Open Source CHAdeMO charger 100-150kW

Toyota Rav4 EV Forum

Help Support Toyota Rav4 EV Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

miimura

Well-known member
Joined
Jan 17, 2013
Messages
1,939
Location
Los Altos, CA
I think turning this thread into a discussion about an Open Source DC Fast Charger would be useful and interesting.

Tony's proposal for a higher power unit that has two handles and can charge two vehicles with variable power is very interesting. This is the way that Tesla SuperChargers work. Tesla gives the first vehicle to arrive as much power as it asks for and the second gets whatever is left over, up to its requested amount. The SuperChargers use a stack of 10kW AC to DC chargers, the same as the ones used in their cars, to leverage production volumes and provide some fault tolerance in case one fails. This is a good architecture.

For any new effort, I think the easiest path to implementing a fault tolerant system like this is to use some existing hardware and integrate it together. eMotorWerks has some DC charger hardware and could be ganged up to build this kind of system. At 25kW each, there is not as much granularity as the Tesla 10kW chargers, but they are likely cheaper and easier to inetgrate. Conceptually, the easiest way to handle the dual output for two cars is to have each charger output go into an A/B switch that can direct all of that charger's output to one of two current delivery buses, one going to each of the two CHAdeMO handles. Each charger in the system is working at the battery voltage of the connected vehicle and the current from each one is additive. As the current ramps down on one vehicle and one charger's current is no longer needed it can be switched over to the other bus and ramp up the current to that vehicle as needed.

From a system integration standpoint, cooling is probably one of the larger concerns. SuperChargers are liquid cooled, mostly because that is how the individual charger modules are cooled.

Thoughts?
 
My latest concept is a regional "JdeMO" network throughout the western US similar to Superchargers:

a. 100-150kW, dual head, just like Superchargers
b. Prepaid for by users
c. custom plug that is not compatible with non-paid users
d. fully compatible with CHAdeMO capable cars
e. "Dumb" stations... it starts and stops on site... no communications that interfere or fail to charge your car
f. between major metro area, LA to SF, SF to Sacto, Sacto to Reno, San Diego to LA, LA to Las Vegas, Seattle to Portland, Sacto to Oregon

Let's do one of the above as "Open Source"
 
miimura said:
Tony's proposal for a higher power unit that has two handles and can charge two vehicles with variable power is very interesting. This is the way that Tesla SuperChargers work. Tesla gives the first vehicle to arrive as much power as it asks for and the second gets whatever is left over, up to its requested amount. The SuperChargers use a stack of 10kW AC to DC chargers, the same as the ones used in their cars, to leverage production volumes and provide some fault tolerance in case one fails. This is a good architecture....

From a system integration standpoint, cooling is probably one of the larger concerns. SuperChargers are liquid cooled, mostly because that is how the individual charger modules are cooled.

First, I want to say upfront that I would not be interested in anything similar to the other home-brew efforts. No 3D printed parts (fine for prototyping, however), no substitutions of logical parts with substandard ones to "save a buck", etc. I personally don't think or operate that way.

First, virtually every one of the homebrew DC chargers that I'm aware of are non-isolated. That's a non-starter. Air cooled is a non starter (for me).

The overall goal needs to be a charger as robust and trouble free as the Tesla Supercharger. Not quirky and "special".

So, my proposals are:

1) isolated charger(s), no exception

2) hack the existing Tesla 10kW charger and use that as the design for Open Source DC charging.

3) hack any of the other designs, and modify them for liquid cooling

4) the chargers MUST be CAN bus controlled, IMHO

5) we need a really good plug design, which is as big a project as the charger

6) some fancy software to communicate the CAN bus messages for CHAdeMO and the car.

7) the formation of a non-profit called "California West Coast Electric Highway"
 
For what it's worth,

1) EMW does offer a 20kW unit that is isolated (The ISOCharge-20000) http://www.emotorwerks.com/tech/electronics
2) AeroVironment has been doing power-sharing for multiple vehicle parallel charging for years at airports for lower voltage (24-96V) ground support equipment using their MVS & GSE series of chargers with 8-16 ports.
3) The ISOCharge-20000 could easily be modified for liquid cooling and is already CHAdeMO ready
4) The ISOCharge-2000 has CAN control coming soon
5) Uugh, another plug...
6) Fancy software easily done with some of the talent on this list (with enough time and money)
7) Don't forget to include a business strategy to allow these chargers to get outside the state, should it be financially possible to do so (i.e. between Austin and Dallas, between Austin and Houston)

I'm not convinced there would ever be enough vehicles in the club to justify having a charger that could power two cars at one time (>100kW). The power that is needed would limit their placement and the amount of times two vehicles would need fast charging at the exact same time would be very rare. Creating a lower power, less expensive charging station with all of the other attributes might be better.

Since our batteries are not as large as Tesla's, there is little incentive to build a charger larger than 60kW, as the charging C-rate needs to be kept within spec of the cells. So 3 of the ISOCharge-20000's would only be needed (~$15k, plus control board, outer weather box, AC fuse/junction box, output cable & plug). Also, perhaps some of Tony's testing can determine the difference in charge times between 40 and 60kWs, as Fuji showed how their 25kW charger takes only ~11 minutes longer to charge a Leaf battery than a 50kW charger.
http://www.americas.fujielectric.com/sites/default/files/DC%20Quick%20Charging%20-%20FEA%20Comparison%20Study%20%20(25kW%20vs%20%2050kW)%207-3-12.pdf
If that's the case, then one could make a 40kW charger as the best balance between cost and utility. Even going to a 20kW option (a mini-fast charger) could be plugged in almost anywhere (240V single phase) opening up a much wider deployment base.
 
What would be really nice would be something like this:
http://www.proterra.com/product-tech/charging-battery-techonology/

I know, too much wait, no clearance.
It's cool though ;-)
 
Kohler Controller said:
For what it's worth,

1) EMW does offer a 20kW unit that is isolated (The ISOCharge-20000) http://www.emotorwerks.com/tech/electronics
2) AeroVironment has been doing power-sharing for multiple vehicle parallel charging for years at airports for lower voltage (24-96V) ground support equipment using their MVS & GSE series of chargers with 8-16 ports.
3) The ISOCharge-20000 could easily be modified for liquid cooling and is already CHAdeMO ready
4) The ISOCharge-2000 has CAN control coming soon
5) Uugh, another plug...
6) Fancy software easily done with some of the talent on this list (with enough time and money)
7) Don't forget to include a business strategy to allow these chargers to get outside the state, should it be financially possible to do so (i.e. between Austin and Dallas, between Austin and Houston)

I'm not convinced there would ever be enough vehicles in the club to justify having a charger that could power two cars at one time (>100kW). The power that is needed would limit their placement and the amount of times two vehicles would need fast charging at the exact same time would be very rare. Creating a lower power, less expensive charging station with all of the other attributes might be better.

Since our batteries are not as large as Tesla's, there is little incentive to build a charger larger than 60kW, as the charging C-rate needs to be kept within spec of the cells. So 3 of the ISOCharge-20000's would only be needed (~$15k, plus control board, outer weather box, AC fuse/junction box, output cable & plug). Also, perhaps some of Tony's testing can determine the difference in charge times between 40 and 60kWs, as Fuji showed how their 25kW charger takes only ~11 minutes longer to charge a Leaf battery than a 50kW charger.
http://www.americas.fujielectric.com/sites/default/files/DC%20Quick%20Charging%20-%20FEA%20Comparison%20Study%20%20(25kW%20vs%20%2050kW)%207-3-12.pdf
If that's the case, then one could make a 40kW charger as the best balance between cost and utility. Even going to a 20kW option (a mini-fast charger) could be plugged in almost anywhere (240V single phase) opening up a much wider deployment base.
For the RAV4 EV, I don't think that 20kW-25kW is worth doing for this application. It's just not fast enough for inter-city travel. I suspect that Tony has seen the CANbus messages from the Tesla BMS in the RAV and knows that it can take significantly more than the 40kW to 50kW that today's CHAdeMO chargers put out. I think that 80kW is a good number because it would likely allow the RAV to charge to 80% in 30 minutes, a common goal of DCFC. It would also allow you to charge two Leafs simultaneously with little slowdown. However, I do agree that 100kW+ electrical supply presents its own problems. The other issue is charge port inlets and handles that are rated for over 125A. To hit 80kW we would need 200A rating on the handle and inlet. Again, Tony has had contact with Yazaki, so he would know what spec is available today.

If you believe that Nissan will have a 150 mile car soon, then this kind of high power CHAdeMO compatible system would be very appealing to those owners. However, I don't think inter-city travel will ever be appealing to Leaf drivers due to the small battery size and driving to charging time ratio.

Your 20kW "mini-fast charger" is a good idea, but it's really a solution to a different problem. ABB is already working on something like this, initially for China, and the Bosch-BMW unit is similar, but configured for Combo1 or Combo2.
 
That seems far fetched to me Tony.

I could be wrong, god knows my fiance always tells me this, but isnt our money best spent on our own cars? What I mean is, how about we improve our own vehicles and our own battery to a point that our vehicle can go as far as it possibly can on a charge. Let the corporations with the big bucks invest their money in the infrastructure, the charging stations (the high power & the low power), the land leases, the actual electric costs ect.

Lets modify our cars so we can accept the CHAdeMo fast charge, a dual j1772 and better yet, a 60 - 85kwh battery. Lets let the billion dollar companies build the electric highways. We will get a lot further pooling our money together on a Rav4 EV project that helps us need less charging stations than building an EV highway. Plus people like me in NY and others in PA and MN will also invest in the improvements.
 
TonyWilliams said:
2) hack the existing Tesla 10kW charger and use that as the design for Open Source DC charging.
How do you propose to source the Tesla chargers? For development purposes, you could get a few from wrecking yards, but that will not be practical for building out the network. If you're proposing to buy them from Tesla, why not ask to buy the whole cabinet already finished? Changing the vehicle communications and using your own vehicle connector is much easier than integrating the whole stack of chargers yourself.

TonyWilliams said:
5) we need a really good plug design, which is as big a project as the charger
Personally, I think maintaining physical and protocol compatibility with CHAdeMO is a good idea. Adding a module inside the car that will authenticate with the charger without requiring net connectivity would be a good solution. This authentication could just be a newly defined extension of the CHAdeMO protocol.
 
jimbo69ny said:
That seems far fetched to me Tony.

I could be wrong, god knows my fiance always tells me this, but isnt our money best spent on our own cars? What I mean is, how about we improve our own vehicles and our own battery to a point that our vehicle can go as far as it possibly can on a charge. Let the corporations with the big bucks invest their money in the infrastructure, the charging stations (the high power & the low power), the land leases, the actual electric costs ect.

Lets modify our cars so we can accept the CHAdeMo fast charge, a dual j1772 and better yet, a 60 - 85kwh battery. Lets let the billion dollar companies build the electric highways. We will get a lot further pooling our money together on a Rav4 EV project that helps us need less charging stations than building an EV highway. Plus people like me in NY and others in PA and MN will also invest in the improvements.
I agree that you should control your own destiny as much as you can. However, nobody that has the "big bucks" has the vision and motivation to build out the infrastructure that we need, so it's just not going to happen without a community driven approach. The Pacific Northwest High Amp L2 project is a good example.
 
The only other company I would trust besides the Tesla unit is Brusa with their 22kw unit and YES, it will work on 60hz power. And I totally agree with Tony, no home brew stuff or "always in testing" designs. While I admire Val and EMw for their work, if ANY of you have personally seen or HEARD their isolated unit working.. then you know what's going on. I would not use it.... yet.. there's alot still to be figured out. When they get it, it would be cool at home in a CHAdeMO unit.

And it's not a commercial product, where as the Tesla and Brusa is. If you're stacking these things up in a cabinet with a giant radiator, then it's what you need to do. I should know as I've gone to a LQ setup personally in my Leaf from a air cooled and it's the bees knees!

Maybe the key here is the company that is starting to supply Kia with those new fancy HO CHAdeMO units. Who knows but there's alot of talk about them and a US distributor might be a sweet deal! ;)

My personal take on this is just this (and don't get mad Tony ;) )I think we have enough standards and proprietary EV charging systems and that is what's hurting the adoption rate. Tesla is doing it right! One standard/connector and done! But its compatible with the past and future(to a point, only so much KWH can be pushed through that connector).

The J1772 is here to stay, let's not F with it. Higher amperage is fine.. 40-70a.. fine.. but make everyone have a J-plug. DONE.

CHAdeMO.. we know it can go up to Tesla SC powers.. and it continues to grow, lets adopt this and the protocol has already been hacked/figured out, if there's now a JEdeMO system out there I think it'll confuse the EV community possibly more. I know you guys have your network but I think CHAdeMO should be just the universal DCFC system for non-tesla vehicle.

CCS Combo.. Let it die already.. just make the stations dual CHAdeMO.. any vehicle with this connector is not going to make it (i3, Spark anyone??) so why waste the damn resources..

And Jimbo... if you want a bigger battery in your Rav.. you should have just bought a s60 :roll: because it's already been done but the cost doesn't outweigh the benefits. People talk the talk... but don't walk the walk.. I know this personally from the Leaf/Brusa mod.

They want it.. but won't pay for it.. :roll:
 
miimura said:
Kohler Controller said:
For what it's worth,

Creating a lower power, less expensive charging station with all of the other attributes might be better.

Since our batteries are not as large as Tesla's, there is little incentive to build a charger larger than 60kW, as the charging C-rate needs to be kept within spec of the cells. So 3 of the ISOCharge-20000's would only be needed (~$15k, plus control board, outer weather box, AC fuse/junction box, output cable & plug). Also, perhaps some of Tony's testing can determine the difference in charge times between 40 and 60kWs, as Fuji showed how their 25kW charger takes only ~11 minutes longer to charge a Leaf battery than a 50kW charger.
http://www.americas.fujielectric.com/sites/default/files/DC%20Quick%20Charging%20-%20FEA%20Comparison%20Study%20%20(25kW%20vs%20%2050kW)%207-3-12.pdf
If that's the case, then one could make a 40kW charger as the best balance between cost and utility. Even going to a 20kW option (a mini-fast charger) could be plugged in almost anywhere (240V single phase) opening up a much wider deployment base.

For the RAV4 EV, I don't think that 20kW-25kW is worth doing for this application. It's just not fast enough for inter-city travel. I suspect that Tony has seen the CANbus messages from the Tesla BMS in the RAV and knows that it can take significantly more than the 40kW to 50kW that today's CHAdeMO chargers put out. I think that 80kW is a good number because it would likely allow the RAV to charge to 80% in 30 minutes, a common goal of DCFC.

Your 20kW "mini-fast charger" is a good idea, but it's really a solution to a different problem. ABB is already working on something like this, initially for China, and the Bosch-BMW unit is similar, but configured for Combo1 or Combo2.

80kW would be too fast of a charge rate for our pack (1.6C). Even Tesla limits it to 1.4C. Using a 50kW charger would be 1C and would recharge to 80% from 30% in ~30 minutes.



I'm glad to hear about ABB's option. I just think that a 20kW, install-anywhere, low cost charger is still valuable when you consider it's 3-4X faster than public EVSEs, offering almost 1mile/minute charge rate. The price difference between it an a 50kW version when considering infrastructure and installation costs are more than the difference in output power alone (maybe 1/4th the price?). Don't get me wrong, I still want both, but use the 20kW installs at restaurants/bars and 50kW chargers on the edges of town.
 
Kohler Controller said:
80kW would be too fast of a charge rate for our pack (1.6C). Even Tesla limits it to 1.4C.
That's because most are driving 85's but what about the 60's on the new SC units?

Where are the YouTube vids of the S60 owners going on a new 135kw SC unit.. I bet they are pushing harder.
 
Kohler Controller said:
80kW would be too fast of a charge rate for our pack (1.6C). Even Tesla limits it to 1.4C. Using a 50kW charger would be 1C and would recharge to 80% from 30% in ~30 minutes.... I just think that a 20kW, install-anywhere, low cost charger is still valuable when you consider it's 3-4X faster than public EVSEs, offering almost 1mile/minute charge rate. The price difference between it an a 50kW version when considering infrastructure and installation costs are more than the difference in output power alone (maybe 1/4th the price?). Don't get me wrong, I still want both, but use the 20kW installs at restaurants/bars and 50kW chargers on the edges of town.

The charger concept that I propose is modular. It can be 10kW, or 150kW. So, that part of the equation is a non-issue.

The 60kWh Tesla has been seen at 105kW, or 1.75C

The 85kWh Tesla has been at 135kW, or 1.588C

The Nissan LEAF (no cooling system) is 2C.

The RAV4 EV won't always have the exact same 2900ma cells, so we have to build for tomorrow. I am confident that we can take 80kW today (250 amps lower pack voltages), which makes is a "125kW" charger on the currently unobtanium 500 volt battery that CHAdeMO is rated for. That's for a single car.

Which is why I keep calling the proposal 100-150kW.

After driving this little 1000+ mile weekend in the RAV4, I can tell you that 25kW CHAdeMO sucks. Great for a LEAF, not so great for a "real" car. The faster we can jam the power in the battery at the lower SOC%, the better.

For two cars that show up, there needs to be enough reserve for a single car to be at full power while the second car gets something.

Again, I foresee REGIONAL travel as the niche, and 25kW isn't even close.
 
miimura said:
For the RAV4 EV, I don't think that 20kW-25kW is worth doing for this application. It's just not fast enough for inter-city travel. I suspect that Tony has seen the CANbus messages from the Tesla BMS in the RAV and knows that it can take significantly more than the 40kW to 50kW that today's CHAdeMO chargers put out.


For the record, the maximum amperage reported by the CAN for "Sport" mode is 530 amps, and 480 amps in normal mode. There obviously is no maximum for DC charging, since it wasn't equipped that way.


To hit 80kW we would need 200A rating on the handle and inlet. Again, Tony has had contact with Yazaki, so he would know what spec is available today.


200 amps * 300 volts for a mostly depleted battery is only 60kW (1.26C). We need a specification for 1.5C at 300 volts, which equals 240 amps. In common industry rating standards, that is 240 amps * 500 volts = 120kW. I recommend that be the minimum, with 300 amps (150kW) to be the maximum.

So, the maximum charge on our Rav4 EV with the current battery of 48kWh-ish is:

1.5C max charge rate, 72kW

20% SOC (300 volts * 240 amps = 72kW)
50% SOC (340 volts * 211 amps = 72kW)
96% SOC (382 volts * 20 amps = 7.6kW)

*******************************

1.6667C max charge rate, 80kW

20% SOC (300 volts * 267 amps = 80kW)
50% SOC (340 volts * 235 amps = 80kW)
96% SOC (382 volts * 20 amps = 7.6kW)
 
jimbo69ny said:
Lets modify our cars so we can accept the CHAdeMo fast charge, a dual j1772 and better yet, a 60 - 85kwh battery. Lets let the billion dollar companies build the electric highways. We will get a lot further pooling our money together on a Rav4 EV project that helps us need less charging stations than building an EV highway. Plus people like me in NY and others in PA and MN will also invest in the improvements.

Clearly, I think we should have all the things you mention. This thread is devoted to an additional thing, not an in lieu of thing.

I wouldn't expect you to finance what undoubtedly will be a western solution, much like all the original Tesla Roadster stations were.
 
miimura said:
TonyWilliams said:
2) hack the existing Tesla 10kW charger and use that as the design for Open Source DC charging.
How do you propose to source the Tesla chargers? For development purposes, you could get a few from wrecking yards, but that will not be practical for building out the network. If you're proposing to buy them from Tesla, why not ask to buy the whole cabinet already finished? Changing the vehicle communications and using your own vehicle connector is much easier than integrating the whole stack of chargers yourself.


Yes, clearly, writing a big check for a current 135kW Supercharger and putting a CHAdeMO plug on it is the easy answer. I'm going to suggest that it's going to be cost prohibitive. One interesting strategy with this is that one plug can remain Supercharger and the other CHAdeMO.

Since the 60kWh Tesla cars may not have Supercharger access, but they do have all the hardware, that means a market exists for some very limited Tesla drivers to use our non-networked units that couldn't otherwise Supercharge.


miimura said:
TonyWilliams said:
5) we need a really good plug design, which is as big a project as the charger
Personally, I think maintaining physical and protocol compatibility with CHAdeMO is a good idea. Adding a module inside the car that will authenticate with the charger without requiring net connectivity would be a good solution. This authentication could just be a newly defined extension of the CHAdeMO protocol.


My current thinking is:

1) one way compatible plug
2) using pin 3 of the CHAdeMO plug as a authentication connection
3) using the CAN reported VIN of each car, just like Tesla
4) any 2 or all 3 of the above
 
TonyWilliams said:
200 amps * 300 volts for a mostly depleted battery is only 60kW (1.26C). We need a specification for 1.5C at 300 volts, which equals 240 amps. In common industry rating standards, that is 240 amps * 500 volts = 120kW. I recommend that be the minimum, with 300 amps (150kW) to be the maximum.
So, is there a commercial off the shelf vehicle inlet and matching cable and handle available for purchase rated at 300A?
 
miimura said:
TonyWilliams said:
200 amps * 300 volts for a mostly depleted battery is only 60kW (1.26C). We need a specification for 1.5C at 300 volts, which equals 240 amps. In common industry rating standards, that is 240 amps * 500 volts = 120kW. I recommend that be the minimum, with 300 amps (150kW) to be the maximum.
So, is there a commercial off the shelf vehicle inlet and matching cable and handle available for purchase rated at 300A?


Of course not. We have to build it.

The CHAdeMO standard is 125 amps.
 
TonyWilliams said:
miimura said:
TonyWilliams said:
200 amps * 300 volts for a mostly depleted battery is only 60kW (1.26C). We need a specification for 1.5C at 300 volts, which equals 240 amps. In common industry rating standards, that is 240 amps * 500 volts = 120kW. I recommend that be the minimum, with 300 amps (150kW) to be the maximum.
So, is there a commercial off the shelf vehicle inlet and matching cable and handle available for purchase rated at 300A?
Of course not. We have to build it.

The CHAdeMO standard is 125 amps.
Well, if you take a page from Tesla's play book, you make the high current pins longer for more contact area to carry the higher current. That's what they did for the Type-2 compatible inlet. The trick is to make the vehicle inlet compatible with the rest of the existing infrastructure. I thought there was some specification for higher power CHAdeMO but all the existing product is remaining at the lowest common denominator of 125A.
 
miimura said:
Well, if you take a page from Tesla's play book, you make the high current pins longer for more contact area to carry the higher current. That's what they did for the Type-2 compatible inlet. The trick is to make the vehicle inlet compatible with the rest of the existing infrastructure. I thought there was some specification for higher power CHAdeMO but all the existing product is remaining at the lowest common denominator of 125A.

Of course. That is why, again, that I suggest a virtual duplication of the Tesla Supercharger, that just happens to have a CHAdeMO inlet of the "JdeMO" specification.

Yes, longer pins that allow users free access, and still be reverse compatible with CHAdeMO.

Yes, that would allow 300 amps.

We need to build the 40 amp / 208 to 277 volt chargers.

We need to build the custom plug.

We need to write the software.

Everything else is just hardware to collect, like a case to put it in, or radiators and fans to cool it, or cables to transfer the power.
 
Back
Top