Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Get J1772 adapter to pop charge port door

This site may earn commission on affiliate links.
I just ran an experiment on my iPhone, and with the UMC plugged into the car, I found no way to get the "open charge port" button to display. In its place was either "stop charging" or "start charging". In order to get the UMC out, I had to unlock the car, but that can also be done from the mobile app, on a different screen.

"Open Charge Port" is only displayed on the remote app when nothing is connected to it.. Why would you need to open it, if you already have a cable plugged in?

When I disconnect my J-1772 handle from the adapter when done, the car then releases the locking pin and you can remove the J-1772 adapter from the charge port, no need to go into the car.

BTW, with the UMC you should be able to press the button on the handle to release the locking pin.
 
I believe any UMC can open any Model S charge port door, in addition to any HPWC and any supercharger connector.

I can confirm that. When I took delivery of my car while my DS was showing me how to open the charging door it popped open the charging door on a coworkers Model S I had parked behind, perpendicular, to get close to the power outlet. It was a bit of surprise to us both.
 
I spent a while poking through the signal during the burst. It looks like it's square wave at 1kHz, so pretty simple (assuming I'm not filtering away interesting information again).

There are 13 pretty clear sync pulses at the beginning, then a 1 clock (1ms) pause, then the data starts. The data appears to be manchester encoded, but not entirely consistent. For the first 50 bits or so, there is a data transition matching with where the rising clock would be every cycle, which makes sense. But then, the signal transitions appear to move half a clock to the falling edge (there is a transition every 1ms, but now aligned 0.5 off from before).

At that point, I decided to stop trying to parse the data and just try to mimic it. I extracted out the timings and generated AVR code from that which passed my parsing code, but still doesn't open the charge port. I suspect the problem is either with RF noise (I'm getting a smear on the waterfall diagram rather than a tight signal), signal strength (first pulses are a lot stronger), or timing (delays on the AVR are running a little long)

just curious, any progress?
 
I fiddled around with the transmitter electronics a bit, but I couldn't seem to suppress the amplitude variation that I think is at least part of the problem. My capacitor selection is pretty slim at the moment, though, so I may just need a bigger one than I have at the moment. Why do I always start these projects just after the stores where I can get parts close for the day? :wink:

Since that part's on hold, I took another look at decoding the protocol and I have that basically figured out now. It is in fact 50 bits of Manchester-encoded data on the rising clock from the sync train, followed by 50 bits on the falling sync clock, then 50 bits on the rising sync clock. Conveniently enough, all three messages have the same data, which seems like a good confirmation:

(1)1101101011010101010110010011000111001001011110011(0)

I'm not 100% sure about the parenthesized numbers since my decoder is a little sloppy on the transitions. Also, without knowing which Manchester convention they're using, it could be the bit inverse as well (though that only matters if you want to try to assign semantic meaning to the bits).

In practice, all this means is that if I can ever get the RF part of the electronics working that the code running on the microcontroller will be a bit more elegant, so not very important, but fun nonetheless :smile:
 
I fiddled around with the transmitter electronics a bit, but I couldn't seem to suppress the amplitude variation that I think is at least part of the problem. My capacitor selection is pretty slim at the moment, though, so I may just need a bigger one than I have at the moment. Why do I always start these projects just after the stores where I can get parts close for the day? :wink:

Since that part's on hold, I took another look at decoding the protocol and I have that basically figured out now. It is in fact 50 bits of Manchester-encoded data on the rising clock from the sync train, followed by 50 bits on the falling sync clock, then 50 bits on the rising sync clock. Conveniently enough, all three messages have the same data, which seems like a good confirmation:

(1)1101101011010101010110010011000111001001011110011(0)

I'm not 100% sure about the parenthesized numbers since my decoder is a little sloppy on the transitions. Also, without knowing which Manchester convention they're using, it could be the bit inverse as well (though that only matters if you want to try to assign semantic meaning to the bits).

In practice, all this means is that if I can ever get the RF part of the electronics working that the code running on the microcontroller will be a bit more elegant, so not very important, but fun nonetheless :smile:

I hooked up AppleTV once. It worked fine, which was good.
 
Any more progress on getting the RF transmitter to pop open the charger port door?

I spent some more time on the electronics side (putting on and removing capacitors, etc), but no luck yet. I'm going to try a microcontroller with a more stable clock next in case it's timing-related, but I'm starting to suspect that the RF parts of this may just be past my expertise in that area.
 
Press and hold J1772 button and pull out the J1772 handle with adapter as one unit.

To remove the adapter/J1772 from the car while charging or not, here's what you do... Simply press and hold the J1772 button which stops charging and unlocks everything (charging port changes from green to light blue). THEN with your other hand underneath everything, grab the adapter and J1772 and pull it out as ONE UNIT (while still pressing the J1772 button with the other hand). It's stupid, but it works. If you release the J1772 button, it locks back up and you have to time it just right to pull it out. I verified this with my Tesla Delivery Specialist so I thought everyone knew.

This technique negates the need to go in the car (or app) to force stop charging to unlock the adapter.

Opening the port to start charging can be done in the car or with the app.

I'd prefer if Tesla had a special click on the key fob to open the charge port (I'd prefer press and hold on the rear hatch button if Tesla software developers are reading).

I read through this thread because I was looking for a better way, but I guess, unfortunately its my way or NOT the highway.
 
I can confirm that. When I took delivery of my car while my DS was showing me how to open the charging door it popped open the charging door on a coworkers Model S I had parked behind, perpendicular, to get close to the power outlet. It was a bit of surprise to us both.

ditto. When I picked up my Model S from the service center, when the DS was showing me how to charge, he said that every single Model S uses the same frequency for the charge port, and then he showed me that when he clicked the open button on the UMC, every single other Model S in the lot's charge port opened up. lol. kinda of lame IMO knowing that any other Model S owner out there can unlock my charge port if they felt like it.
 
You are exactly right about the charging port lock ITSEL - have a look at the thread on my converter lock where I talk about that exact thing. The button electronically tells the charger to stop, in readiness for removal of the charger handle.

- If you are nearby with the fob, the port unlocks when the charger stops.
- If you are not nearby, the port does NOT unlock when the charger stops (whether from the charge ending or from the button being pushed)

Because of this effect, you can protect your session from disconnection. That thread talks about it.
 
A key fob could be built out of a small processor (ATMEL AVR TINY, for example), a small battery and one of these transmitters: RF Link Transmitter - 315MHz - SparkFun Electronics.
Maybe it could be integrated in the J1772 lock: CapturePro CP1 | Power12 Electric Vehicle Aftermarket

I'd been imagining it could even potentially be integrated into the adapter and automatically power up using the power from the sense pin.

I really feel like my transmit code is close (my SDR decoder decodes equivalently for the real adapter and my transmitter), but I haven't had a chance to debug it more in a while and I'm not sure when I'll have time again. Most of the information is in this thread, but if anyone wants more detailed code/data to give it a whirl on their own, I can certainly package that up.
 
Had a look at a EU charge cable, and the transmitter in the charge plug.
I had a 433,92Mhz receiver laying around (Aurel RX-MID), and connected it it to a PSU and scope.
Used pin9 of the module (data out), nothing more.

scope_3.png


When pressing the button on the charge plug, the above data is repeated at least 5 times.
I haven't looked that closely yet, but it doesn't look like a rolling code.
 
Had a look at a EU charge cable, and the transmitter in the charge plug.
I had a 433,92Mhz receiver laying around (Aurel RX-MID), and connected it it to a PSU and scope.
Used pin9 of the module (data out), nothing more.

View attachment 29789

When pressing the button on the charge plug, the above data is repeated at least 5 times.
I haven't looked that closely yet, but it doesn't look like a rolling code.

Do you think this can somehow be translated into a string of 1s and 0s? Forgive my RF ignorance. :)
 
When pressing the button on the charge plug, the above data is repeated at least 5 times.
I haven't looked that closely yet, but it doesn't look like a rolling code.

It is not a rolling code because any charger needs to be able to open the door - Superchargers, someone else's UMC, etc. It's well-known from some of the gatherings that a Supercharger button press will open every charge door within 20 parking spaces. :)

As a result, it should be a relatively simple transmitter project. Others have used the transmitter module from torn up/broken UMC's.
 
Had a look at a EU charge cable, and the transmitter in the charge plug.
I had a 433,92Mhz receiver laying around (Aurel RX-MID), and connected it it to a PSU and scope.
Used pin9 of the module (data out), nothing more.

View attachment 29789

When pressing the button on the charge plug, the above data is repeated at least 5 times.
I haven't looked that closely yet, but it doesn't look like a rolling code.

That's a very familiar looking plot - you can see the sync pulses at the beginning and I think I see one of the repeats of the manchester-encoded data. This is what I tried just sending back out to a basic transmitter, but was never quite able to get to work (the resulting RF looked "dirty" on my SDR, but I don't know enough RF to know how to fix it). If you press the buttons on the key, you'll see identically formatted data.

From my analysis, I was pretty sure that all of the control signals are fixed in this format on this frequency and the authentication was handled by a separate system that I didn't really look into (some TI chip that's intended for automotive authentication that I saw in the pictures of the fob on the FCC site).

I really think that if you have enough skill with RF to make a simple transmitter module actually output a clean signal with that OOK pattern, it should "just work". The posts of mine that have the best data are Get J1772 adapter to pop charge port door - Page 4 and Get J1772 adapter to pop charge port door - Page 5 though most of that should be unnecessary if you just re-transmit what you're showing on your scope.