Amps and efficiency on the UMC

[LukeT]

Having tried various current settings on a 3 pin plug it looks to me like efficiency (according to teslafi) drops off quite dramatically at the top end. So much so that while moving up from 7 to 8 to 9 amps is probably beneficial, going from 9 to 10 drives such a reduction in efficiency that the extra current used equates to effectively no extra charge.

I base this on an efficiency of 86-88% at 9A against c.76% at 10A. This is on only a few nights' observations but the numbers seem consistent from charge to charge. Have others found smiler? Are my numbers representative?

Certainly at 9A I get no noticeable significant heat out of mobile charger, cables, plugs, sockets etc.

 

[MrBadger]

Interesting. Are those readings at similar state of charge? Your findings goes against my understanding that higher current gives higher efficiency.

I'll probably end up with a current monitor and/or energy meter on my charge point and link it to the Tesla api and/or Tesla wall charger to do some form of home brew smart charging.

 

[LukeT]

Interesting. Are those readings at similar state of charge?

I'll probably end up with a current monitor and/or energy meter on my charge point and link it to the Tesla api and/or Tesla wall charger to do some form of home brew smart charging.

I would say so. They tend to be plugging it in around 50-60% and charging up to 80%. Various snapshots of short charges tend to show the same numbers although occasionally dramatically different, but I've not seen dramatically different except where the charge only lasts a few mins.

 

[mongo]

Having tried various current settings on a 3 pin plug it looks to me like efficiency (according to teslafi) drops off quite dramatically at the top end. So much so that while moving up from 7 to 8 to 9 amps is probably beneficial, going from 9 to 10 drives such a reduction in efficiency that the extra current used equates to effectively no extra charge.

I base this on an efficiency of 86-88% at 9A against c.76% at 10A. This is on only a few nights' observations but the numbers seem consistent from charge to charge. Have others found smiler? Are my numbers representative?

Certainly at 9A I get no noticeable significant heat out of mobile charger, cables, plugs, sockets etc.

The charger does not have a huge efficiency drop off.
88% of 9A = 7.92, 76% of 10A = 7.6
The only way this data makes sense is if that additional charge current causes the AC to turn on to cool the pack. Otherwise, the extra current should only be going to the pack.

 

[LukeT]

The charger does not have a huge efficiency drop off.
88% of 9A = 7.92, 76% of 10A = 7.6
The only way this data makes sense is if that additional charge current causes the AC to turn on to cool the pack. Otherwise, the extra current should only be going to the pack.

Yeah I agree the implied decrease in charge rate makes no sense. To get the numbers accurate would of course need a big number of data points and I only have 2 or 3 at 10A and a few more at 9A (which has become my habit).

However, I think heat from cables is non-linear to charge, so can imagine how it may be insignificant at one rate and quite significant with a relatively modest increase, such that the increase in energy input when increasing amps may drive only a much smaller increase in energy getting usefully into the battery. Equally, cables aren't everything - maybe other parts are worse for this?

Seems unlikely pack cooling kicks in on 2kw in a car that takes 120 at a supercharger?

 

[mongo]

Yeah I agree the implied decrease in charge rate makes no sense. To get the numbers accurate would of course need a big number of data points and I only have 2 or 3 at 10A and a few more at 9A (which has become my habit).

However, I think heat from cables is non-linear to charge, so can imagine how it may be insignificant at one rate and quite significant with a relatively modest increase, such that the increase in energy input when increasing amps may drive only a much smaller increase in energy getting usefully into the battery. Equally, cables aren't everything - maybe other parts are worse for this?

Seems unlikely pack cooling kicks in on 2kw in a car that takes 120 at a supercharger?

Cable loss is resistance times the square of current, and cable loss should be less that charger loss.
Even if all the loss was due to resistance: If you had 10% loss at 9A, then .1 = 9^2*R = .001234 at 10 Amp = 12.3% loss
9*90% = 8.1 10*87.7 = 8.77 still higher power delivered

At the lower power region, higher current should be better due to fixed loads from being on.

Were you at the same SoC for each test?

 

[LukeT]

Were you at the same SoC for each test?

Over 50% always (I think) and definitely below 80%. But as I said, very small sample size.

Would be interested to know what efficiency readings others tend to see at 10A anyway.

Also, what is TeslaFi" efficiency" exactly? I'm not sure, not looked into that.

 

[Artiste]

The science is over my head but I’m interested in charging by UMC. I’ll almost exclusively be charging at work because it’s free so I haven’t even ordered a home charging point. But I will probably have to top up at home and I’ll be plugging the UMC in via a TP Link smart socket so I can monitor energy usage.

I’ve heard conflicting reports that very slow charging, such as by UMC, is either good or bad for the battery. Is there any consensus?

 

[mongo]

I’ve heard conflicting reports that very slow charging, such as by UMC, is either good or bad for the battery. Is there any consensus?

It's fine to do fast or slow with the UMC. If you do a lot of slow you msy need to go to 90%+ occasionally to reset the charge level meter.

 

[zap fizzle]

This is an interesting post.

Is it possible there is an error in measurement?

EV AC charging is generally most efficient at maximum rated current, or perhaps just under it. This is because rectifiers are generally engineered to be most efficient at around their rated current, and because faster charging means less time spent with auxiliaries such as coolant pumps and computers powered up.

There is some complexity around this - eg how many chargers a Tesla has (there have been different designs over the years) and how they're configured when on single phase vs 3-phase, but it should still hold true.

Other people have examined charging efficiency and found it poor at low amps and increasingly roughly linearly up to just short of maximum.

I spent a week on UMC off a high quality metal-cased bs1363 socket recently, also experimenting with a bs1363-to-commando adaptor which allowed current to be increased above 10 amps (DANGEROUS). 10 amps produces moderate heat in the plug, wheres 8 amps gives a stone-cold plug. This heating is from the fuse in the plug; every other part of the system remains stone-cold. It's a few watts of heating at most - not sufficient to reduce charging efficiency on its own; as a thought experiment, imagine even 20W of sustained heat inside a plug - not pretty - it's only a tiny bit of waste.

At 12A, the plug is HOT after an hour and still heating. Definitely an impending fire.

I could imagine the "extra amp is sufficient to bring the coolant pumps on " scenario, but in my case there was no coolant pump activity at any speed, even when the car had recently been driven hard. On my home 32A charger I often note the coolant pumps running during charging.

 

[David99]

Not sure how the M3 charger is built. I know the Model S/X is made out of 2 or 3 separate units. Each unit is most efficient at full load. If you dial down the charge current, you might have one of the units at full load and the second one at partial load which makes the total efficiency a little lower. Doing tests with different currents might reveal where the switch points are. Theoretically they would be at 1/3 points.

Charging st the highest current would be the most efficient since the charger runs entirely at full load and the total charge time is shortest which means the secondary losses (coolant pump, Fans) are the least.

 

[WannabeOwner]

Having tried various current settings on a 3 pin plug it looks to me like efficiency (according to teslafi)

I have never known how TeslaFi calculates it (might be on their support forum?). Given that TeslaFi cannot see the "supply side" ... perhaps what is needed is an actual measure supply side compared with the kWh gained by battery?

 

[LukeT]

This is an interesting post.

Is it possible there is an error in measurement?

EV AC charging is generally most efficient at maximum rated current, or perhaps just under it. This is because rectifiers are generally engineered to be most efficient at around their rated current, and because faster charging means less time spent with auxiliaries such as coolant pumps and computers powered up.

There is some complexity around this - eg how many chargers a Tesla has (there have been different designs over the years) and how they're configured when on single phase vs 3-phase, but it should still hold true.

Other people have examined charging efficiency and found it poor at low amps and increasingly roughly linearly up to just short of maximum.

I spent a week on UMC off a high quality metal-cased bs1363 socket recently, also experimenting with a bs1363-to-commando adaptor which allowed current to be increased above 10 amps (DANGEROUS). 10 amps produces moderate heat in the plug, wheres 8 amps gives a stone-cold plug. This heating is from the fuse in the plug; every other part of the system remains stone-cold. It's a few watts of heating at most - not sufficient to reduce charging efficiency on its own; as a thought experiment, imagine even 20W of sustained heat inside a plug - not pretty - it's only a tiny bit of waste.

At 12A, the plug is HOT after an hour and still heating. Definitely an impending fire.

I could imagine the "extra amp is sufficient to bring the coolant pumps on " scenario, but in my case there was no coolant pump activity at any speed, even when the car had recently been driven hard. On my home 32A charger I often note the coolant pumps running during charging.

It is I think quite possible it's a data anomaly, simply because of the small number of data points. I will gather data of charges at different currents and see if a clear trend is there once I have more.

Tonight I'm at a different house with a garage with a tendency to trip so dropped it at 7A overnight. It tripped at 9A after an hour or so.

Not sure how the M3 charger is built. I know the Model S/X is made out of 2 or 3 separate units. Each unit is most efficient at full load. If you dial down the charge current, you might have one of the units at full load and the second one at partial load which makes the total efficiency a little lower. Doing tests with different currents might reveal where the switch points are. Theoretically they would be at 1/3 points.

Charging st the highest current would be the most efficient since the charger runs entirely at full load and the total charge time is shortest which means the secondary losses (coolant pump, Fans) are the least.

This is a new Model S. The UMC goes up to 32A although I haven't tried that yet. I think it may be a new generation of UMC.

 

[MrBadger]

But I will probably have to top up at home and I’ll be plugging the UMC in via a TP Link smart socket so I can monitor energy usage.

Check the max power of that. Sometimes these remote controlled switches have reduced power capacity, like 10A/2400W is quite common and TBH I would not want to be at the limit of something for extended periods of time.

 

[Artiste]

Check the max power of that. Sometimes these remote controlled switches have reduced power capacity, like 10A/2400W is quite common and TBH I would not want to be at the limit of something for extended periods of time.

A good point, but I have checked. Last winter I ran a 3KW heater off one, sometimes for several hours a day. It consistently showed a power consumption of 2960W, and it never tripped or got warm to the touch. 10A should be no problem.

 

[LukeT]

I have never known how TeslaFi calculates it (might be on their support forum?). Given that TeslaFi cannot see the "supply side" ... perhaps what is needed is an actual measure supply side compared with the kWh gained by battery?

Nor me. It claims to measure "kwh used" and "kwh added" but I don't see how it measures kwh at the plug. We also have a journey on TeslaFi with "efficiency" over 100%, so I do wonder...