Maximum achievable charging power with an inverter

[pablo_mtl]

Hi. I'm having some questions about using the inverter for charing.

After searching the forum for my questions, I decided to share my interrogations about the charging mode of the open inverter.

(First, let's say I'm talking about "boost" mode, since most rectified domestic AC voltage will end up below most modern commercial EV pack voltage of 350-420VDC)

If one brave fellow had a good inverter hardware (think about something in the range of power of most modern EVs, 80kW to 160kW peak), with a working V2 or V3 control board, coupled with a correctly setup motor of the same power, appropriate cooling, cabling size, etc... A viable configuration.

Q: What would be the theoritical limiting factor of maximum charging power in boost mode with such a setup?

Let's not discuss about if it's right or wrong to do that and create a large non-isolated charger, nor discuss if there is sufficient power available on a single phase of 230-240VAC at home to feed this theoritical charging setup. I know that it is safer, perfectly-isolated, less noisy, more EMI/EMC friendly and more proven design to simply use an OEM or an aftermarket onboard chargers; my question is simply about the science of it all on the open inverter hardware: using the inverter+motor combo as a powerful charger.

For conversation sake, let's suppose :
- there is an infinity of AC power available on that 230VAC phase,
- there is a perfectly sized isolation transformer present in the room to keep it all nice and isolated,
- there is also a GFCI in this circuit,
- there is an isolation monitoring device in the vehicle ensuring DC-link is 1-2megaohm from chassis/earth,
- there is plenty of cooling for the inverter&motor
- this is all happening in a faraday-garage


I know in boost mode, only one of the three IGBT pairs is used, so the limit would probably come from there, but the official wiki page explaining charging mode is stating "At least for lower power charging (say below 5kW)". Where is that limit coming from? What prevents a good motor-inverter combo (capable of delivering a 50-60kW for many minutes) from delivering 15-20kW of charging power (one third of its limit)?

Thanks for enlightening me.

[johu]

It is probably more capable than I stated, I have used the setup myself at 4 kW, until the fuse tripped, that's where 5 kW comes from.
Not sure what is the lower limit, motor winding or inverter

[arber333]

Do you think a jump from 5kW to 20kW is a high power charging?

According to my own experience with my 3ph DIY EMW charger derivative i came to some conclusion.
Much depend on good inductor. Few years back I used my own wound toroidal inductor and i couldnt get the system to work really well above 10kW in buck mode from 600Vdc to 360Vdc. Then last year i got Prius Gen2 inductor and i just put it in my setup. Suddenly i could reach 18kW with air cooled charger for sustained ammount of time!

There is also unwritten rule about IGBT rating. Whenever i used IGBTs rated less than 2x of the charger amp requirement they failed in a relatively short time. So i used more than 2x amp requirement. Like if charger required 80A max i would use 200A IGBTs at least if not 300A.

Remember EMW charger developed from buck setup very similar to Openinverter buck charger version. But later PID current regulation vas added as well as some form of CP protocol with BMS loop.

Since you talk about GFCI i wanted to comment on that as well. When i kept my charger floating from car chassis by 4 rubber mounts i could charge from any EVSE just fine. Granted i didnt connect EVSE GND to Car GND obviously and i used laptop 12V PSU for charger 12V supply...

[pablo_mtl]

Thanks for these inputs.

While 5kW is close to what most normal EV have onboard (6.6ish kW), I would like to be able to achieve 20kW charging (similar to what the infamous "T-brand" vehicles have onboard).

For the sake of experimenting, I might go the "poor man's Chademo" way and build an off-board charger. But since I do not have random parts to make a frankenstein setup like you guys (I unfortunately never had time to jump in the EMW-charger adventure back in these days), I have some whole inverters here that could be entirely repurposed with for that charging job. Hence, this thread about the theorical limit of using inverters in boost mode.

It also means that I could use something better than motor windings to act as inductors... like, ...well, real inductors!

..now I would need to go inductors shopping.

[celeron55]

Well, I've been using the prius gen3 inverter as a 20kW charger for about a year now, so that certainly is achievable. It doesn't get a lot of use because at home I use my separate single phase charger, but it's useful on longer trips.

[rstevens81]

The best place for inductors is probably going to be Yaris/auris inverters as these are dirt cheap plus you can salvage what you want from the rest of the inverter.

I keep going round in my head if want to use a Prius inverter in single phase application as used OEM real chargers are so cheap.

[pablo_mtl]

celeron55, thanks for this feedback. I've read your interesting thread about using the Prius Gen3 inverter directly with 3-phase power. Nice work! If I wasn't living in a 1-phase world, I would definitely go that route to build a dedicated charger.

rstevens81, well, OEM chargers are "so cheap" when compared to that old time when the only viable option for our type of projects was a shiny new $2500 3.3kW Brusa NLG513... But in 2020 I'm still having trouble figuring a good OEM charger solution for 20kW-ish under $100/kW-ish.

ex:
- a working Gen2 Tesla 10kW charger + control hardware + shipping = 1k$,
- a Chevy Volt 3.3kW OEM Lear charger = 350$,
- a LEAF/Kia Soul 6.6kW OEM charger = 700$,
- etc.

[arber333]

celeron55, thanks for this feedback. I've read your interesting thread about using the Prius Gen3 inverter directly with 3-phase power. Nice work! If I wasn't living in a 1-phase world, I would definitely go that route to build a dedicated charger.

rstevens81, well, OEM chargers are "so cheap" when compared to that old time when the only viable option for our type of projects was a shiny new $2500 3.3kW Brusa NLG513... But in 2020 I'm still having trouble figuring a good OEM charger solution for 20kW-ish under $100/kW-ish.

ex:
- a working Gen2 Tesla 10kW charger + control hardware + shipping = 1k$,
- a Chevy Volt 3.3kW OEM Lear charger = 350$,
- a LEAF/Kia Soul 6.6kW OEM charger = 700$,
- etc.

Maybe 2x tesla chargers on 3phase?
I use 3x 3.3kW Outlander chargers distributed on 3 phases. Hidden benefit if this is i have 2 spare DCDC converters .

For 20kW charging for long trips i will adapt my EMW charger and try to connect it directly to battery via 175 Andersen connector. Now i used 50A connector but i was using less than 50A. Maybe it would be worth it to adapt DIY charger to chademo protocol...

[New Electric Ireland]

ex:
- a working Gen2 Tesla 10kW charger + control hardware + shipping = 1k$,
- a Chevy Volt 3.3kW OEM Lear charger = 350$,
- a LEAF/Kia Soul 6.6kW OEM charger = 700$,
- etc.

If you hunt around 20kW for 500 USD with two Tesla PCS should be achievable;

viewtopic.php?f=9&t=1281&p=21074#p21074

[pablo_mtl]

Maybe 2x tesla chargers on 3phase?

Unfortunately, 3-phase isn't a road-side option in my country. (On the upside, there is an infinity of 30A@240V-single phase EVSE everywhere). But at Home, two 10kW Tesla OBC/PCS would work.

If you hunt around 20kW for 500 USD with two Tesla PCS should be achievable;

That would work. I already started investigating the Tesla Gen2 OBC path, but if would be a matter of finding the cheapest/easiest option between 2X Gen2 OBC vs 2X M3 PCS. Even Tesla Gen3 OBC would be an option if they were not sold at such a high price.

[arber333]

Unfortunately, 3-phase isn't a road-side option in my country. (On the upside, there is an infinity of 30A@240V-single phase EVSE everywhere). But at Home, two 10kW Tesla OBC/PCS would work.

Well than your best option would be to get US version of Tesla charger and use one of Damiens boards to hijack it.
viewtopic.php?f=10&t=78
https://www.ebay.com/itm/TESLA-MODEL-S- ... 2953408407

[pablo_mtl]

Well, yes, that was my original conclusion before I considered using a $100 Prius inverter dedicated for off-board charging. Hence, this current interesting thread about >5kW charging mode with an inverter. (cheaper and more interesting than purchasing $1k of hardware).

[Jack Bauer]

If it were me and cost per amp were the driver I'd go this route : viewtopic.php?f=14&t=825

[pablo_mtl]

Thanks Damien! after reading that interesting development thread link you posted, I see that for charging mode, the Yaris/Auris Gen 3 inverter seems like a strong contender: both cheap and an all-in-one solutions by using internal buck-boost DC-DC. (having a few non-Toyota inverters here, I was focusing on trying the "classic" charging mode of the openinverter, using motor phases, but I think the Toyota hardware is a way more elegant setup for what I want to achieve).

Your video of bench-cooking one of them is simply another proof that these are very capable units!

I'll find one locally and try to build an off-board-charger with it. Although we don't have Yaris/Auris cars here, I think the "Prius C" inverter is the same hardware. And just like the Yaris, it's available for cheap because it is of no use for the taxi-car repair industry.

Cheers!

[Romale]

I think it would be appropriate to ask a question here instead of creating a new topic: what is the minimum input voltage the charger should have or maybe there is a rule not lower than X percent of the battery voltage?

[Romale]

I just tried to connect the em57 kit to a 130 volt battery and start the charging boost mode from a small 12 volt accumulator to check and configure, the inverter wrote in the web interface it was in boost mode, the contactors closed immediately as soon as I turned on the ignition, but the charging current was 0 amperes in fact (foc firmware 4.97.r)

[johu]

I think it would be appropriate to ask a question here instead of creating a new topic: what is the minimum input voltage the charger should have or maybe there is a rule not lower than X percent of the battery voltage?

I have charged the 360V battery from 60V power supply. Efficiency won't be great, in practise I wouldn't go much above 1:2

[Romale]

thank you ))

[Romale]

is there any special order for starting the boost charging mode?
I used a 60 volt power supply.
I connected his + to the middle phase of the motor, and his - to the negative terminal of the battery. I closed +12 to the outputs forward and backward, turned on the 60 volt power supply, turned on the inverter power. controller status: boost. however, no charging current has occurred. I can't figure out where I'm wrong. I had high hopes for this system and such a crash (((

[johu]

I connected his + to the middle phase of the motor

I think there's your problem. L2 is the module being clocked for boosting, so to have an inductor your input voltage can be on any phase BUT L2. Right now you're just shorting out your power supply

[Romale]

it's very strange. I can't figure out exactly where the error is and why should I have a short circuit for the power supply?
this did not happen, there was simply no load in the form of a charging current.
in fact, I did everything exactly as indicated in the wiki for using an inverter with a motor to charge the battery, just simplified the number of contactors and used a power supply instead of a diode bridge.
maybe there is an error in the wiki and it is necessary to supply + to the battery, and - to one of the phases?
below is a picture of how I connected and a screenshot of how to connect, I don't see the difference.

[johu]

Your connection is correct and in spot values you can see the IGBT is clocked with 715 digits, I think that is just around 20% duty cycle. Much too low...
I haven't tested charge mode in a while, maybe try older software like 4.97.R?

[Romale]

maybe try older software like 4.97.R?

I started just with the old firmware and it didn't work on it. later I tried foc.5.17.r and now 5.15.r is installed.. I also tried to raise the voltage of the power supply to 80 volts (for a 130 volt battery this should be enough) there is probably some kind of settings field in an unknown place that makes its own adjustment and does not allow charging to work

[johu]

Ok, then it's not a bug. 4.97 charged well in good old Polo
Can you dump http://192.168.4.1/cmd?cmd=json while trying to charge?

[remy_martian]

it's very strange. I can't figure out exactly where the error is and why should I have a short circuit for the power supply?
this did not happen, there was simply no load in the form of a charging current.
in fact, I did everything exactly as indicated in the wiki for using an inverter with a motor to charge the battery, just simplified the number of contactors and used a power supply instead of a diode bridge.
maybe there is an error in the wiki and it is necessary to supply + to the battery, and - to one of the phases?
below is a picture of how I connected and a screenshot of how to connect, I don't see the difference.

IMG_20220603_000319_227.jpg

Screenshot_20220529-174259_1.png

Might be a silly question, but you do have the inverter positive connected to the HV Battery positive through a closed contactor, correct?

If that answer is yes: What happens when you move the "charger out" positive wire from the blue to the green or red motor phase?