OpenEVSE - Open Source Charging Station

[essaunders]

Has anyone tried to contact them to see if it's availble with 70A rated wiring/cable? It would be more, but that's what I am looking to do next.

Update: I called them, they do not currently offer this in the 70-75A version, as they do not offer an EVSE with that capability yet, and these are really just replacement cables for their own products.

Boooo

So the connector is 70A, but the cable on 30A? (just want to clarify...)

but still, looks like a good price.

 

[chris1howell]

I don't think it's being looked at to replace the Arduino, the timing on the code has to be right, and running a unix variant on the Rasberry probably precludes its use for that

Hmm, I would hope that the clock would be accurate enough - one of it's target uses is for mini multimedia purposes which typically needs a good clock...

How accurate does the clock need to be to generate a stable pilot signal?

It's not the clock that's the issue, it's the fact it's running an operating system... Even that "GERT" board has a similar (if not almost the same as the Aruino) ATMEL microcontroller on it that OpenEVSE uses. Operating systems tend to have interrupts and don't make great finite state machines for a time sensitive task, like generating the 1KHZ pilot signal, for example. You also need the hardware on the OpenEVSE as well, for the interface to the pilot signal. With its Ethernet interface, possible WiFi interface, and web server that can be run on a $35 PCB, it's great to control things with, but would probably not do great at real time sensitive control. I coud be totally wrong here, but I don't think so. OpenEVSE is running a dedicated program, and not an operating system, that's the big difference.

GERT board details: http://www.element14.com/community/docs/DOC-48859?ICID=raspberrypi_gertlearnmore" onclick="window.open(this.href);return false;

The problem is I/O, there is only 1 PWM pin which may be able to drive the 1khz pilot, but there is no analog to digital converter to read the pilot. I think the best solution is to use the OpenEVSE board as is and send state data and commands over i2c or serial. That way all critical EVSE functions are handled in the OpenEVSE hardware.

http://www.raspberrypi.org/phpBB3/viewtopic.php?f=5&t=5391[/url]"]RasPi is a completely digital board, with no A/D or D/A available. If you need hard real-time control functions, your easiest solution is to use an Arduino or similar, and interface that to the RasPi if needed.

 

[Nekota]

Anyone looking at the Rasberry Pi? One of the distributors released the Gertboard GPIO kit which looks to be way overkill for what the Open EVSE requires, but with a full Linux distro running behind it the sky is pretty much the limit in terms of customization.

I heard that that there's a project that lets one port code more or less directly from the Arduino to the Rasberry Pi as well...

The RPI is a nice inexpensive linux computer and would be very easy to interface to provide the controls for EVSE without the Gertboard GPIO kit. The RPI has 18 GPIO pins available with a 3.3V logic interface and a 'Wiring Library' to provide an Arduino like interface to the GPIO was recently released. But I would think an mBed would make a better match for the EVSE as it is smaller, has more GPIO and serial interfaces, a real time clock and is much smaller making it easier to integrate into a package. The development tools for the mBed are easier to use than the RPI but that will probably change as the RPI gets more open software. I just added a RPI to my collection of small computers including the Arduino, mbeds, pogoplugs, 'central axis' NAS and the Linksys WRT54G routers.

 

[GlennD]

I replaced my garage unit. I wanted more room so that my delay timer would fit inside. As an added benefit there was room to use a contactor instead of a relay.

 

[GlennD]

I ordered the Levitron cable to see what it was like and also since it was claimed 25 feet long. I received it today.

The connector is ITT and the cable is UL listed. It comes with the same RU tag my $280 shipped cable had. This is the real deal and it is currently a bargain. The cable has the same ratings but it seems to be a little less flexible. This weekend I am going to swap cables to gain 5 feet.

Anyone want an Avnet cable?

 

[GlennD]

I just measured the cable. Excluding the connector it is 24FT 4IN. If you include the J1772 connector it is over 25Ft.

 

[mitch672]

I just measured the cable. Excluding the connector it is 24FT 4IN. If you include the J1772 connector it is over 25Ft.

Glenn, I just bought an ITT 75A J-1772 on eBay for $250 (the same seller also has one up now for $225, BIN $275 ($9.99 shipping). Yes, they are only 20', my solution is to get 4' of liquid tight, connect to my new 75A OpenEVSE I am starting to build this weekend. There will be a junction box at the end of the liquid tight where the J-1772 cable will be connected in... I will only need to strip 4-6" this way, as opposed to 12" or more of the j-1772 cables jacket, terminating to the 75A SquareD contactor inside a rather large 12 X 16 NEMA enclosure. This also allows the start of the 20' cable to be near the garage floor, so it will be like having a 25' cable (hint for anyone needing a few extra feet)

I will post some pictures during construction, so far the 75A J-1772 cable and contactor are in, enclosure and other parts arriving from Digikey tommorow or Friday. Why a 75A OpenEVSE? Think "Tesla Model S" in your future

 

[GlennD]

As thick as the 30A cable is I would only go with the 75A cable if I had a pressing need for it. In my portable unit I have the 16A cable and it is like large drop cord.

 

[mitch672]

The 75A cable is 1" in diameter, not much larger than the 30A cable, which is 3/4" anyway, the ITT 75A cable contains (2) #6 for the hots (red and black), the ground conductor is #8 (green)' the pilot and proximity wires are each #16 (orange and blue)

I only want to build one more OpenEVSE, and this is it

Sold my Schnieder EV2430WS (thanks "plugmein") and my 30A portable OpenEVSE, this will be my only one, other than what comes with each vehicle.

Mitch

 

[mrbigh]

Witch wire color is the pilot????
Is it there an standard color coding? or it does change by different manufacturers?

the pilot and proximity wires are each #16 (orange and blue)

Mitch

 

[mitch672]

Witch wire color is the pilot????
Is it there an standard color coding? or it does change by different manufacturers?

the pilot and proximity wires are each #16 (orange and blue)

Mitch

Orange on the ITT J-1772 cable:

 

[GlennD]

Since my town house (non electric) has only 100A service 30A is as much as I can reserve for a car charging.

 

[GlennD]

This weekend I changed out my J1772 cable. Here is a comparison of the ITT 30A cable and the Levitron 30A cable. Both are about the same diameter.

Both use ITT's J1772 connector and both use cable with a UL stamp. Both cables have #10 power and #16 signal wires.

The ITT cable conductors are wrapped in paper and then covered with a very heavy outer jacket that looks like silicon rubber. It is more flexible and it looks like it would be more flexible in cold weather. It is around 19 feet long.

The Levitron cable conductors have 3 paper sudo wires to make it round and it is then wrapped in paper. It then has a thinner rubber jacket. In all it is constructed like high quality extension cable. My cable was just over 24 feet long. It does not look to be as flexible in cold weather.

For my use when in Southern California 40 degrees is considered cold, Both cables are fine. I think the ITT might be more durable with its thicker jacket but the extra 5 feet of the Levitron means I can charge my car without backing in.

 

[kubel]

Sorry if this is a stupid question, but does the polarity of the CT matter? The two pins don't appear to be separately labeled so that's why I was wondering.

 

[mitch672]

It's just a coil, the polarity does not matter.

 

[garygid]

If one measures only AC leakage, the polarity of the coil would
not matter. However, if there is a DC component, the readings
could be "strange", depending upon how the measurements
are made. Right?

 

[Jklafehn]

Fabulous job Chris!!

I just read this entire thread. Most of it went over my head but it was very interesting to see all the work that has gone into this project!

Im really looking forward to receiving my boards!

Im building two double evse units for each of my rental properties.

 

[Deleted member 3147]

If one measures only AC leakage, the polarity of the coil would
not matter. However, if there is a DC component, the readings
could be "strange", depending upon how the measurements
are made. Right?

AFAIK, the coil cannot detect DC current; the physics just don't work (though I am unsure of the details, I'm not an EE). I do know that DC ammeter clamp meters(e. g. http://www.fluke.com/Fluke/uken/Accessories/Current-clamps/i30s.htm?PID=56297) use Hall Effect sensors, not coils.

 

[eaf]

He meant to say that if charging current was DC, a ground leak would create a disbalance that would either give positive or negative spike on the coil's wires, and the polarity would matter. But given that charging current is AC, if a leak is present you'll see a series of negative-positive-negative-... spikes 20ms apart, so no matter how you connect that coil, within 20ms, the schematics will register one spike or the other.

 

[GlennD]

This is my latest Open EVSE. I built it to permanently mount at work. I am tired of unpacking my portable unit every day.

This is an Adruino based unit. It consists of a stack comprising an Arduino Uno, A Open EVSE shield, and an Adafruit RGB shield. It is mounted in a large Adafruit enclosure. It auto selects L1 & L2.