Page 1 of 5

Using EV Chargers

Posted: Jan 02 2019 10:53am
by methods
This will be slow to roll out... but I now am familiar with how to do this.

From Easiest to Hardest:

Basically you take any charger with a 240V input and zip-tie it to your bike. Add a diode and a couple relays and you are set. You will of course need the correct inlet connector. I can help you with this.

All that is required here is CAN and good timing. More on this later.

You need special goodies for this.... but a quick search of the web can point you in the right direction.

Same as CHAdeMO but wrapped in more riddles.


I can get all of those to turn on and provide a charge.

Most EV's are in the 400V range. The DC Fast Chargers I work with only go down to about 200V. You can actually spoof this for a 50V pack pretty easy... but it is inefficient.

EV Chargers (and I have to be careful here...) do not necessarily work like a CV/CC Power Supply.

With CCS... you can set bounds and limits similar to a CV/CC... like... Max Voltage, Max Current.
Trip limits


If you want to charge a 50V EV battery off of a 200V charger... AND I DO NOT SUGGEST YOU DO UNLESS YOU REALLY KNOW WHAT THE FRUCK YOU ARE DOING... you will need to run at less than 25% efficiency.

Lets say you want a 50A charge.
So... you request 200V @ 50A

Requesting 50V will just be rejected (most of the time)

So... You need to provide "back-pressure" that meets spec

200V / 50A = 4ohms

So... hooking up a 4ohm toaster capable of 10KW would do it
(Thats 10KW... not 1KW... pay attention!)

So in that case, if you have ~about~ a 50V pack you could calculate the following:

(200V - 50V)/50A = 3ohms

This is just ohms law and has to do with sum voltages.
If you have 50 pushing back on 200 then there is 150 to drop... and you gotta drop it somewhere...

So that would be 7.5KW of waste for a 2.5KW charge
Fine by me... Especially if it is cold out.
That is about what an average EV heater burns and if you are sitting in the car waiting for a charge... eh...

Hey - its cold out!
Get used to the idea that Electricity gets burnt for heat!

What would make more sense is to configure your pack to be 200V


Just remember...
That your ID is being logged...
So if you screw up and damage a charger...
I may end up coming to your garage to take away your tools.


If you are pretty hardcore you can contact me directly and I will help you to make good decisions around leveraging Charge Infrastructure.

If you are nOoB status... dont even think about it.
I mean it.
Fire... BIG FIRE... your stuff broken quick :cry:


Re: Using EV Chargers

Posted: Feb 02 2019 2:01pm
by methods
Now hopefully we have dispelled any myths. . . The type held and propagated by folks who have no practical sense of Power and Electricity.

... NO ...

You can not practically "grab a charge" from a DC Fast Charge Station with a couple resistors and some jumper cables. :lol:


What you can do tho. . .
And what I have demonstrated against a Zero Motorcycle. . .
Is run a proper DC-DC converter.


To succeed I suggest you do your testing up to 420VDC.

Do NOT aim at 200VDC.
Aim higher

We make Power with Voltage and anyone who says different (Eh HEM... LFP...) is not being practical in any way.


Nearly any AC-DC converter out there which is Wold Compatible will accept something like a 380V DC input. Swap a few caps and diodes and very quickly. . . you will have a DC-DC converter which takes 420V at its input and produce 50.2V at its output.

... Have I done it? ...


Since when have you seen me NOT blowing shit up thru experimental proofs??

We are rolling this out R E A L S L O W

If you are in a big rush, go someplace else.
This aint the post for you


What we want to do is make a compact DC-DC for your Ebike which will accept CCS1, CCS2, CHAdeMO, or Tesla DC Fast Charge.

... Why?

Because we can.


While you are being patient please get rigged up with some J1772
Just 3D print the inlet

Wire only the following:


Those should be sufficient to carry 50A... so... 10AWG will work.

For comms you will need



Proximity is the easy part and you will quickly figure it out.

this is a bit more tricky...

Understand that it is NOT a DC stimulus
It is AC
And that is what the Diode is all about

You will need a DPDT relay, a few resistors, a mosfet, and an Arduino.
I now use PI... but an Arduino will suffice.


J1772 is everywhere
Roll by Google and check it out

Go download a 3D file and get to printing

Print one for me and I will do an article on setting it up for Arduino.


WHY do we need an Arduino?
Because I say so, thats why.

Anybody can flip switches to initiate a charge session.

What we are doing here is setting you up for DC Fast Charge. ... . which ... will take A LOT MORE... and this wont be any sort of plug-n-chug activity.


We are talking about sniffing packets, determining protocol, meeting electrical requirements, meeting timing requirements, properly interlocking, HiPot testing, and ...

Getting it to ACTUALLY WORK
Without undo risk or bullshit


Off to go do manual labor like Tesla did. . . when the masses just could not understand the basics. . . and time had to be lost to the machine.


Re: Using EV Chargers

Posted: Jun 02 2019 10:27am
by methods
I may have to move this one to the non-reference area.

I was unable to initiate J1772 charge, let alone Level 3/4/5 charge.

This was against V1 Jump Bikes
(FYI - Jump has pushed the V2 bikes out to Santa Cruz. The front disk brake actually stops the bike but the front headlight is going to get one of our kids in town hurt. I really wish jump (aka Uber) had been a little more thoughtful about night riders)
I carefully checked all the settings on the J1772 charger.
IMG_20190423_210640_compress19.jpg (39.66 KiB) Viewed 3676 times
I poked my PayPal card at the Charge Station and it accepted my payment method.
Charge station said - "Ok, poke it in"
IMG_20190423_210910_compress52.jpg (48.85 KiB) Viewed 3676 times
As you can see from the text writings it clearly say "FREE CHARGE SESSION"
... so... I was thinking like "rad!"

I could not figure out where the plug goes on this Electric Vehicle.
I tried the most obvious spot
Cup-Hole in the basket

I figured maybe guru at Uber made Wireless Charge happen?
There is no dice for this.

I tried all the other obvious spots where wireless charge should work.
There is no dice for this either

I think the next step is to kidnap one of these Jump bikes and bring it back to laboratory. Once at evil layer we can find smallest box that accepts 240VAC on one end and puts out a few hundred watts of low-ish voltage DC on the other end.

Show Jump Bikes how to be Radical.
Add a J1772 port
Partner with suppliers of J1772 charge session.


This must be complicated yes?
Not even Zero Motorcycles does it.

I could do this
1 weekend

... then ...
Once we pass "beginner test" of harnessing J1772 standard 240V Input... we can move on to Its big brother, CCS1

This is very intuitive process
One step at time
344344ec39f259aad7ac3f0566ace070.png (59.47 KiB) Viewed 3676 times
First we learn this one
Three big wires
2 small wires

Ok... Two of the big wires must be L1 and L2.... right?
The 240V we seek for power transwer

Third wire must be grossly oversize Earth Ground
The two little wires, they reference this grossly huge Earth Ground.
One is the Control line - where we talk yes?
One is the proximity line - where we see we are connected yes?
images.jpg (6.09 KiB) Viewed 3676 times
Only after you can operate the "old TV" do you go get the new TV and hook it up.
Of course we put old TV and new TV together in one spot - yes?

So nothing has changed
Just two new big pins
That is DC+ and DC-

Those each go to a contactor
Those contactors go to either end of your battery (or your DC-DC converter)

Now the Control Line (not proximity, and referenced to Earth) carries the communications.
These communications are dominated by Qualcomm of course.... but the good news is...

Just like any comms port, you can bang any bytes you want thru it right?
Isn't that was Tesla does?
Does Tesla drag the CAN lines out or do they modulate comms over power?


... slow and steady
Slow and steady... only just the fluffiest of cream off the top...

I spend 2 years digging and digging and digging
You read 2 posts and have caught up
You are welcome


Re: Using EV Chargers

Posted: Jun 02 2019 11:18am
by methods
Meh... I suppose we should actually get to the technical content.


That is a hot-link to Wikipedia
Here is a local copy
J1772_signaling_circuit.gif (16.57 KiB) Viewed 3669 times
You can make this one right?
Using standard methods kit yes?

Few resistors, diode, switch
Go hook it up and see :flame:

Top two big pins are the 240VAC
Bottom big pin is the GND
The other two pins... one has to be Prox and one has to be Pilot

Not going to hurt them - two tries max

... Now...
With your PayPal Debit Card...
and your little black box...

You can get the juice flowing anywhere.
Get that going
Dont kill yourself
Dont start any fires
Dont upset any stations...

Then we talk about turning on the big 50KW++ monsters yes?

(Note - the duty cycle of the +/-12V signal coming in tells you how much power you can draw. Just... dont draw more than a KW and you can ignore this. All the stations can do a KW)


Re: Using EV Chargers

Posted: Jun 02 2019 11:24am
by methods
Warning. . .

This is not a kids site.
I have been spending my days doing stuff like this:
That is a validation run against a 100KW, bi-directional, monster capable of 1.2KV and 334A. In this picture I am making sure its 100KW safety switch blows open.


My perspective is currently blown out of the water.
I have been hooking up cables the size of my thumb to potentials unheard of.



When you were fooling with 36V I was sparking the 100V wires.
Now you may be sparking the 100V wires -> These are 1000V wires.
Please be vary careful

This is not a linear thing
Danger goes up with the square minimum


Re: Using EV Chargers

Posted: Jun 02 2019 2:58pm
by liveforphysics
Tesla superchargers are like you described, but have a step where the vehicle VIN is requested, and the station uses a GSM modem to verify that VIN is approved to charge before it starts.

You should share how to tap many J plugs to get useful charge power. I try to only stop at locations with 4 x jplugs, and none of them shared power. This enables drawing >25kW, which makes reasonable charge rates for a motorcycle (still way too slow for a car).

Re: Using EV Chargers

Posted: Jun 07 2019 9:43am
by methods
Thanks Luke

- So if I hear you clear, you want a PCB board with a PI or an Arduino that can hook up to 4 J1772 stations at once?


And to address the speculation on how power is divided -> You will need to be a little tricky there at times. (PURE SPECULATION HERE - OF COURSE) The stations do the best they can right? So when you first pull up, if it is just you, they may say you can have X amps. If a second guy shows up, and the power feed is not enough, then the PWM will probably change right? Politely asking you to share with the other guy. Maybe now you are at X/1.2 Something like that.

Zero had trouble with this in early testing yes? I heard a squawk about that at lunch one day. They (Zero) would only listen to the available power at time of start. Then when power sharing kicked in, they would over-draw.

... Again pure speculation based on lunchtime banter.

... ...

Can you speak to how the Cable Cooling works in the SuperCharger?

I know previously we spoke of 3 ways to get it done
Baseline Liquid Cooling
Compressor Based Cooling
High Temperature materials and Mesh

I have never cut a cable in half
Have you seen one cut in half?

Re: Using EV Chargers

Posted: Jun 07 2019 10:02am
by methods
For the record...

LFP had "J Connectors" as he calls them zip-tied all over his motorcycle long before I was even thinking about them... so... I can say >3 years ago he was talking about stuff like this.

ChargeRate.png (66.59 KiB) Viewed 3558 times
For Ebikes I am proposing that we can ignore the PWM current limit.

I have not read the spec so I do not know if the spec says that the PWM can change throughout the cycle or if it has to come out up front with a continuous rating.

From my limited perspective...

For J1772 I figure most guys are going to have a retarded box
This box is going to take AC in
This box is going to be rated for a CC/CV out
This box is going to try and push the current limit until CV is reached

... No doubt any number of boxes can be tuned with an Analog input, CAN input, UART input, SPI input, etc.

So for Ebikes we assume Dont Care
For E-Motorbikes we assume a bit more smarts in the AC-DC converter

... This is where we are starting yes?
(Remember I am coming in from out in the cold. This is not my area. . . but it needs to be. . . so if we dont converge on something which can be referenced I will remove this thread)


Re: Using EV Chargers

Posted: Jun 07 2019 10:08am
by methods
And for the readers...

We are just asking questions and reasoning on answers.
We are just looking at available information and acting on it.

There are no secrets around J1772 - but it is the place where all folks start... The baseline filter

... There are however quite a few secrets to be riddled on around this higher power charging that is going on. Luke knows that if you try and pass 50KW, or 100KW, or 150KW, or 200KW, or 250KW that shit gets hairy quick. All sorts of squared problems start to materialize.

Just the other day I tried to run 100KW into a short circuit. :kff:


Re: Using EV Chargers

Posted: Jun 07 2019 10:20am
by methods
Starting in ... l_2015.pdf
I see parts in production

I read that as COTS - Commercial Off The Shelf

No IP there
I will start by making a demo board.


Re: Using EV Chargers

Posted: Jun 07 2019 10:21am
by methods
Clearly old news... 2015

Octopart does not like it, moving on


Re: Using EV Chargers

Posted: Jun 07 2019 10:25am
by methods
Starting 7 years in the past
Good thing I know embedded


Re: Using EV Chargers

Posted: Jun 07 2019 10:28am
by methods
Green light all the way

So far I can tell that I will need SPI

I can create SPI from Arduino, from PI, from any USB Dongle-Box
Looking good so far

My past experience with SPI is that it can go to 1MB
If you wish to go long distance, you use Linear Technology SPI Driver coils. These make SPI which is robust thru differential drivers. Little nugget chips with few perfs.

... Moving along


Re: Using EV Chargers

Posted: Jun 07 2019 10:32am
by methods
From this screen shot you can see that I am starting from a "Clean Machine"
CleanMachine.png (42.02 KiB) Viewed 3549 times
I dont even have an IDE installed on this machine.

This is an ASUS FX504 Gaming laptop
I paid $850 bux for it
Lets see what I can make it worth :P


I use Eclipse for C coding
I have been using Eclipse since it was some busted junk that just came out.
I have used it thru very large projects. Multi-Year projects. It works.


Re: Using EV Chargers

Posted: Jun 07 2019 10:35am
by methods
Here is the Makefile

Code: Select all

# Add include path & extra define
EXTRA_CFLAGS += -Werror -Wall

ifneq ($(KERNELRELEASE),) # kbuild part of makefile

obj-m += qcaspi.o
qcaspi-objs := qca_spi.o qca_framing.o qca_7k.o qca_debug.o


KERNELDIR = ../linux


	install -d $(DESTDIR)/root
	install qcaspi.ko $(DESTDIR)/root/

	$(MAKE) -C $(KERNELDIR) M=$(PWD) clean
	@rm -f Module.symvers
These guys use Linux obviously
I am a Windows guy
I also recently worked at a spot where everyone was using Linux

Just to prove I dont need to bite anybodies IP...
I will just go ahead and solve this riddle for the Windows Machine

... I will do it in Arduino if I can get it to work. Otherwise I am going to compile for a PI.

Re: Using EV Chargers

Posted: Jun 07 2019 10:38am
by methods

When you are walking into a :kff: storm like this you can go about it 2 ways

Code: Select all

 * qcaspi_mod_init - Module initialize function.
 * This is the entry point of the kernel module. It is called when the module
 * is loaded. It registers SPI and network devices.
 * Return:   0 on success, else failure

static int __init
	struct spi_board_info *spi_board = NULL;
	struct spi_master *spi_master = NULL;
	struct spi_device *spi_device = NULL;
	struct qcaspi *qca = NULL;

	printk(KERN_INFO "qcaspi: version %s\n", QCASPI_VERSION);

	if ((qcaspi_clkspeed < QCASPI_CLK_SPEED_MIN) ||
	    (qcaspi_clkspeed > QCASPI_CLK_SPEED_MAX) ||
	    (qcaspi_legacy_mode < QCASPI_LEGACY_MODE_MIN) ||
	    (qcaspi_legacy_mode > QCASPI_LEGACY_MODE_MAX) ||
	    (qcaspi_burst_len < QCASPI_BURST_LEN_MIN) ||
	    (qcaspi_burst_len > QCASPI_BURST_LEN_MAX)) 
I can invest hours, days, weeks, months into understanding this code and getting it working.
I can invest hours, days, weeks, months into the actual datasheet and write my own code

NO DOUBT all this code is great and NO DOUBT it does a thousand things I dont need it to do in a super complicated way.

next we get the Data Sheet for the part.

That is the magic decoder ring for the above garbage.


Re: Using EV Chargers

Posted: Jun 07 2019 10:41am
by methods
No idea what this is ... 4602ff.pdf

And remember
This is the Technical Reference Area

We are not here to show you how to do it.
We are here to show you how it is done.
Then you can do it yourself

I am just tearing into this like I have torn into any big project over the last 20 years.
Figure out what it is
Solve it


You need to be competent in Electrical Engineering, Software Engineering, Project Management, IP Protection, Public Speaking, Writing, . . .. LOL

This is going to be fun


Re: Using EV Chargers

Posted: Jun 07 2019 10:47am
by methods
They say this:
CCS: Single charging interface capable of supporting both AC and DC charging
− Includes physical connector definition (SAE J1772)
− low level signaling (SAE J1772)
− high level communications between PEV and EVSE (SAE J2931 and ISO/IEC 15118)
• CCS Relies on HomePlug Green PHY for high level PEV/EVSE Communications
− Use of HPGP is mandatory for DC charging
− Optional for AC charging
− HPGP-equipped AC chargers are referred to as “Smart AC Chargers”
Right now I am trying to come up with the Google Search Terms which will get me the most recent chip set. It is a nOoB Mistake to go tearing into a 7 year old design when some nice new SOLVED nugget may exist.

From the writings I can see that these chips are ball grid array. That sucks right off the bat, and we are not in this for profit, so we seek a Dongle, a Nugget, a Demo-Board, or something which will not force us to set up a BGA process in my living room.


Just prefer not to
And.. none of us can afford to turn-key shit like this. I am not going to pay $1000 for a PCB board which I can make 50 of for the same price.



Re: Using EV Chargers

Posted: Jun 07 2019 10:48am
by methods
First Spec - not available
WIP - Work in progress


Re: Using EV Chargers

Posted: Jun 07 2019 10:50am
by methods
Now someone wants to sell me shit

No thanks guys
I did not buy CHAdeMO spec
I am not going to buy CCS spec


Re: Using EV Chargers

Posted: Jun 07 2019 10:52am
by methods
This is required Charge Infrastructure for the WORLD to move forward on.

This is not some IP for your mom to retire on.

I am methods
I watched this garbage roll out for years on end
If I can not get it going... then you dont have the show on the road

... Making someone buy a spec is not any kind of safety precaution. Safety comes from CLEARLY CALLING OUT

the API



Looking for that now
I want to hook up to a Level 5 charge station

What hardware do I need?
What API am I hitting against?
What is MVP for communications?

That is all I want to know

(see how this works? You gotta know what you are after)


Re: Using EV Chargers

Posted: Jun 07 2019 10:54am
by methods
. . .

If you are a company trying to deal with this crap there are a lot of ways to go about it.

I am the kind of guy you can hire
Who will go in there and pull feathers


You can hire other types of guys


You may just want to go to each company which provides Charge and try to System Integrate with them - YES?
No doubt tho
You dont want to show up with an ARduino duct-taped to some Alkaline batteries and some code you got from your friend


Be ready
I am getting ready
My ISH will work :confused:


Re: Using EV Chargers

Posted: Jun 07 2019 10:58am
by methods
Look at all this crap . . . sigh

All that is good to have in the back pocket
What we want on the White board is

1) Simple Diagram
- Your side looks like THIS

2) Simple Flow Chart
- Your code flows like THIS

3) Hello World Handshake
- Get this working, then we talk

All big projects start out like that

What hardware will I need?
What is the basic dance we will do?
What can I use as a Unit Test or as a baseline test or as a Hello World or as a Handshake


Re: Using EV Chargers

Posted: Jun 07 2019 11:01am
by methods
There are 50 EV companies out there now

That means this has been solved at least 75 times... because.. a lot of companies already failed

Each of them took a long time to solve it
No doubt they want to make money off of that


How many guys are floating around who already solved it?
Who know what hardware looks like?
Who know what software looks like?
Who know basic dance already?

... NOT ME

So... Lets see if I can do it.

I have never seen this flow diagram
I have only seen how cars interact with stations

I have never seen any of this software
I have seen some logs here and there... of endless BLAH BLAH between stations and cars. I paid no attention to any of that. Looks like bunch of GarbleGoop to me

I am minimalist
I solve MVP
What is MINIMUM I have to do to turn on Level 5 Juice?

I want 2,000A right now at 420V
Turn on please
... Poking button... TURN ON PLEASE!!!

I have my 50W resistor attached
I wish to produce smoke which is magic
I catch this smoke in glass jar
Sell on internet - get Rich quick!


Re: Using EV Chargers

Posted: Jun 07 2019 11:03am
by methods
So. . .


We have Forward Engineer
We have Reverse Engineer

I can do both

Forward Engineer spends 1 week minimum just reviewing materials
I print many things
I read and read and read

Reverse Engineer spends 1 week minimum just building tools
I focus on collecting known-good information

Both cost me $10k out of pocket

That is the first thing to understand


... It is summer tho...