triangles
Well-known member
After nearly 6 months my JuiceBox is finally working right! JuiceBox tech support did try to help but started ignoring me after about 2 months. It took much fiddling, a little help from tech support, and some help from an electrical engineer friend to finally exercise the demons from this EVSE. Mostly it was my EE friend...
First a little background: I ordered my juice box right when PCB ver. 8.9 came out (August 2014). Since I didn't get my car till mid November it sat un used untested until then. At first I could not get my JuiceBox to "boot" at 240V. It would power up normally on 120V. I have the standard kit and I added the second current sensor, screen, and remote. The unit seems to boot up fine but when the relay is engaged things go wrong. very infrequently it would come up normally and charge my car. Sometimes the screen will go blank but it will still charge my car. Sometimes the screen will invert. Sometimes it will display what appears to be a generic "boot" screen of the screen manufacturer. Other times the screen will be blank and it wont charge. What ever the issue is it seemed to be getting worse. I assumed it was tripping on the GFCI self test at startup as I never checked the GFCI when I built it since I didn't have my car yet. In the pictures/video the screen is loose since I chose to install the screen in the lid of the enclosure. Video of crazy screen inversion at boot
Pictures of inverted screens:
A pic of my JuiceBox during testing:
Not shown I added a "power" switch that turned the unit on and off by interrupting power to the circuit board's power supply. The switch also has a blue LED ring around the button. I wired this to turn on when the contactor was energized so that I could have a quick visual cue that the EVSE was supplying power to the car. I had it working such that I would have to power cycle about 3-4 times to get past the boot sequence. Also 9.5 times out of 10 when plugging in the car I would get a false GFCI trip (this is why I ended up with a dead 12v battery once since the 12v battery powers some coolant pumps and the car thinks it should be getting juice but isn't. This eventually kills the 12v battery).
To keep this post from getting too long, here's what I discovered and the solution. From my emails to tech support I learned that something was crashing the Arduino processor. I concluded it was electrical "noise" being picked up by the wires going to the LCD. My EE friend suggested it was at least partially due to the unnecessarily long wires going to the J1772 connector(Black and red wires you see snaking around the enclosure). He suggested I shorten these wires and keep them as far away from the processor/lcd area of the board as practical. He said I should also shield the LCD wires and ground one end of the shield to the enclosure. Since I didn't have any terminals on hand I decided to fold the loop of extra wire around the contactor(relay). My EE friend said I should keep the wires as parallel and close together as possible. I ended up using 4 zip ties to bundle the wires around the contactor. I removed the green wire (it's not used for anything) going to the LCD and wrapped the remaining wires in aluminum foil. I stripped off a length of insulation from a wire and laid it in with the foil to add a means of grounding the shield. I then wrapped the foil/ground wire with electrical tape so it wouldn't short anything out. I attached the ground wire to the upper corner screw holding in the PCB. Sorry I didn't take any pictures since I wasn't too confident this would work. I put the JB back together and to my amazement I was not able to get any faults. I've only power cycled a dozen times and plugged in a few since so I'm not 100% certain it's fixed but it sure seems so.
I can't fault Emotorwerks too much since I strayed from their design a bit. They could have handled tech support a bit better than simply ignoring me when their initial suggestions didn't remedy the problem.
I went with a JuiceBox because of price and I wanted a method to see how many KWh go into the car when charging. The shortcoming is that it resets to 0 every time the EVSE turns the power going to the car off and then on again. So if the car charges and I have a go time set, I better get to the car before the go time if I want to see how many KWh it took to charge the car. The ability to manually set the max charge rate is also a nice bonus.
First a little background: I ordered my juice box right when PCB ver. 8.9 came out (August 2014). Since I didn't get my car till mid November it sat un used untested until then. At first I could not get my JuiceBox to "boot" at 240V. It would power up normally on 120V. I have the standard kit and I added the second current sensor, screen, and remote. The unit seems to boot up fine but when the relay is engaged things go wrong. very infrequently it would come up normally and charge my car. Sometimes the screen will go blank but it will still charge my car. Sometimes the screen will invert. Sometimes it will display what appears to be a generic "boot" screen of the screen manufacturer. Other times the screen will be blank and it wont charge. What ever the issue is it seemed to be getting worse. I assumed it was tripping on the GFCI self test at startup as I never checked the GFCI when I built it since I didn't have my car yet. In the pictures/video the screen is loose since I chose to install the screen in the lid of the enclosure. Video of crazy screen inversion at boot
Pictures of inverted screens:
A pic of my JuiceBox during testing:
Not shown I added a "power" switch that turned the unit on and off by interrupting power to the circuit board's power supply. The switch also has a blue LED ring around the button. I wired this to turn on when the contactor was energized so that I could have a quick visual cue that the EVSE was supplying power to the car. I had it working such that I would have to power cycle about 3-4 times to get past the boot sequence. Also 9.5 times out of 10 when plugging in the car I would get a false GFCI trip (this is why I ended up with a dead 12v battery once since the 12v battery powers some coolant pumps and the car thinks it should be getting juice but isn't. This eventually kills the 12v battery).
To keep this post from getting too long, here's what I discovered and the solution. From my emails to tech support I learned that something was crashing the Arduino processor. I concluded it was electrical "noise" being picked up by the wires going to the LCD. My EE friend suggested it was at least partially due to the unnecessarily long wires going to the J1772 connector(Black and red wires you see snaking around the enclosure). He suggested I shorten these wires and keep them as far away from the processor/lcd area of the board as practical. He said I should also shield the LCD wires and ground one end of the shield to the enclosure. Since I didn't have any terminals on hand I decided to fold the loop of extra wire around the contactor(relay). My EE friend said I should keep the wires as parallel and close together as possible. I ended up using 4 zip ties to bundle the wires around the contactor. I removed the green wire (it's not used for anything) going to the LCD and wrapped the remaining wires in aluminum foil. I stripped off a length of insulation from a wire and laid it in with the foil to add a means of grounding the shield. I then wrapped the foil/ground wire with electrical tape so it wouldn't short anything out. I attached the ground wire to the upper corner screw holding in the PCB. Sorry I didn't take any pictures since I wasn't too confident this would work. I put the JB back together and to my amazement I was not able to get any faults. I've only power cycled a dozen times and plugged in a few since so I'm not 100% certain it's fixed but it sure seems so.
I can't fault Emotorwerks too much since I strayed from their design a bit. They could have handled tech support a bit better than simply ignoring me when their initial suggestions didn't remedy the problem.
I went with a JuiceBox because of price and I wanted a method to see how many KWh go into the car when charging. The shortcoming is that it resets to 0 every time the EVSE turns the power going to the car off and then on again. So if the car charges and I have a go time set, I better get to the car before the go time if I want to see how many KWh it took to charge the car. The ability to manually set the max charge rate is also a nice bonus.
