Electric Car Experiences

[free_electron]

Would be nice to have a DC trickle-charge input with diode protection in the pack, so it could make use  of low-rate solar as available.

That would be extremely lossy. At least use "ideal" diodes using mosfets. Feeding DC is not a good idea . The on-board charger is not MPPT. it expects 1 voltage (or 2) and is tuned for that operating point.

My home is set up with a Tesla solar system and 4 powerwalls. I leave the car plugged in at home. The Solar gateway (the box that interconnects the inverters, the storage, the grid and the home) can talk to the car.
Depending on power availability (solar or storage supply) and the home power draw, it steers the car to raise or lower the charge current in such a way that it does not pull off the grid. You can set a lower limit. for example : i need the car 4 hours from now for 200 miles. If it sees it won't get there it will pull from the grid as well.

IF you combine that with a SPAN load center (intelligent fuse box) they can even do load shedding like temporarily turn off your water boiler or shift the boiler to nighttime. If there's excess power they can route that to other storage (as opposed to dumbly dumping it into the grid). A warm water boiler is also a energy storage. Heat the 80 gallon tank during excess power. that water stays hot for 24hrs in a well insulated boiler. So it's plenty for a shower in the morning. They can also start washer/dryer/dishwasher during peak power availability.

I don't have a SPAN panel but i have set my boiler on a time clock. It only gets power from 11AM to 3PM. That's peak solar production and plenty to heat the water. I have a cheap-ish (700$ costco) LG washer/dryer with wifi where i can set start times. The daily load is run after 3pm. There's still plenty of sun and that runtime is short.

[Siwastaja]

parking lot capacity; it's limited by physical charge point capacity which is limited because of the expense of very high power fast chargers

You're from Finland. Engine block heaters are everywhere in parking lots. A regular 220v socket at 20 ampere will but 32Kwh into the battery during an 8 hour working day. That is more than enough for your daily commute !

Those are wired for ~6-10A average and ~10-16A max, but they are fine for overnight charging at 8A limit, which most stock Schuko EVSEs that come with the car do. Also cheap retrofit boxes with decent Type 2 socket (and more advanced current limiting / current sharing) are available - the most expensive part of underground cabling is already done, although with varying current capacity. But I think our "engine block heater parking lots" are a great demonstration to the rest of the world how simple and cheap the concept is. We managed to put them everywhere even though we could have technically survived without. Any parking lot / shopping center / whatever owner, anywhere in the world, can take some inspiration and do the same with a tad thicker cable and modern EVSEs, which are still dirt cheap, for decent ~10kW charging infrastructure, which is not only good for overnight, but also shorter visits, shopping, etc. Of course this is cheapest when done at the same time the parking lot is first built, required underground cabling and the poles are cheap to install at that time and more expensive to retrofit.

[mikeselectricstuff]

Would be nice to have a DC trickle-charge input with diode protection in the pack, so it could make use  of low-rate solar as available.

That would be extremely lossy. At least use "ideal" diodes using mosfets. Feeding DC is not a good idea . The on-board charger is not MPPT. it expects 1 voltage (or 2) and is tuned for that operating point.

A volt or two out of 400v at a few amps - not a huge deal. Remember I'm talking trickle charging here.
The diode is a critical safety component to stop the full pack power being available externally ( connected before  contactors) , so keep it simple, maybe use two in series.

[free_electron]

EV battery packs don't like trickle charging ... Keep in mind that an EV pack is (can) being conditioned ( kept at correct temperature ) by actively heating or cooling it.  A Few amperes is not going to cut it ...
Even on 110volt @ 12 ampere they actively manage the pack.

[Ville]

I should install a logger for my ioniq 5 12v battery again.  My earlier results are from 2024.  And there's a lot of software-updates since that. (if Hyundais term "ICCU" says something to you, it might be a little hint why I logged things)

Basically idle consumption went down to ~0,4 watts after several hours (If I remember right, it took like 4 hours for the lowest level).  I have my notes somewhere that I can't find now, but I tested a lot of different situations and how long it took to calm down after.  Like walking by the car with the key, or opening just the trunk to get something etc.

So, it's Ioniq 5 2023 and the time was ~july 2024.  12v battery starts to charge from traction battery at ~12,7 volts, highest charge current while parked that I got was 34 amps or so.  It seems to try to keep the bigger battery active time as short as possible.  That can't be good for the lead battery.  I got the car in october, so I had then had the battery for 8 months, so it and the car was like ~10 months old.  While parked for several days, it charged the 12v battery about once a day.  But like I said, it seemed to trigger it by the voltage reading and the age and all the torture of the battery would affect it a lot.  In last scheduled maintenance they swapped the 12v battery, as they tested it had gone bad enough.  I didn't yet had any problems with it, or never actually have.

I'll need to find my mooshimeter and hope it is still alive.  It does the logging just fine, if I babysit it daily to check it hasn't crashed.  Having two channels is nice for that 12v battery, as the low voltage input has good enough resolution to measure current by using battery ground cable as a shunt.

It starts to get kinky if we need to find out what happens inside the big battery while it's disconnected.  There should not be anything else than the bms and it should be sleeping when theres nothing happening.  Battery percentage before and after parking wont be reliable enough to tell anything, as environment temperature etc might bump it for couple % anyway.

[Siwastaja]

EV battery packs don't like trickle charging ... Keep in mind that an EV pack is (can) being conditioned ( kept at correct temperature ) by actively heating or cooling it.  A Few amperes is not going to cut it ...
Even on 110volt @ 12 ampere they actively manage the pack.

The only reason why they don't "like" is an artificial decision by the manufacturers: just unwillingness to support something they think no one wants. It's not even hard to engineer correctly, just that manufacturers avoid any such "extra" work when it's not mandated. If EU mandated it, engineers would solve it in a day. Just turn off every unnecessary thing and you are done. Charging very slowly does not require any magic compared to being in "off" state - li-ion is fine slow-charging at any temperature, also slow charge doesn't generate any meaningful amount of heat, so neither heating nor cooling is needed. Decent BMS and other control computer stuff can easily run with less than 1W of power if designed as such. A wide-range high-efficiency 1-10kW AC charger is the hardest part, but ironically that's probably easiest for the manufacturers as it's all solved problem and they can outsource the genius-level power electronics stuff.

It's purely an attitude problem with manufacturers, nothing else.

We should start mandating cars that are capable of true V2G and that means low-quiescent mode where the battery can be used in whatever usage patterns. The Japanese figured this out in 2010 or something but it seems to be a massive blind spot still in Europe. Even given all the war and energy crisis speak. And this is truly a low-hanging fruits, just stop making artificially shitty cars for no reason whatsoever and let the free market operate and come up with all clever ways to utilize that resource.

[woody]

We should start mandating cars that are capable of true V2G and that means low-quiescent mode where the battery can be used in whatever usage patterns. The Japanese figured this out in 2010 or something but it seems to be a massive blind spot still in Europe. Even given all the war and energy crisis speak. And this is truly a low-hanging fruits, just stop making artificially shitty cars for no reason whatsoever and let the free market operate and come up with all clever ways to utilize that resource.

Biggest problem is that European car manufacturers never really wanted EV as that was incompatible with their vested interests. They write the law, so Europe did not want EV. And now they are years behind the competition and sentenced to fighting the rear guard ICE fight while the rest of the world buys Chinese EV.

[free_electron]

The only reason why they don't "like" is an artificial decision by the manufacturers

i can't speak for all as i only know one system. It has nothing to do with "like" but everything with practicality.

If you take an input of 110 volt at 12 ampere that is "trickle charging". You get barely 4 miles per hour of charging. It is completely impractical. The on-board charger can do it (the on-board charger eats anything from 110v single phase to 240volt 2 or 3 phase) and i have used it once cause i was overnight somewhere remote and they did not have a 240 volt outlet. Plug it in around 6pm and by 9am i had about 50 miles of range extra. I didn't need it cause i had a supercharger nearby, but it was an interesting experiment.

Going even lower in current is just not practical. If you drop to 6 amp you barely get 25 miles, for 15 hours of charging.... if you charge that during business hours (at work) you get like 10 miles for an entire workday. you can't even get home.

So yes, you can trickle charge. Is it useful ? only in absolute emergency or if you have lots of time.

Daily charging for me is 240 volt at 24 ampere. That gives me about 18 miles per hour. Plug the car in at 8pm and by 8 am you have added 200 miles to the battery. That's 2/3. Since you had some residual the night before it is now essentially full. Daily charging should not exceed 80% of capacity to extend pack longevity.

Since i work from home the car is always hooked to the charger and the energy Gateway talks to the car to modulate its charge current based on solar availability and power draw from the home. Priority is home -> powerwalls (for nighttime use) -> car -> grid. If there is excess powerwall energy during night ( low power draw from the house if heating/cooling is not needed) i can dump the powerwalls into the car if needed.

[nigelwright7557]

I struggle a bit with automatic transmissions so I need a manual gear change/clutch  car.
I can drive an automatic but when it comes to an emergency stop I still hit the none existent clutch.
I found when driving an automatic to keep left foot back against the seat base and drive with the right foot.

[Siwastaja]

If you take an input of 110 volt at 12 ampere that is "trickle charging". You get barely 4 miles per hour of charging. It is completely impractical.

It is completely practical if the car is plugged in "all the time" anyway, and it participates in behind-the-meter peak shaving / self-use / as excess PV sink, then it's entirely possible it's charging at 1kW for prolonged times, or charging at 500W, or, given bidirectionality, discharges at 345W, or whatever. It's useful energy being put in the pack, even if it isn't enough alone for all charging needs.

Of course 1.3kW is too low for bulk charging, even overnight, for those who drive a lot. I know people who happily charge at 1.8kW for 99% of their needs, though, so it's close. And it's exactly why losses matter. The fact is that some people do charge very slowly and find that viable. Now for them it makes a huge difference if the waste is 1W, 100W or maybe 400W. At 10kW it doesn't seem as important, until you realize that it's a fundamentally weird idea to "charge as fast as possible", often there is no need. Li-ion lifetime and efficiency gets better the lower the charge power is, there is no lower limit. Using as low power as possible to achieve the target would make sense, if car manufacturers didn't play shenanigans (weird timeout modes in DC charging, excessive quiescent consumption).

[Ville]

If you take an input of 110 volt at 12 ampere that is "trickle charging". You get barely 4 miles per hour of charging. It is completely impractical. The on-board charger can do it (the on-board charger eats anything from 110v single phase to 240volt 2 or 3 phase) and i have used it once cause i was overnight somewhere remote and they did not have a 240 volt outlet.

The problem is, that you need to pause the charging to change between 3 and 1-phase charging.  That might bite you some day, when the battery healt report shows that you charge your car in some way that the latest news tell is bad.  I'm planning to keep my car for a long time, the one before my ioniq was 19,5 years old (and I bought it brand new) when I basically gave it to my dad.  Someone might be way more interested in the resale value of their car than me.

But the reasoning why I would like to charge slower, is that the energy prices will go up and having solar at home is really easy way to get around the problem.  Finland is 3-phase country.  Really.  I'd say 99% of households have 3-phase in actual use in their houses/apartments etc for stoves in kitchen and in the sauna.  So of course we also charge our cars with 3-phase, mostly 3x16A at 230 volts, aka 11 kW.  Standard for charging says it has to be at least 6 amps, no matter the voltage or the amount of phases, so our lower limit is 4 kW.  My nominal 6,6 kWp solar makes just under 50 kWh in a good day.  I just checked logs for the last ~perfect solar day and using the 4 kW as the treshold, I could charge ~purely on solar for 7 hours and get ~35 kWh of free energy for my car.  If I could lower the charging rate and change the treshold to say 1 kW, I could have 4 hours more free charging time at 2,5 kW average power and get 10 kWh / 50km/h more out of it.

For a "perfect solar day" that 35 -> 45 kWh isn't really that important for most of the time and changing between 1 and 3 phase charging would be needed just twice.  But there's a lot of half cloudy days when the energy production fluctuates a lot and optimizing the energy harvesting for the car would need a lot of changes from 1-phase max 3,7 kW to 3-phase min 4,1 kW.  If the bms counts every pause as a new charging occurence, there would be a lot of those on the battery healt report.  Oh, and the lowest charging rate for us would be 6A x 230V = 1380 watts.  Why would it need to be possible to charge at half of that in north america, if that is so bad for the battery/charger/etc?  Why isn't the standard tied to power instead of current, which now makes the minimum to be anything from ~700 to ~4100 watts depending where you are.

[woody]

I have the same problem. Some days are completely sunny and others, like today, switch between lots of sun and lots of rain. The solution I use:

The car can charge 5A over 3 phases as a minimum and that is what I set. Overcast days I do not charge. Sunny or partly sunny days I charge during the day. If there is lots of sun the 30 panels (2012) generate enough for the car. Which means I use all of my solar power. If clouds pass a bit of grid power is used. IMO that is what the grid is for.

My biggest challenge is that the car battery is often fully charged (80%) which means the solar power has nowhere to go.

[Siwastaja]

If you take an input of 110 volt at 12 ampere that is "trickle charging". You get barely 4 miles per hour of charging. It is completely impractical. The on-board charger can do it (the on-board charger eats anything from 110v single phase to 240volt 2 or 3 phase) and i have used it once cause i was overnight somewhere remote and they did not have a 240 volt outlet.

The problem is, that you need to pause the charging to change between 3 and 1-phase charging.  That might bite you some day, when the battery healt report shows that you charge your car in some way that the latest news tell is bad.  I'm planning to keep my car for a long time, the one before my ioniq was 19,5 years old (and I bought it brand new) when I basically gave it to my dad.  Someone might be way more interested in the resale value of their car than me.

But the reasoning why I would like to charge slower, is that the energy prices will go up and having solar at home is really easy way to get around the problem.  Finland is 3-phase country.  Really.  I'd say 99% of households have 3-phase in actual use in their houses/apartments etc for stoves in kitchen and in the sauna.  So of course we also charge our cars with 3-phase, mostly 3x16A at 230 volts, aka 11 kW.  Standard for charging says it has to be at least 6 amps, no matter the voltage or the amount of phases, so our lower limit is 4 kW.  My nominal 6,6 kWp solar makes just under 50 kWh in a good day.  I just checked logs for the last ~perfect solar day and using the 4 kW as the treshold, I could charge ~purely on solar for 7 hours and get ~35 kWh of free energy for my car.  If I could lower the charging rate and change the treshold to say 1 kW, I could have 4 hours more free charging time at 2,5 kW average power and get 10 kWh / 50km/h more out of it.

For a "perfect solar day" that 35 -> 45 kWh isn't really that important for most of the time and changing between 1 and 3 phase charging would be needed just twice.  But there's a lot of half cloudy days when the energy production fluctuates a lot and optimizing the energy harvesting for the car would need a lot of changes from 1-phase max 3,7 kW to 3-phase min 4,1 kW.  If the bms counts every pause as a new charging occurence, there would be a lot of those on the battery healt report.  Oh, and the lowest charging rate for us would be 6A x 230V = 1380 watts.  Why would it need to be possible to charge at half of that in north america, if that is so bad for the battery/charger/etc?  Why isn't the standard tied to power instead of current, which now makes the minimum to be anything from ~700 to ~4100 watts depending where you are.

Indeed the obvious longer-term direction should be a direct DC connection to a hybrid inverter, basically, the EV battery "replacing" the stationary battery, or, used alongside with it with a hybrid inverter with two-battery support. Like, a 15kW hybrid inverter can easily run at +/-15kW with minimum power around only 100W or so. It can keep your house meter at zero, charging and discharging based on whether you have excess PV or more loads than PV.

But for this to become reality, car manufacturers need to get their bidirectional DC CCS ports actually working, and then hybrid inverter manufacturers need to come up with decent products (PV+battery+CCS DC) - only a few exist as we speak, and they don't work because cars don't work. Then we could dream about all sorts of fancy things like a phone app where you tell that you need the battery at minimum of 70% at 7am and it would actually optimize charging based on own solar production (and forecasts thereof), electricity markets (variable pricing), reserve markets and whatever.

This seems to be quite difficult to car manufacturers. Even just standardizing the German bidirectional DC port has taken 15 years, when it should have been a 15-day job, so I guess type2 unidirectional AC coupling, with the PWM current limit signal, is going to be it for the unforeseeable future. And the Chinese seemingly cannot make a reliable hybrid inverter + stationary home battery happen either, so

[free_electron]

looks like you have the wrong setup. nothing works nicely together.

Tesla car, powerwall, gateway, inverters and panels.

The app knows i have a car and storage and solar. The gateway regulates powerflow dynamically based on power sources (solar / storage / grid) and requirement from home and car.
I can set rules like :
- if there is excess solar : dump it in the car at a rate that does not impact home power consumption.
- i need to go tomorrow at 8am for 200 miles , make sure the car has enough, no matter what.

There is the root problem : nothing works together. It's all bits and pieces from various sources , including the bits and pieces that make the car. This part is valeo, that part wabco, that part continental, that part clarion or blaupunkt , another part delco and so on. they buy bits and pieces and they may specify certain things but it is not a cohesive system.

[Ville]

I told earlier that i'm going to log my ioniq 5 (2023) 12v battery.  I have data for ~13 days now, with voltage and current in 1 minute intervals.  No averages etc, just a sample every minute.

I had the opportunity to let the car stay ~untouched for 4 days because of the holidays and saw a strange thing happen 68 hours after parking it.  Theres ~2 watt blips about every 40 minutes (probably something else that just wont register with the 1 sample per minute log, but anyway) that just ended at 68 hours.  Base load is about 0,2 W.  Which seems pretty low, when someone mentioned 300 watt idle loads for some evs, which just can't be true, as that would mean that the 12V 60Ah battery would die in under 3 hours.. Did someone miss the milli-part of the idle watts or something?

I tought I had the logging setup last time using the negative battery lead as a shunt, but it was the positive one that was way easier to connect to.  I tried to calibrate the shunt voltages against clamp meter with as high stable current consumption I could and I think I got pretty close with several values between 5 and 9 amps.  Highest charging was 39 amps and highest consumption 26 amps.  Battery voltage was 11,94 - 14,90 volts.  Last time (summer 24) I saw pretty regular threshold for charging at 12,7 volts, but this time it seems more random.  Probably better for the battery?  Median voltage was 12,81v.

Charging the actual drive battery does not really affect the 12v charging cycles, if something, the still somewhat random charging isn't so brutal and I saw it peak at just ~10 amps while connected to grid.  >20 amps happens often when the car is just parked unplugged.  I would have thought that the 12v charging would be forced or at least the threshold to be bumped higher when theres external power available, but Koreans had thought otherwise.

When parked, there's about 2-50 watt consumption after leaving the car for ~10 minutes, where the higher values are for things just shutting down and me using the power trunk lid etc. 10-60 minutes stays at 0,5 watts + some 2 watt blibs. Hours 1-68 it stays at just 0,2 watts with those 2 watt blips every 40 minutes and after that the blips disappear.

Earlier logs (summer 24) with different sw/fw on the car it woke up for 4 hours pretty much every time when it saw the keys even near it.  Then there were steps for 20-50 watts for 2-3 minutes, 2-5W for 3 minutes, and couple steps through the 4 hours to finally drop to about 0,4 watts.  Things have clearly improved since that.

[Siwastaja]

Which seems pretty low, when someone mentioned 300 watt idle loads for some evs, which just can't be true, as that would mean that the 12V 60Ah battery would die in under 3 hours.. Did someone miss the milli-part of the idle watts or something?

You missed that it's not from 12V lead acid, it's architecture where HV battery is kept on permanently (or goes in sleep state after days of inactivity, so always-on for those who drive (almost) daily), DC/DC feeding into the 12V. Tesla introduced this whole concept already with the first Model S with its notorious, well-documented initial 150W-200W always-on idle draw.

Many others (probably your Hyundai) do the right thing and turn off the main DC contactor when the car is turned off. This decision forces the manufacturer to optimize the idle consumption because otherwise it kills the 12V battery, especially if it's lead acid, it would destroy by sulfation in no time. But for Silicon Valley stuff where they slap together random Linux computers and can't find a software professional capable of doing that properly, this decision means minutes of boot time of the systems that are so important for their customer segments. So they choose between two unacceptable things, and hope to fix the car later. As I found out with some Googling, the Polestar was already significantly improved to ~50W levels from the 300W number I worked with. Still of course significant, and requires the HV contactor to be always on.

All we need is EU or China to forbid keeping the HV contactor on unnecessarily, for safety reasons, along with the deathtrap door handles, and manufacturers will be forced to fix their shit.

[uer166]

Which seems pretty low, when someone mentioned 300 watt idle loads for some evs, which just can't be true, as that would mean that the 12V 60Ah battery would die in under 3 hours.. Did someone miss the milli-part of the idle watts or something?

You missed that it's not from 12V lead acid, it's architecture where HV battery is kept on permanently (or goes in sleep state after days of inactivity, so always-on for those who drive (almost) daily), DC/DC feeding into the 12V. Tesla introduced this whole concept already with the first Model S with its notorious, well-documented initial 150W-200W always-on idle draw.

Many others (probably your Hyundai) do the right thing and turn off the main DC contactor when the car is turned off. This decision forces the manufacturer to optimize the idle consumption because otherwise it kills the 12V battery, especially if it's lead acid, it would destroy by sulfation in no time. But for Silicon Valley stuff where they slap together random Linux computers and can't find a software professional capable of doing that properly, this decision means minutes of boot time of the systems that are so important for their customer segments. So they choose between two unacceptable things, and hope to fix the car later. As I found out with some Googling, the Polestar was already significantly improved to ~50W levels from the 300W number I worked with. Still of course significant, and requires the HV contactor to be always on.

All we need is EU or China to forbid keeping the HV contactor on unnecessarily, for safety reasons, along with the deathtrap door handles, and manufacturers will be forced to fix their shit.

Teslas that aren't 7+ years old sleep pretty well: idle/vampire draw as measured is in the order of 20mA*12V going many years back. Even newer ones have a special isolated sleep DCDC that supplies that without having to charge the 12V lead acid/16V li-ion auxiliary pack. Likely they do much better than other EVs and can sit asleep for many many months.

However for the arguments about using the onboard AC charger and minimizing idle draw *during* charge this doesn't apply, I'm not sure if that was ever a priority but a lot of things have changed since I've had first-hand context. It would indeed be nice if you didn't waste 200W or whatever while AC charging, someone should actually go out and measure specific numbers.

Direct diode connection to the pack from some external source is craziness: there's no regulatory pathways to enable anything like that unless it's reinforced-isolated from the pack, which would require a separate solar-input DCDC. Would be a cool yet expensive feature!

[uer166]

Also I re-read your post: Teslas, even older ones, *never* had always -on contactors: they would almost immediately open them and run all auxiliaries from 12V, and then they would turn them on to recharge the lead-acid occasionally via DCDC.

This "short--cycling" is what caused the battery wear and is the thing that was improved massively in any of the cars newer than about 7 years.