Internal Charger Ch

[TheDon]

I know this was most likely asked and answered but I can't find it....

When charging with the Mobile Charger (14-50, 10-30, ...) how much charging loss is there. I know that the car tells me how many kW that were added to the battery, but how much is wasted in heat, running the charger and the like.
(I have a M3 AWD 2019)

Thanks....

 

[edigest]

The actual efficiency will vary depending on a number of factors so I doubt there is a simple formula. But, in general, higher amperage is more efficient and therefore lower cost overall. For example when charging at 12A I see charge efficiency between 70% and 80%. Charging at 32A, I see just over 90% efficiency consistently.

 

[ZOMGVTEK]

With the regular mobile connector included, the most efficient point is likely around 20-25A. This is mostly speculation, and the real world total difference in losses from say 20A to 32A is unlikely to be significant at all. The mobile connector itself probably consumes 2-4W, the main losses are the vehicles charger, pumps, various systems that need to be on, and the conductor losses. Most chargers efficiency peaks somewhere around 40-80% load, depending on the design. But there’s other factors to consider as there’s significant losses from the car being ‘awake’, and the conductor losses are moderate as well. Usually the efficiency is crummy initially, gets fairly good quickly and flattens out, and starts to dip a touch as it gets near the limit. But the car takes a few hundred watts sitting there ‘on’, so longer charge times means longer on time. Plus there is the consideration of lithium plating and the physical time that makes very low currents silly anyways.

If you’re saying you don’t need the car charged quickly and would prefer to save a few pennies, I’d say set it to 28A. The difference is most likely absolutely trivial and I don’t think anyone accurately measured it yet.

 

[TheDon]

... For example when charging at 12A I see charge efficiency between 70% and 80%. Charging at 32A, I see just over 90% efficiency consistently.

How are you calculating/obtaining the efficiency %? Is there something from the car I can use to figure that out?

 

[TheDon]

... If you’re saying you don’t need the car charged quickly and would prefer to save a few pennies, I’d say set it to 28A. The difference is most likely absolutely trivial and I don’t think anyone accurately measured it yet.

The 14-50 is limited to 32A and the 10-30 at 24A. This charge is most likely done overnights so either 28-32A? Just trying to get a handle on REAL charging costs. The car says how much it stuffed into the battery. I don't see how much the overhead of cooling, cable/connector and charger losses all add up to. If it's like 10% then it's like $0.83 for me to do a larger charge.

Is there a way to get data from the car to get me a better idea of what my car is doing?

 

[edigest]

How are you calculating/obtaining the efficiency %? Is there something from the car I can use to figure that out?

TeslaFi reports the efficiency of each charge.

 

[darth_vad3r]

The actual efficiency will vary depending on a number of factors so I doubt there is a simple formula. But, in general, higher amperage is more efficient and therefore lower cost overall. For example when charging at 12A I see charge efficiency between 70% and 80%. Charging at 32A, I see just over 90% efficiency consistently.

There actually appears to be a pretty decent ballpark simple formula actually.

Use about 95% inherent onboard charging efficiency and subtract 250-350 W of constant overhead.

This is a table I whipped up to illustrate this theoretical model.
You can see 120 V / 12 A is 71-78% efficiency with this model.
240 V / 32 A is 90-92% efficiency.

Both of these predictions match your numbers

For 240 V:

Same thing at 120 V:

DISCLAIMER: These are models. Not measurements.

 

[ewoodrick]

The "Mobile Charger" isn't a charger, it's a connector. There's essentially no loss in it. It's really a glorified extension cord.
The charger is in the car.

 

[TheDon]

The "Mobile Charger" isn't a charger, it's a connector. There's essentially no loss in it. It's really a glorified extension cord.
The charger is in the car.

Yea - I mistyped. I think you got the idea though.
As an aside - there are losses in the "glorified extension cord" as well as in every connector. The in-car charger also has a loss as does the wiring going to/from the charger. Then there is the loss in heat from the battery itself - which is just about zero from experiential measures. There is micro heat in the battery at supercharging levels but well within operating limits.

 

[TheDon]

TeslaFi reports the efficiency of each charge.

So there must be data within the car to get their data.

 

[ZOMGVTEK]

Until someone measures the energy going into the pack while also measuring the input power at various voltages and currents, we won’t know for sure. It’s unlikely the peak efficiency is at 48A, but it very well may be around 32A. Although the wall connector at 32A will be a touch more efficient than the mobile connector at 32A. And you pay for the losses in the conductors from the meter all the way to the car, so that would need to be taken into account as well ideally. The total losses are non trivial, but it’s likely the difference between 24A and 32A is very minimal. I wouldn’t think twice about charging on a 14-50 at 32A. It’s possible backing the current down a touch may be marginally more efficient, but best case I don’t see it being more than a few pennies per charge. Worst case it’s a few pennies more than running it full speed. The difference is likely around that ballpark.

Basically, don’t charge off a regular 120V 15A outlet if you can avoid it. That’s going to be SLOW, inefficient, and put marginally more wear on the car and battery due to the massively longer duration. But 24A or 32A at ~240V is quite reasonable.

 

[darth_vad3r]

So there must be data within the car to get their data.

Yes, there is data available from the API, as well as I think they might make some other assumptions about losses.

The API tells you the for the charger: the current (no decimal points), the voltage (no decimal points), the power (no decimal points, but you can multiply the first two numbers to get a more accurate number).
It also tells you in terms of what's headed to the battery: the charge rate in miles/hr, which if multiplied by the Wh/mi constant for your car gives you power rate in W that the car is reporting to you as "delivered to the battery". You can compare this with the API's reported drive state power, which shows negative during charging (again, in kW, no decimal places, so the above method of rate in mph x Wh/mi is a better number to use).

Dividing power delilvered to battery by charger power gives you some measure of efficiency (in the car, excluding things outside the car, and possibly also excluding things inside the car too).

See a preliminary chart I posted here (Better to charger higher or run battery lower?) based on quick single sweep API data points done at an L2 charger at 30, 20, ... 10, 5 Amps. I've gathered more data since then but haven't processed the raw data into a table or chart yet. I collected 3 samples at each current setting from 30, 29, 28... 7, 6, 5A.

Until someone measures the energy going into the pack while also measuring the input power at various voltages and currents, we won’t know for sure.

See above on how you can use the API to get Tesla's measurement numbers at these two points.

It’s unlikely the peak efficiency is at 48A, but it very well may be around 32A. Although the wall connector at 32A will be a touch more efficient than the mobile connector at 32A. And you pay for the losses in the conductors from the meter all the way to the car, so that would need to be taken into account as well ideally. The total losses are non trivial, but it’s likely the difference between 24A and 32A is very minimal. I wouldn’t think twice about charging on a 14-50 at 32A. It’s possible backing the current down a touch may be marginally more efficient, but best case I don’t see it being more than a few pennies per charge. Worst case it’s a few pennies more than running it full speed. The difference is likely around that ballpark.

What makes you think 48 A is unlikely to be the peak? Do you figure the loss in the wires at that point would be incrementally worse enough vs at 32 A to overcome the otherwise increased efficiency from reducing the effects of the flat overhead losses?

I extended that table I posted above to 48 A and with that model I get 92.0-92.8% efficiency at 48A/240V, and 90.4-91.7% at 32A/240V.
This is a theoretical model in the car, so, for 48A to be worse overall, including external-to-the-car wiring, the 48A would have to cut over 1% off the efficiency vs 32A. Is that a reasonable expectation? Too high? Too low? Anyone know?

 

[darth_vad3r]

If I use I^2 R losses using ~0.03 ohms*, I get 32 W loss at 32 A, and 72 W loss at 48 A.

The extra 40 W loss is 0.35% of the power at 48 A.

*I used 8 AWG 0.6282 ohm resistance per 1000 ft at 20C and assumed ~50ft.

6 AWG would be even less resistance.

I could be way off. Someone want to check my math/method?

EDIT:
Google says the math is right, don't know if the method or input numbers are right:
(32A)^2 * ((0.6282/1000) ohm/ft) * 50ft) - Google Search = 32 watts
(48A)^2 * ((0.6282/1000) ohm/ft) * 50ft) - Google Search = 72 watts

 

[ucmndd]

The internal charger in the Model 3 uses either two or three individual 16A charging boards depending on trim level. Peak efficiency will be at multiples of 16... i.e. 16, 32, or 48 amps.

 

[darth_vad3r]

The internal charger in the Model 3 uses either two or three individual 16A charging boards depending on trim level. Peak efficiency will be at multiples of 16... i.e. 16, 32, or 48 amps.

Have we seen evidence that they are turned on in series and not in parallel? The EU 3x16 48A setup is always in parallel right, because of 3 separate phases ... I wonder why couldn’t the NA version also set the chargers up to operate in parallel to split the load?

 

[ucmndd]

Have we seen evidence that they are turned on in series and not in parallel? The EU 3x16 48A setup is always in parallel right, because of 3 separate phases ... I wonder why couldn’t the NA version also set the chargers up to operate in parallel to split the load?

Good question. I don’t actually know.

There have been some observational tests done on the Model S/X (I’m sure you could find them with a little searching). One showed that the first generation 40 or 80A dual charger setup was most efficient at either 40 or 80A.

Another test was done with the later generation charger that used either 2 or 3 24A modules and found peak efficiencies at the multiples of 24 (24, 48, and 72).

I assume the same behavior would be observed with the current 16A modules, but admit that’s speculation.

 

[davewill]

Because there's a fixed overhead while charging, I'm confident the most efficient rate is also the fastest. The longer the charge takes, the more energy the overhead eats up. All of this speculation about other rates is pretty silly.

 

[ucmndd]

Because there's a fixed overhead while charging, I'm confident the most efficient rate is also the fastest. The longer the charge takes, the more energy the overhead eats up. All of this speculation about other rates is pretty silly.

It’s not silly, it’s been measured - and at least for the multi-module later generation 48/72A Model S/X chargers, full speed is not the most efficient. The difference is frankly negligible, but it is measurable.

Model S Gen2 Charger Efficiency Testing

Ideal Charge Rate??