not aware they exist yet.

The software was always designed to detect an out of tolerance voltage fluctuation, stop charging, wait a few minutes, then resume charging. Presumably, that design adequately protected everything. But maybe they're not sure it's enough.

btw - my heat pumps are less than 1 year old.

 

[FlasherZ]

If I knew the thresholds, then my troubleshooting money would be well spent, because I would know what to look for. The survey on this site indicated that 35% - 40% of Model S owners are experiencing reduced current charging, or error messages at the conclusion of a charging session. The products that Tesla sells need to work with a wide majority of consumers and home electrical environments.

I would argue that the products that Tesla sells also need to work safely, and that perhaps the charging loads are stretching certain configurations of the home grid. As I mentioned previously, without upgrading my service to my home there was no way that the Tesla was going to be able to charge (at 40A or 80A) without dimming my lights, because I was already stressing my transformer and service cables even without the Model S. My previous home electrical service would have made the car back down, I'm sure of it, because of the constrained conductors and smaller transformer.

Some owners are even getting error messages and reduced current at public charging stations.

A transformer that is too small, or a loose connection will cause this as well. What makes you think that public charging stations are immune from capacity issues or bad wiring?

Yet you seem sure that "Tesla's new thresholds are pretty reasonable". Would that imply that their previous thresholds were reckless? I wouldn't think so.

Considering they had no thresholds previously (other than the 120V extension cord voltage drop), I would say that I'm happy Tesla has added capability to protect my home. They did not back off charging current previously AT ALL and made the assumption that when you set it at 40A, you could draw that without a concern.

Let me try and make it clear, since you keep making reference to the new adapters in relation to the thresholds used: new adapters will not address the voltage fluctuations in most cases. It will only address the cases where there are loose connections at the plug and receptacle that the UMC plugs into. The new adapter will not address cases where the HPWC has a loose connection, or where service conductors are too small and almost start the home on fire (previous case was documented on this forum), or where the car begins to smoke a transformer.

The software was always designed to detect an out of tolerance voltage fluctuation, stop charging, wait a few minutes, then resume charging. Presumably, that design adequately protected everything. But maybe they're not sure it's enough.

In my experience, it was not. It would stop for a few moments in the case of a complete power interruption (and resume after expiration of a random 15-minute timer), but it would not adjust based on a voltage reduction that would signal a high-resistance problem. And I'd still like Tesla to address the fact that the ring will turn red and fail to try recharging again after my co-op's "blink-blink-blink" short-circuit protection needs more than 2 tries to close the circuit again.

btw - my heat pumps are less than 1 year old.

I've seen bad capacitors right out of the box. I've seen compressors that have been shifted such that their oil gets into the coolant circuit. That doesn't mean there isn't a problem.

I don't get your resistance to finding the voltage fluctuation within your home when you've already cited a very informal survey that suggests that a supermajority of people don't experience a back-off. Narrowing the problem doesn't take anything more than switching off breakers (so this "cost" you speak of is little more than some time), and it'll help you identify if there is a problem with one of your appliances. Why insist that Tesla eliminate a safety net that will help prevent home fires?

And finally, as I've mentioned several times here - just because it works doesn't mean it's safe. Just because Tesla previously made the blind assumption that any charging setup set at a given current setting could handle it without any validation, doesn't mean it can't improve that in the future.

If the Tesla chargers are backing off their current, you have an electrical system that is experiencing an issue either due to strained infrastructure or a badly behaving appliance. This isn't Tesla's fault. Find it, fix it, and the car won't back down anymore.