How can we help you today?

Start a new topic
Discussions OpenEVSE Forums Tips and Tricks

Adding a timer to a generic charger?

 I have a ClipperCreek charger, but I need to add a timer. It appears that you can cause charging to stop by simply disconnecting the "control pilot" wire. Is this correct?


My plan is to add a jack to the charger to reroute the control pilot to an external enclosure and place a signal relay in that enclosure to  open the control pilot when I want to disable charging. The relay will be under control of a little computer (e.g., Raspberry Pi zero W)


Yes, I will void the warranty.

Yes, I could kill myself if I'm not careful.

Yes, this is NOT about an openEVSE product (but it may still be relevant.)


If it is infeasible to modify a generic charger in this way, I will need to buy the OpenEVSE kit and cannibalize the old charger fro its cables, Which seems like a waste.


14 CommentsSorted by Oldest First

Disconnecting is not the best way to implement a timer but it will work.


The right way to implement a timer is hold the pilot at +12v or +9v when a vehicle is connected. As soon as the Station is Ready to begin the session then the 1000hz pilot signal is activated.


Cannibalizing a dumb station with OpenEVSE is never a waste, you would add current measurement, kwh display, adjustable current, temperature monitoring and throttling, upgrade firmware...etc

2 people like this
I am curious about disconnecting the pilot. If you push the button down on the charger jack while it is in the car this disconnects the pilot right? On my 2016 volt an alarm goes off to tell you that the plug is still in the car. instead of disconnecting the pilot could you reintroduce the +12 v and put the car back into a ready state which would shit off the current flow without having the alarm going off. i am looking for a dumb charger to cannabalize  or buy an open evse and utilize it if it can do what I want with it.

 

I am not an expert, which is why I asked in the first place. There are five pins on the SAE J1772 connector: two hot wires, one ground wire, a "control pilot", and a "proximity".  The pin I am referring to is the "control pilot." The pin you are referring to is "proximity." I was surprised to discover that the cable from the SAE J1772 connector back to the controller only has four wires, at least on my charger. The "proximity" wire terminates in a passive circuit within the connector itself.  Look carefully at the schematic on the Wikipedia article at
     https://en.wikipedia.org/wiki/SAE_J1772
The state of the system with the control pilot disconnected will be seen by the car as equivalent (I hope) to being plugged in to a charger that is itself disconnected from its electrical circuit. Reconnecting the control pilot will (I hope) appear to the car as if the charger has been reconnected to its electrical circuit.

If I did not already own a dumb charger, I would certainly purchase an OpenEVCS instead of messing around with another dumb charger, and In fact I will buy one in a heartbeat if my experiment fails.

 

If you push down the button the proximity circuit changes resistance from 150 ohms to 480 ohms. It does not disconnect the pilot. 


Just disconnecting the pilot is a panic scenario, it could cause the relays to open under load as the vehicle has not received the proximity command which requires the vehicle to stop pulling current.


Changing the pilot from 1000 khz back to a steady 12v DC is the correct way to notify the EV that the charging station is no longer ready to supply power.



So, is this correct? 


Pilot wire to switch common-

12V DC to one pole of switch-

Controller to other pole of switch-


First switch position connects pilot from car to controller

Second switch position connects pilot from car to 12V DC

image

Matt, It is not correct. The pilot is part of a voltage divider network so feeding 12v would not work. If you feed 12v - 1000 Ohm resistor - Vehicle, the Vehicle would lock into the Connected State with EVSE Not Ready.


Disconnecting the pilot from the charging station would put the charging station back in the "Vehicle Not Connected" state while the vehicle would be in another state. State changes would not be registered by the other end.



Thanks. That helps a lot.


Matt

So just to make sure I'm clear, is the attached diagram what I want to do?


Here's me story. I have 2 OpenEVSE chargers running on one 40amp circuit. Until recently, I had a Leaf and a Rav4. Neither took more than 4 hours to charge so I would have one start at 10pm when the rates went down, and the other start at 2am. There was never a problem with them tripping the breaker since they would never run at the same time.


Recently I bought a Bolt, so I now have the potential to have an overlap of my chargers. I would like to put a current sensor on the line that feeds the Bolt charger and when it senses current, have it put the other charger in "wait" mode until the Bolt is finished.


To me, the easiest way to do this is to put the pilot in "wait" with a relay, triggered by the current sensor. Am I on the right track or do you see an easier way?


Thanks for your help.


Matt

IMG0484.jpg
IMG0484.jpg
(2.22 MB)

I would do it with WiFi, MQTT and Node-Red.


Setup each station to publish to MQTT server.

Setup a flow in Node-Red to whatever logic you wish. Share available current 50/50... 60/40... 70/30... Finish 1 start the other.

Output available watts for Station 1 and Station 2 to a MQTT Topics.

Subscribe to each topic on each station using the Divert feature.


Each station will adjust current based on the MQTT feed. 

This is way beyond me but I sent it to my son to see if he can make sense out of it.  I'll let you know if he thinks he can help me with this.



So if I am unable to do the MQTT thing, is my design for putting the pilot on "wait" a workable option? 

It might possibly  work, but it would likely fault out the vehicle. The State transition would not be correct. With a relay the pilot would go to 0v as the relay is in-transit between PWM and DC and DC back to PWM.


Another way would be to send a serial command to the controller to Sleep or Enable. Whatever device you are planning to monitor the first station could send $FE^27 to enable the Station or $FS^31 to put the station to sleep. 

Thanks. We're still working on the MQTT angle. Hopefully that will be the solution.

My son is working on the MQTT angle and he's wondering if there is a way to update the "state" more frequently. Is there a way to have it update every 5 seconds?

Login or Signup to post a comment
More topics in Tips and Tricks
See all 23 topics