E-MotorWerks JuiceBox - an open source 15kW EVSE

[valerun]

hey guys,
new leaf owner here and just bought the premium juice box (which works great btw).

The Roving WiFi rn-vx board wasn't in stock when I bought the kit so I am trying to add it now however after following the instructions on how to install it I am having a hard time to program it.

I uploaded EMW_EVSE_WiFi_programmer_Public but then when I connect to the serial port all I see is:

bash-3.2# cu -s 9600 -l /dev/cu.usbserial-A602W9F7
Connected.
Starting
Free memory: 833
Failed to start wifly
Terminal ready

nothing seems to be working from here. The board lights up and I was also able to connect to it over the ad-hoc wifi and enter command mode, type commands, etc.

not sure what to do at this point, can anyone please help ?

thanks !
-g

Hi gabb -

looks like Arduino is not talking to your rn171-xv. Most likely reason is some break in the rx/tx lines between them. First off, can you make sure you followed instructions on page 44 of the Build Manual (https://docs.google.com/a/emotorwerks.com/document/d/127joe-QgHYRQVD4Zl2-oZO3A9gRjJZC4ReH_2jIM_x0/edit" onclick="window.open(this.href);return false - installing U31 and other components required for level translation between Arduino and the module?

V

 

[gabb]

hey guys,
new leaf owner here and just bought the premium juice box (which works great btw).

The Roving WiFi rn-vx board wasn't in stock when I bought the kit so I am trying to add it now however after following the instructions on how to install it I am having a hard time to program it.

I uploaded EMW_EVSE_WiFi_programmer_Public but then when I connect to the serial port all I see is:

bash-3.2# cu -s 9600 -l /dev/cu.usbserial-A602W9F7
Connected.
Starting
Free memory: 833
Failed to start wifly
Terminal ready

nothing seems to be working from here. The board lights up and I was also able to connect to it over the ad-hoc wifi and enter command mode, type commands, etc.

not sure what to do at this point, can anyone please help ?

thanks !
-g

Hi gabb -

looks like Arduino is not talking to your rn171-xv. Most likely reason is some break in the rx/tx lines between them. First off, can you make sure you followed instructions on page 44 of the Build Manual (https://docs.google.com/a/emotorwerks.com/document/d/127joe-QgHYRQVD4Zl2-oZO3A9gRjJZC4ReH_2jIM_x0/edit" onclick="window.open(this.href);return false - installing U31 and other components required for level translation between Arduino and the module?

V

thanks for your help. What you say it makes sense, I kind of figured the problem might be there as well.
I did follow the instructions. I only installed the U31 ic converter and the Q31 regulator (the only components provided with the wifi module - not sure what are the 'other components required' you're referring to are?)

 

[valerun]

thanks for your help. What you say it makes sense, I kind of figured the problem might be there as well.
I did follow the instructions. I only installed the U31 ic converter and the Q31 regulator (the only components provided with the wifi module - not sure what are the 'other components required' you're referring to are?)

Can you post / send to [email protected] a photo of your board showing area around U31?

Thanks!

 

[joewaters]

I'd like to see RTC on the next gen Juicebox firmware, since we have plenty of power outages here. I have had a couple of ICE days in the last 2 weeks because my power flickered and the JB timer did not kick back in.
// need some RTC functions here...
//---------------- END timing functions -----------------------------------------------
(looks like plans were in the works, but didn't make it out of the gate.)
adding the DS1307/DS3231 is easy enough in hardware, I just don't have time/skill to write the firmware.

 

[gabb]

thanks for your help. What you say it makes sense, I kind of figured the problem might be there as well.
I did follow the instructions. I only installed the U31 ic converter and the Q31 regulator (the only components provided with the wifi module - not sure what are the 'other components required' you're referring to are?)

Can you post / send to [email protected] a photo of your board showing area around U31?

Thanks!

 

[dsinned]

Its very difficult to see in your pictures, but some of the solder joints at the pins along the edge of the WiFI module's board to board connections look rather poor. One in particular looks like a solder "ball" formed, which you might want to touch up. Can you try to refocus the pictures so we can better see these kind of details?

Also, you need to be more careful with your soldering iron as it looks like some of the insulation melted on the test wire for the GFCI circuit.

As an aside to Valery, I don't see a WPS "button" in these pictures, only a silkscreen reference for it. Where is it?

 

[gabb]

Its very difficult to see in your pictures, but some of the solder joints at the pins along the edge of the WiFI module's board to board connections look rather poor. One in particular looks like a solder "ball" formed, which you might want to touch up. Can you try to refocus the pictures so we can better see these kind of details?

Also, you need to be more careful with your soldering iron as it looks like some of the insulation melted on the test wire for the GFCI circuit.

As an aside to Valery, I don't see a WPS "button" in these pictures, only a silkscreen reference for it. Where is it?

yes, the soldering should be good, I checked and double checked.
The U31 convertor seems to be working, it gets 5v on one side and it outputs 3.3 that goes to the wifi's module '3v3' pin.

one thing I noticed is that when I manually trigger the module's WPS I do see garbage on the terminal (see pic bellow)

that tells me that at lease the TX path should be working.

any other ideas of what I can try to debug this further ? I am stuck with this disassembled at the moment for about a week now and I'd like to be able to use it to charge my car again

thanks,
-gabriel

 

[gabb]

any other ideas of what I can try to debug this further ? I am stuck with this disassembled at the moment for about a week now and I'd like to be able to use it to charge my car again

FIXED IT
so it was a cold soldering point from arduino's pin4 to the voltage conv. It looked great with the magnifying glass but there was no contact underneath. sigh ... I can now put it all back together and start charging again...

 

[valerun]

any other ideas of what I can try to debug this further ? I am stuck with this disassembled at the moment for about a week now and I'd like to be able to use it to charge my car again

FIXED IT
so it was a cold soldering point from arduino's pin4 to the voltage conv. It looked great with the magnifying glass but there was no contact underneath. sigh ... I can now put it all back together and start charging again...

Awesome! Thanks for reporting back!

Val

 

[valerun]

Hi All -

Great news - next revision of JuiceBox Open Source EVSE is here. All orders from tomorrow will be this new version

It is based on the new PCB version - V8.9 - and has a few nice additional features:

1. All-Surface-Mount PCB. We have moved 95% of the parts to SMT. This means lighter and more reliable board (easier cable routing, higher EMI immunity, lower susceptibility to moisture, etc). This has also reduced our cost a bit which allowed us to add all these nice features below

2. Perhaps the biggest one is Hardware GFCI protection. We have added an additional layer of GFCI protection on top of the interrupt-based software layer. Now, even if the main processor freezes, the GFCI event will open the relay and de-energize the output. The feature is implemented using a D-type flip-flop IC which will stay latched until reset by the MCU. This is an important safety improvement and is one of the steps we took towards getting the design UL-certifiable

3. On-board temperature monitor. This will provide over-temperature protection in case unit overheats for any reason (placed in the direct sunlight, in the overcurrent event, etc)

4. PCB heater. Yes, we have added a small 6-watt heating element to the bottom of the PCB. It would turn on below 0 degrees F and turn off once the processor core heats above that temperature. This will result in more reliable operation in extremely low temperatures (we did NOT receive any complaints from our customers in Canada / midwest even with a previous version so this is mostly a preventive measure)

5. Good news for our kit builders - all WiFi communication components and WiFi module sockets are now pre-placed on the board. All you have to do now in order to add WiFi is insert the module into the PCB sockets and (optionally) solder on the WPS button

6. A bit more technical change - we have moved to a zero-drift op-amp for GFCI / voltage measurement circuit. This allowed us to get rid of the vast majority of the false GFCI trips (a problem previously reported by ~3-4% of customers) and get rid of complex calibration process.

7. All resistive components have been replaced with 1% tolerance parts. This will means higher precision of the voltage and current measurements.

8. Better placement / configuration of the pilot line connection

9. Pads for large cable terminals were added to the bottom of the board. This is the provision for future addition of laminated copper busbars to eliminate cable alignment issues (busbars are not quite ready yet but the board is ready for them as soon as we get them out).


In addition to hardware-based changes, the following pure software changes were made in the latest V8.9.1 firmware:

1. More consistent display states for states other than A and C (A and C were already well covered in the previous versions)

2. More precise current measurement

3. Immediate WiFi reporting on plug-in (so you don't have to wait 1-10 minutes to see the first data from your unit)


As usual, all related design files can be downloaded from our JuiceBox folder at http://emotorwerks.com/JuiceBox_Public/" onclick="window.open(this.href);return false;. Everything should be there by now - the only thing we are missing is schematics which is being updated right now.

Thanks,
Valery.

 

[valerun]

a couple of additional pics. This is a Base Edition unit with current sensing and external WiFi antenna.

The first photo is of the SQL database data as reported by this unit - charging one of our Leafs...

 

[dsinned]

AWESOME!!!

My JB (ordered last month) already has the new antenna mast; a big improvement for the Basic unit.

Onboard temperature sensing and higher precision AC voltage and current sensing are welcome additions too!

Changing to a zero offset OPAMP and eliminating R26 should simplify calibration for users that do not have a good quality multimeter. (My JB was quite problematic with respect to proper AC voltage sensing, requiring R26 calibration in addition to a simple firmware based "voltage sensitivity" mod. Glad I got the FTDI option!)

One suggestion to make future PCBs even better is provision for a set of four (2x2) standoff holes on the board to accomodate placement of the LCD opton (facing up). This would permit Basic JB owners to replace the metal cover plate with a sheet of clear plexiglass, so they can view of the LCD mounted "inside" the enclosure. It also would permit Basic users to opt for the remote key fob option. Also, a fairly thick plexiglass cover, with no other cutouts required in the enclosure, would maintain watertight integrity for outdoor use! I actually made this cover mod to my fully optioned enclosure and it worked quite well.

 

[valerun]

AWESOME!!!

My JB (ordered last month) already has the new antenna mast; a big improvement for the Basic unit.

Onboard temperature sensing and higher precision AC voltage and current sensing are welcome additions too!

Changing to a zero offset OPAMP and eliminating R26 should simplify calibration for users that do not have a good quality multimeter. (My JB was quite problematic with respect to proper AC voltage sensing, requiring R26 calibration in addition to a simple firmware based "voltage sensitivity" mod. Glad I got the FTDI option!)

One suggestion to make future PCBs even better is provision for a set of four (2x2) standoff holes on the board to accomodate placement of the LCD opton (facing up). This would permit Basic JB owners to replace the metal cover plate with a sheet of clear plexiglass, so they can view of the LCD mounted "inside" the enclosure. It also would permit Basic users to opt for the remote key fob option. A fairly thick plexiglass cover, with no other cutouts required in the enclosure, would still maintain watertight integrity for outdoor use! I actually made this mod to my fully optioned enclosure and it worked out quite well.

Thanks!

BTW temp sensing and [some degree of] higher precision voltage sensing is possible on V8.7 boards, as well. We will be issuing V8.7.11 firmware revision soon with those additions.

 

[dsinned]

Outstanding!!! Can't wait!!! Thanks!!!

Will the temperature of the internal JB be displayed on the LCD as well?

I was checking my fully optioned JBB (JuiceBox Basic) last night while charging my RAV4 at 7.75kW. The surface of the enclosure was "warm" to the touch, but not out of line with expectations. The AC input - output cords (with 8AWG power conductors) were just barely over ambient temperature. It would be really nice to have an LCD readout of the internal temp via a firmware upgrade to V8.7.x and a bit more resolution to the AC voltage sensing which affects the LCD's "kW" readout!

 

[Dave]

I have my v8.9 installed and it is charging the Leaf great fron 240v, 27.5A, but the GFCI seems to not be working.
Everything is hooked up correctly and I have verified power to the GFCI test circuit. J cable hot leads are routed through current sensor correctly.
I used a GFCI tester and couldn't get it to trip while charging vehicle.
I used a 6k resistor with my DMM and measures 800ma leaked to ground from J side while charging and it didn't trip.
I have verified all grounds to PCB, supply, J cable.
Sorry to post here, but I have had no luck with your email support.
Any suggestions?

 

[GlennD]

I have my v8.9 installed and it is charging the Leaf great fron 240v, 27.5A, but the GFCI seems to not be working.
Everything is hooked up correctly and I have verified power to the GFCI test circuit. J cable hot leads are routed through current sensor correctly.
I used a GFCI tester and couldn't get it to trip while charging vehicle.
I used a 6k resistor with my DMM and measures 800ma leaked to ground from J side while charging and it didn't trip.
I have verified all grounds to PCB, supply, J cable.
Sorry to post here, but I have had no luck with your email support.
Any suggestions?

I have a box with 10K and 5K resistors to check EVSE GFCI circuits. You are checking one hot (120V) so a 10K draws 12Ma and a 5K draws 24Ma. My 8.9 board worked fine and tripped at 24Ma and not at 12MA. The box goes on one hot lead and ground.

You can not use a normal 5Ma GFCI tester since a car is set for 20Ma.

I can not figure how you had 800Ma.

 

[valerun]

I have my v8.9 installed and it is charging the Leaf great fron 240v, 27.5A, but the GFCI seems to not be working.
Everything is hooked up correctly and I have verified power to the GFCI test circuit. J cable hot leads are routed through current sensor correctly.
I used a GFCI tester and couldn't get it to trip while charging vehicle.
I used a 6k resistor with my DMM and measures 800ma leaked to ground from J side while charging and it didn't trip.
I have verified all grounds to PCB, supply, J cable.
Sorry to post here, but I have had no luck with your email support.
Any suggestions?

Hi Dave -

Am on a trip now but will shoot a short video tomorrow with this test.

BTW 800ma seems very high - on a 6k resistor, this is 4,800V...

PS. Couldn't find your email in my box - perhaps I am looking for a wrong name. When did you send your email? Did anyone reply at all?

Thanks,
Valery.

 

[dsinned]

Outstanding!!! Can't wait!!! Thanks!!!

Will the temperature of the internal JB be displayed on the LCD as well?

I was checking my fully optioned JBB (JuiceBox Basic) last night while charging my RAV4 at 7.75kW. The surface of the enclosure was "warm" to the touch, but not out of line with expectations. The AC input - output cords (with 8AWG power conductors) were just barely over ambient temperature. It would be really nice to have an LCD readout of the internal temp via a firmware upgrade to V8.7.x and a bit more resolution to the AC voltage sensing which affects the LCD's "kW" readout!

Hi Valery. I also wanted to know the answer to the question (stated above) asked a week ago. Please advise. TIA.

Btw, I switched my JBB hardware over to the 4D System's 1.7" OLED-160-G2 and it worked just fine as a replacement to the 1.44" LCD. However, in the JB menu setup mode, the display scrolls horizontally repetitively after a delay of ~5 seconds, but it does not do this on any other display screen. There is now nearly 1/2" more horizontal line length in landscape mode, but hardly any more vertically. I am waiting for a 4GB Micro SD-HC to load a slightly larger, Microsoft Windows font, (Arial Narrow, 10 char/in), to see if this helps to enlarge the information displayed any better.

***UPDATE***
The reason the 1.7" uOLED-160-G2 "scrolls" the Menu Setup screen is because if that page remains static for longer than 5 seconds, i.e. no menu commands entered via the key fob, the display automatically goes in to screensaver mode. The 1.44" uLCD-128-G2 does not do this, because there is no screensaver mode embedded in that particular display. I feel the screen saver feature is desireable, as is a larger display in the JB application.

Although ~$30 more expensive, the 1.7" OLED is simply a better choice. In default landscape mode, the extra line length (160 vs. 128 pixels), has advantages in the JB application. Both displays have the same built-in "system" font, consisting of characters of 7x8 pixels. Assuming one pixel separation between characters (in a string), there is space for 20 characters per line with the uOLED. With the uLCD, the limit is 16.

In the latest version of JB firmware, V8.9.1, space for 4 extra characters per line is necessary in order to support the internal temperature readout at the end of the first line of the display as discussed in the post below.

 

[Dave]

Figured it out.
The load has to be placed on the vehicle side of the current sensor.
I tried and the GFCI trip instantly.
Still a little unclear on the on board testing procedure, but hopefully the new instructions clear it up.
Dave

 

[dsinned]

For those of you with the new V8.9 PCB in a Premium Enclosure (or Basic with LCD option) please check something for me. Are there three digits in parentheses at the end of the first line displayed on your LCD? I believe the first digit (0, 1, or 2) represents the "state" of your J1772 to JB connection. The next two digits represents a temperature in degrees C. I believe that number is the internal temperature of the JB's Ardunio microcontroller.

I came to this conclusion after studying the latest V8.9.1 firmware source code and I think the temperature readout was a newly added feature.

Here is an example of how the first line appears on the JB (with LCD) before and after the new PCB was released:

Old vers: V8.7.9 08:55 (2)
New ver: V8.9.1 08:55 (2,55) ; where the last two digits indicate an internal JB temperature of 55C, (or 131F).

As a side note, the way the new temperature management feature works, if an OTP (over temperature protection) condition occurs, when the internal temperature of the Arudino microcontroller exceeds 90C, the JB will suspend active charging for at least 10 minutes, thus preventing overheating. Conversely, if the temperature drops below -20C, a 6W heater energizes to try to keep the JB above freezing. Above -10C, the heater turns off.

To best of my knowledge, no other EVSE has implemented these feature!

Valery, is this correct?