openinverter forum

Potentially silly question time.

My conversion is going to be fairly tight on weight so I'm keen not to lug around anything unnecessary. It occurs to me that unnecessary could include the AC charger.

If I fitted a Chademo or CCS socket I'd be able to charge while out and about but I'd then need a DC charger at home. A home Chademo charger looks relatively straight forward to do, the comms are well documented and the physical interface well defined, I already have an AC adaptor and a Volt gen2 charger, it would only be 3kW but that would do me. The trouble is Chademo plugs are extortionately expensive.

CCS plugs are available at reasonable prices, and it is more future proof, but from what I've read the comms protocol is painfully complicated. There do seem to be a few open source libraries available but I'm not clear on how much of the stack they actually implement.

So the question - can anyone adivise roughly how much work it would be to get from the open source libraries currently available to a working home CCS EVSE software implementation? Not talking about the electrical side at all here, just the software communication stack. It doesn't have to be fully featured, just enough to get say an i3 Lim to close the contactors.

Not after anything precise here, are we talking hours, days, weeks, months or years?

the openinverter tesla charger software supports chademo. ie you can use a telsa obc as a chademo charger

There is a 3d-printed CHAdeMO plug available, used it for many years. Another route is inverter charging: https://openinverter.org/wiki/Battery_Charging

Used that on Polo for many years. Besides some extra relays and some relatively small magnetics to be nice to the grid this doesn't add any weight. I didn't use the technique in Touran because it is kind of hard to make contact to the AC bus of the Leaf inverter and with magnet motors it can get a little noisy.

Finally if you use a Prius inverter you can use that for charging

Thanks for that. Seen the 3D printed plug and its probably the way I'll go if I use use Chademo.

The original plan was to use the Prius boost converter but it doesn't quite work for me. The mains voltage here can get up to 250V at times (I've occasionaly seen 255V on sunny days! - its a weak rural network with a lot of solar on it) so the charger would need to buck as well as boost.

Part of the attraction of Chademo or CCS is the fast charging while out, for a car that will have a fairly limited range it would make it a bit more flexible.

So is a software stack for CCS a bit out of reach at the moment?

You could always add a CCS or maybe the Chinese Chademo socket (whatever you can get a cheap plug for) in addition to your actual Chademo socket and run any protocol you like through it. You just need some room for the socket and contactors.

Could do but would prefer to just have the one socket.

The options seem to be Chademo with a 3D printed plug or CCS. I'm not too keen on the printed plug. The charger will be outdoors in all weathers and rain/sun plus 3D printing plus HV DC just don't sound like a good combination.

So back to the original question is a CCS EVSE software stack just a pipe dream at them moment?

I believe there are some open source stacks, though it's not really clear to me if they work yet, or what they support:

https://github.com/SwitchEV/josev
https://github.com/SwitchEV/iso15118
https://github.com/SwitchEV/RISE-V2G

https://github.com/Martin-P/OpenV2G

We've proven that we can establish the low level communication quite easily with qualcomm chips, so the next step will be trying out the above software stacks, or writing a new one.

Thanks for the links; I'd seen most of them but not all.

I've had a browse but can't get a clear picture of how well they work and whether they support the features I'd need and had been hoping someone might have a better idea of their capabilities.

No real rush as I've got plenty of other stuff to work on and am a fair way off from having to make a decision. I'm erring towards the CCS option though so I'll probably keep a watching brief until I have a bit of spare time and then have a look at the open source stacks.

Pete9008 wrote: ↑Tue Oct 25, 2022 1:24 pm I've had a browse but can't get a clear picture of how well they work and whether they support the features I'd need and had been hoping someone might have a better idea of their capabilities.

I guess you have seen it already, but if not: In https://openinverter.org/forum/viewtopi ... 2&start=50 we are discussing another open source implementation. I would it call "real-open-source", in contrast to the josev, where they have an open-source and a "pro" version.
The current state is, that it "talks" to some of the public chargers, and to some others not yet. I'm continuously improving it, and appreciate any input. If we have somebody who is able to create a "measuring device" which measures the voltage on the CCS inlet and gives it to a python program, and to create an "actor device" which controls the relay between the battery and the charger, controlled by python, and maybe get from the BMS the voltage and intended charging voltage/current/power, then, I guess it is possible to make the first real-life charging session after two or three intensive weekends

uhi22 wrote: ↑Thu Dec 01, 2022 4:25 pm I guess you have seen it already, but if not: In https://openinverter.org/forum/viewtopi ... 2&start=50 we are discussing another open source implementation. I would it call "real-open-source", in contrast to the josev, where they have an open-source and a "pro" version.
The current state is, that it "talks" to some of the public chargers, and to some others not yet. I'm continuously improving it, and appreciate any input. If we have somebody who is able to create a "measuring device" which measures the voltage on the CCS inlet and gives it to a python program, and to create an "actor device" which controls the relay between the battery and the charger, controlled by python, and maybe get from the BMS the voltage and intended charging voltage/current/power, then, I guess it is possible to make the first real-life charging session after two or three intensive weekends

Thanks for the heads up. I have been watching (and have already posted on) that thread. I actually bid on a couple of homeplug modems on ebay this morning but unfortunately didn't win the auction

Unfortunately don't have any accessible CCS hardware here but still keen to get duplicate your test setup and see what's possible.

I also have a MAX22530 eval board here (bought for something else but no longer needed) which might be suitable for the voltage sensing and would be fairly easy to drive from a Pi board?

Edit - forgot to say - some excellent work you are doing on the CCS, many thanks for sharing

The MAX22530 is impressive thing, i was not aware that such devices with built-in DC/DC exist. It should be possible to drive it from the Pi's SPI. Nevetheless, I thought into an other direction, more old-style A 5V-to-5V DCDC, an arduino pro mini which reads analog voltage via internal 10 bit ADC, and sends it via serial line. Finally an optocoupler, and a usb-to-serial adaptor. This setup fits to both, Raspberry and Windows-Notebook, which makes the work more comfortable. In the end, the singlechipsolution is more suitable for broader roll-out, this is clear.