Another option may be to try to characterise the charging curve of each EV to discriminate which one is charging. Not sure if it is something that can be distinguished with slow AC charging
Or this is another option to integrate with Home Assistant: https://smartcar.com/
Hi
I was curious to understand which method is more popular [1], which are the main differences (advantages/disadvantages) [2] and which are the main limitations [3] of the following means to control the Wallbox Pulsar Plus through Home Assistant?
A. OCPP: https://github.com/lbbrhzn/ocpp
B. Portal API (Reverse engineering): GitHub - hesselonline/wallbox
Also, I am curious to understand what would make you chose Wallbox instead of any other EV charger provider [4]
I am worried if Wallbox is providing a proper support to integrate Wallbox chargers with your existing systems and this is why I am asking 
Iâll start the ball rolling as I have tried both integrations, only deciding to move away from OCPP to Hesselâs integration within the past month. The supposed pros & cons of OCPP are as followsâŚ
Now to the pros & cons of Hesselâs Wallbox Portal APIâŚ
Now to the reasons why I chose Wallbox Pulsar Plus as opposed to any other charger. In late 2019 when I was looking for a charger, I considered the followingâŚ
There are three main reasons I chose WallboxâŚ
So, after 3 years, do I regret my choice?
Now, thanks in large part to Hessel for all his brilliant work on this integration, reverse-engineering the Wallbox Portal API, I am pretty happy with what I can do to control my charger. There is still one thing on my wish listâŚ
Oriol - I expect that Wallbox doesnât really like to support this âunofficialâ means of access through a reverse-engineered API. However, it provides much better results than OCPP, so I hope you donât decide to lock out Hesselâs integration without any notice. That would be very disappointing indeed!
Interesting idea. Now of course I want to know what car is connected at the start of the session, so it would have to detect the difference in ramp up. That again would require the charging power reporting to work, which still does not âŚ
I also suspect the ramp up for the cars to be very close and that together with a 30 second update on the power may make it hard.
I choose the Wallbox (in order of importance) for:
As @GrantK I have had a lot of WiFi connectivity issues in the beginning, that seems to be solved.
I find the app annoying to use (but I essentially donât use it, so not important).
The Homeassistant implementation is key to the user experience (many, many Kudos for @hesselonline).
I must say, the Wallbox has never let me down charging a car!
Hi @GrantK
That was an exhaustive and excellent contribution. Thanks for sharing. I will share your notes internally so that the different actors can act on them.
Regarding this:
" I expect that Wallbox doesnât really like to support this âunofficialâ means of access through a reverse-engineered API. However, it provides much better results than OCPP, so I hope you donât decide to lock out Hesselâs integration without any notice. That would be very disappointing indeed!"
â my last aim would be that this happened. I truly believe in the idea to build shared value through more open systems and easy integrations with other providers. What I cannot promise is stability of this unofficial means as we continuously update our portal so some âendpointsâ (or however they are called) you are using may change. This is why we are exploring to launch a more official API or improve the OCPP means so your contribution will for sure help on this analysis.
Thanks
Last weeks contact about the Charging Power bug is that itâs not a breaking issue and priority is set to low, which means that resolving this bug can take months or will never be fixed (based on my experience of low priority bugs)
@Oriol_FP Maybe you can pool some strings internally to get it fixed sooner
Thanks,
Rien
@RienduPre - In your previous post hereâŚ
You said this bug was âHigh Priorityâ.
Now you say itâs âLow Priorityâ.
Which one is it?
They said it was high priority, but since itâs not a breaking âchargerâ issue itâs now downgraded to low
Oh, very sorry to hear that. It probably means a very long wait as you say. I cannot understand why it is so difficult to fix. If it works on Bluetooth, why not WiFi?
Hereâs my Charging Power chart from Shelly EM. It needs a little scaling to give the correct figures as itâs from another phase 120 degrees away from the voltage signal, but the raw data is there, and itâs updated every second. Maybe Iâll replace the Wallbox data with Shelly EM data and give up waiting.
Hi @RienduPre
I donât have the bigger picture neither the influence to ensure a greater prioritization of this issue. I just added your comments and complaints to the related open issue in the service desk.
For me it is not a high priority anymore either ⌠since it became obvious that even if the reporting works the number that is reported is very unreliable anyway (6.4 kW is reported as 7.3 kW, that is not even in the ball park).
Indeed the Wallbox documentation for the MID option spells out that that is to achieve a âmore precise way of measuring energy consumptionâ. This of course requires a second power meter (first for Power Boost function) on the RS bus (at a cost). But then any power meter on the charger connection directly connected to Home Assistant will do the job (@GrantK :-)).
Has anyone successfully connected Home Assistant to the RS485 bus of the Wallbox Power Boost meter to read grid voltage and current. Who is the bus master in this setup, the Wallbox or the Carlo Gavazzi EM? If I get that to work I can use the power meter I now use for the grid for the charger and do not even need more h/w (only need to move from main distribution to garage distribution box).
What I do not understand in this picture is that the âadded energyâ reading seems to be exact. How is that possible if the power reading is so far off?
"What I do not understand in this picture is that the âadded energyâ reading seems to be exact. How is that possible if the power reading is so far off?"
I wouldnât say the âAdded Energyâ reading is exact. Here is an example comparing the Wallbox session data with readings from my utility meterâŚ
Wallbox 6.342 kWh
Utility Meter 8.02kWh
Thatâs a 26% discrepancy, not even close! The voltage sensor in my Pulsar Plus reads 6.3% low.
Itâs pretty obvious that Wallbox donât calibrate their sensors. As we discussed before, your readings are too high and mine are too low. Maybe weâre expecting too much, but I would have thought for something which costs around 1000 Euros, the accuracy could be better!
@GrantK the utility meter measures like a 20-40W of self-consumption (RPi, halo, etcâŚ) of the charger (when charging may be more, not sure) which the internal meter of the charger is not considering. With that fact, will the numbers become reasonable? For how long was charging in that session? Furthermore, not sure if you have more loads different than the charger (protections consume?) below that utility meter that may be creating the difference
@Oriol_FP - good questions, and here I have the answersâŚ
My Pulsar Plus was using around 3800 Watts for 2 hours and 5 minutes. This is the maximum possible AC charging rate for my Nissan Leaf. If we subtract 40W to allow for internal usage by the Pulsar Plus, it makes a difference of just over 1%; not enough to make any significant difference.
When itâs not running, Pulsar Plus consumes between 4 and 6W according to my Shelly EM, which is monitoring that circuit. During the 2 hour period mentioned above, Pulsar Plus was the only thing running on that particular circuit, which has its own meter, separate to the rest of the house. My utility meter shows energy consumption in 30 minute blocks, so itâs easy for me to calculate the exact usage.
Hereâs another significant thing to considerâŚ
When I originally installed my charger, it was a Wallbox Pulsar, not the Plus model because that wasnât available at the time. Later, Transnet swapped the original Pulsar for a Pulsar Plus at my request. During the time I was using the original Pulsar charger, I noticed that the readings were much higher than Pulsar Plus, under exactly the same conditions. Here is an excerpt from the email I wrote to our local distributor TransnetâŚ
The kW power delivered vs. Amps relationship is quite different to the original Pulsar charger. By my estimate, the original Pulsar read about 7% high whereas the Pulsar Plus reads about 10% low. So you are actually getting more kWh delivered than you think when using the Pulsar Plus. This isn't a problem; just something users need to be aware of if they are trying to achieve a particular SoC for the battery in their EV.
Another indication that Wallbox donât calibrate their sensors!
Hi @GrantK
Thank you for your analytical reply. Looks that you are closer to the most trustful position so, again, as I am not the electronics role accountable for this I can just promise you that I will forward this to the role I know (or expect) is more affected by this. Thanks for letting us know and for the details.
I was upgraded to a Commander 2 due to charging power issue and WiFi dropping. It has an Ethernet connection.
Here is the charging power now!
Looks like it still has the same problem @Deedzy ?
Iâm guessing the network connection is now solid, but the indicated charging power still fluctuates, dropping to zero many times during a charging session. When approaching 100% itâs normal to notice that, as I can see on your chart at around 2am, but the rest of the time, charging power should stay at the rate you set. Do you agree?
Agree it should stay pretty consistent. If I monitor it with Bluetooth it seems consistent. The charging power doesnât seem to drop to zero as much with this box and yes no network connection dropped at all.
