Speak EV - Electric Car Forums banner

1 - 20 of 57 Posts

·
Registered
Joined
·
72 Posts
Discussion Starter #1
I took some close up photos of the internals of my Pod Point charger today as I was wondering what type of MCU was used with a potential idea of writing some custom software to implement hooking in for charge timing etc. (Not going to happen now).

Thought some people here might like them. The main 2 photos are here.

P1070773-2.JPG


P1070768-2.JPG


With the rest and the same with higher resolution at https://photos.google.com/share/AF1QipNBDkZgPle3RuXa0NJGDh8qGFtl8u212eTisayAyXyrfID1f0yGKbQMTIuk2PPW0w?key=a1AzdzF2Rll0OEw1WnF6STRsbEFQTzlLcGNKSmZB
 

·
Registered
Joined
·
2,861 Posts
Interesting that they, like Chargemaster, use two relays in series despite apparently being of suitable specification to require only one. Must be some safety requirement behind this.
 

·
Registered
Joined
·
72 Posts
Discussion Starter #3
Well I had a further look at my Pod today to try and see a route to hack / modify it easily to enable either remote control or a switched control which I could hook up to another MCU for charge time control.

Usual Disclaimer: Dangerous High Voltage AC! It has also been pointed out to me that the Low Voltage Side might not be double isolated.

No luck yet, however found the following:

The LOCK pins are correct, connecting these together simulates a lock (Or you could wire a lock switch in using these pins), closing these turns the lock on. However if the car is plugged in while locked and then the pod becomes unlocked it doesn't seem to resume automatically and requires re-plugging in.

The RX, TX header is UART running at 38400 baud, Unfortunately this also doesn't result in a lot of information apart from during charging it says the current KW being pulled by the car.

Interestingly the terminal block at the bottom left is for a current sensor as roughly described in the pod point manuals. There's also a temperature sensor header too.
 

·
Registered
Joined
·
153 Posts
The easiest way I can see of doing this would be something a bit like those smart charging cables. It just takes in the CP signal advertised from the existing chargepoint, and regenerates a different one to the vehicle. So, even though the existing protocol controller is advertising 32A is available, the adjusted PWM signal sent to the vehicle might only offer 6A of that. Look up J1172 hydra for an open source thing that does similar man-in-the-middle spoofing of CP signals, only in your case, you don't need quite as much switchgear.
 

·
Registered
Joined
·
72 Posts
Discussion Starter #5
The easiest way I can see of doing this would be something a bit like those smart charging cables. It just takes in the CP signal advertised from the existing chargepoint, and regenerates a different one to the vehicle. So, even though the existing protocol controller is advertising 32A is available, the adjusted PWM signal sent to the vehicle might only offer 6A of that. Look up J1172 hydra for an open source thing that does similar man-in-the-middle spoofing of CP signals, only in your case, you don't need quite as much switchgear.
That could work. I've been following the Hydra as I followed the designer on twitter for a while.
 

·
Registered
Joined
·
91 Posts
I'd just take control of the CP line between charger and vehicle. If this doesn't follow the requirement of IEC61851-1(I think) then the EVSE wont turn on the mains supply to charge the vehicle. Just stop the vehicle affecting CP and it wont be able to charge. Once you allow the vehicle to affect CP then the charge process can be initiated. Simplistically, put a switch under your control in the CP line between vehicle and EVSE.
 

·
Registered
Joined
·
33 Posts
I have done the reverse engineering a while ago and patched OpenEVSE to limit the charge current to excess PV. You might find this github repository gimx/open_evse and the attached picture useful. It basically takes the CP signal out before the +-12V driver and feeds the modified signal back in again. I have done this on the old pod point layout as well. Btw. all safety features remain functional.
 

Attachments

·
Registered
Joined
·
72 Posts
Discussion Starter #8
I'd just take control of the CP line between charger and vehicle. If this doesn't follow the requirement of IEC61851-1(I think) then the EVSE wont turn on the mains supply to charge the vehicle. Just stop the vehicle affecting CP and it wont be able to charge. Once you allow the vehicle to affect CP then the charge process can be initiated. Simplistically, put a switch under your control in the CP line between vehicle and EVSE.
That might actually be even easier, Just a Relay between the two. I have a box so might try that tomorrow. It's even simpler and then can be a nice tidy setup as I'll get a PCB fabbed with the relay and driver circuitry on.

I have done the reverse engineering a while ago and patched OpenEVSE to limit the charge current to excess PV. You might find this github repository gimx/open_evse and the attached picture useful. It basically takes the CP signal out before the +-12V driver and feeds the modified signal back in again. I have done this on the old pod point layout as well. Btw. all safety features remain functional.
If I do any circuitry it looks pretty easy to replicate and have it go between the pod point and the car. I don't feel like soldering onto my pod point at all as it's still in warranty. I'd also be a bit overkill with using an RPi Zero.

However :O , Old standard wiring colours! Did Pod Point install that?
 

·
Registered
Joined
·
33 Posts
I'd also be a bit overkill with using an RPi Zero.
OpenEVSE is arduino based, it's just an old ArduPilot board that I had laying around. The match in name only hits me now, while typing it :)

Your original post mentions an MCU though
control which I could hook up to another MCU for charge time control.
If you don't want to solder onto the PCB you would only need to add the level shifters. However for a time shifted charging it's enough to have a relay in the CP line that opens the circuit and puts the car side on 12V.

However :O , Old standard wiring colours! Did Pod Point install that?
Make an educated guess.
 

·
Registered
Joined
·
91 Posts
You even have access to the CP signal on the Podpoint PCB - it looks to be on the connector next to the mains output. Try getting the car charging and then remove the CP wire from the connector - car should stop and then re-insert the wire and the car should start up.

Warning: once the charge controller and vehicle have communicated successfully, the charge controller will apply the mains voltage to the connector next to the CP connector. Whilst the CP is at low voltage there is mains close by. You would be wise to bring the CP signal out of the box to a switch and do your test(s) with the PCB covered and no chance of you accidentally touching anything dangerous.

It will be overkill to use a RPi but it depends on how far you want to grow your idea. You could get a mains timer that can throw a relay to control the CP signal. Or use a RPi to drive a relay controlling the CP signal but open up an exciting world using either Wifi or ethernet to add further control and/or monitoring.

Enjoy and stay safe.
 

·
Registered
Joined
·
923 Posts
Be aware that not all EVs behave nicely if the CP is just interrupted during a charge. In theory they should all just stop charging and try to load the CP to the plugged in state, just as they would initially when the CP is sat at a steady +12 VDC. Mind you, I believe that the vehicle OBC should throw an error and remove the load if the CP is taken to the error state (-12 VDC), but in practice I've not seen any response to the -12 VDC error state, and suspect that most vehicles keep the cable fault detection diode in circuit all the time (so they never see the -12 VDC fault state signal). As long as the EVSE responds to an error state by taking the CP to -12 VDC I can't see that this is a problem.

Changing the CP duty cycle on the fly seems fine, but some vehicles seem to like to see the 1 kHz CP signal at least until their OBC shuts off, and may throw a wobbly if the power contactor in the EVSE turns off, leaving the CP running. My testing showed that my old Prius plug-in would randomly throw up charging errors if this happened (it happens during power cuts, as the CP signal used to run on for a very short time until the power supply reservoirs caps discharge). Sometimes it would just reset and start accepting a charge when the power came back on, very occasionally the car OBC would lock out and throw a charging error on the screen in the car, which had to be cleared by turning the car on then off). I fixed this by adding power fail detection, to turn off the CP immediately when a power failure was detected. This seemed to work fine, from the limited testing I did.

I've stuck to making sure that, in as far as is possible, the EVSE follows the protocol carefully, and signals to the OBC that it's shutting down by taking the CP to +12 VDC, before opening the EVSE contactor. This seems to work OK, and hasn't (yet) thrown up any errors, but my testing has been far from exhaustive.

My experience suggests that lots of field testing is needed to make sure an EVSE can handle a wide range of vehicles without problems.
 

·
Registered
Joined
·
72 Posts
Discussion Starter #12
Hmm That is interesting. I'll do some observations and testing. I've had a fair share of charger errors on my Zoe so know what values to check via OBD for states.

If I find that my Zoe doesn't like the CP being disconnected a route I could take is use two relays to begin with. One will first essentially "lock" the charger which I did find cut off the car from charging fine. And then when the charge needs to start it will unlock the charger ,then connect the CP pin.

@Torsten , Mainly RPi as I want the wi-fi connectivity. I could use an ESP or such but I have 20 Pi and it'll then be good to add more too later (Such as possibly interfacing the UART from the Pod Point to record charge rates).
 

·
Registered
Joined
·
72 Posts
Discussion Starter #13
Well got the code done already, just got to wait for the relay module to arrive as I can't find the one I had.

Using a Particle Photon to control the relays, I'll be using 2. One to connect the CP signal and one to control the LOCK on the Pod Point.

To charge It'll turn the relay on for the lock first which will be wired as Normally Closed essentially turning the lock off and initialising the pod point. Then it'll turn on the relay to connect the CP together initialising the charge.

To terminate the charge instead of just disconnecting the CP (As @Jeremy Harris hinted could be an issue) the relay will "lock" the charger first and wait 5-10 seconds for the lock to enable and disconnect the car. Then the CP will be disconnected ready for next use.

By using IFTTT to control the photon I can use the button widget on my phone to override the charger for if I need it to charge in the day, and if we then move from Octopus Go to Octopus Agile can integrate into that to start the same way.

Wiring will likely be Wednesday so will report then.
 

Attachments

·
Registered
Joined
·
1,298 Posts
Excellent write up, looks easy enough.

I agree with the conclusion, IoT enabled chargers are needed. I think your solution is as close to smart charger as possible, only possible improvement is to vary charging rate like Zappi according solar generation.
 

·
Registered
Joined
·
175 Posts
withKLockWritten (2).jpg


here is a version which says K lock level with the con11 double header.
Ryan kindly found this on google when I reached out for help. Thanks Ryan.
my pod point landscape.jpg

here is my pod point early 1.3 version. Mine doesn't have the label and doesn't have a header connected on con6 at the bottom.

I suspect con11 is the key lock (as that is where the writing is level with) but does anyone here own a pod point with a keylock installed and it be a similar variant of this board? Please isolate power to your unit before checking.

All the rest of this project is ready to go (I don't know which is a 12v rail either here as nothing is mark but I can easily wire up a 3 pin plug and use a usb charger to power the particle photon and relays instead.
 

·
Registered
Joined
·
175 Posts
I've a K-Lock version, picture incoming
#
@Chilledbud Ok, but you don't have anything attached to the pins indicated left of the text K-lock. I assume its con11 but i'm not 100% just because on @Ryanteck 's pcb it was the two pins at the bottom of the 4x2 connection.

I sent an email to pod point support to check but I'm doubtful they will help.

Id also like to know what the jumper on the bottom of the green 4x2 is for? mine doesn't have one.
 

·
Registered
Joined
·
175 Posts
got zero response from pod point email. yesterday i tried a jumper on con 11 (which i believe is key lock from other pictures) without car attached to evse, LED didn't change colour.

This morning got braver tried shorting con 11 with a jumper whilst leaf was charging to see definitively whether it could interrupt a charger. Absolutely nothing. I also tried putting the jumper on the bottom of con6 (green connector) but i think this might be tethered vs untethered setup.

Anyway still no joy. I feel that version 1.3 pcb with no mention of K-lock has no key lock and thus cannot be smartened with the hack from @Ryanteck

next steps is to look for later edition pcb on ebay or just sell mine and buy a hydra from Nick Sayer.
 

·
Registered
Joined
·
175 Posts
Correction the jumper in con 6 at the bottom may set the pod point to 7.2kw as stated in another thread on here. It didnt appear to change anything on mine but my car can only take 3.6kw anyway.

May try testing the cp disconnect without the lock since my early variant has no k lock writing on the headers and the ones i suspected do not lock my device. Also the leaf may be a car more tolerant of errors than the zoe ryan has
 
1 - 20 of 57 Posts
Top