Charging from Solar PV panels: iSmart software repeatedly resets Charging Settings and stops charging prematurely

[GadjoDave]

Agghh! That's a real pity, as my friend's workaround sounded like it might do the trick.
Maybe a future MG firmware upgrade might offer the option of setting charging current (or power) in smaller user-selected increments?

I guess another workaround is to cut out the Zappi altogether when I want to charge from PV surplus, and use the granny charger from the domestic 13A socket (which I think works out at around 2.3 kW). My Powerwall shoulld make up any shortfall if the sun goes behind a cloud or the kettle boils...The charging speed will be slower compared than if I was using the Zappi (up to 5 kW); but at least it shouldn't quit prematurely.

I appreciate that EV technology is still in its infancy in this country. Whether it takes 10 h or 20h to charge my car from PV panels is very much a first world problem, and I shouldn't get too hung up on it; but it seems that with just a bit more development and agreement of standards, a system that is periodically unreliable and currently relies on workarounds could work more efficiently and seamlessly...

Slightly off-topic, but might be of interest: I highly recommend two books by American writer Steven M Casey on true tales of design, technology, human error and the law of unexpected consequences: "The Atomic Chef" (Aegean, 2006) and "Set Phasers On Stun" (Aegean, 1993). Out of print, but should be available from libraries, or you might find a second hand copy on line if you're lucky.

 

[shpub]

I doubt it because allowing those different levels start to introduce other issues like does the charger unit support it. Not all do. All that has to be checked. That costs money. Part of the problem is that there is no British Standard User to do all the things that the designers assume would not happen.

I would try your friend's solution anyway but equally if you make more money exporting than an overnight charge would cost, I would just simplify things. Just my opinion though.

It takes years to agree standards as there are many issues involved when you try to get many big corporations to agree when they have patents and other interests that bias the discussions. Companies also do not care if a change allows them to grab more market share at the expense of isolating small fry.

 

[pmknowles]

A couple of things
Very few (if any) AC chargers know the SoC of the car so SoC canot be used as a control method. Setting scheduled charging (either in the app or in the car) can stop the car charging. I now use EVCC to control my Wallbox Pulsar Max - I only have solar panels and a 9.6kVA lithium battery - I start charging at 80% SoC of the lithium battery and stop at 20%

 

[MickeySw]

Very few (if any) AC chargers know the SoC of the car

Correct

so SoC canot be used as a control method.

Wrong, best practice is to set a SoC limit in the car and let the EVSE do the scheduling bit.

 

[pmknowles]

Correct

Wrong, best practice is to set a SoC limit in the car and let the EVSE do the scheduling bit.

I meant the value cannot be used by the charger (or a system controlling the charger). Charging can only be controlled by the car refusing to accept charge. The charger doesn't know if the car is fully charged, is on a schedule or there is a fault - just that it won't accept a charge. Life would be so much easier if the SoC was available to the charger.

 

[MickeySw]

Life would be so much easier if the SoC was available to the charger.

Can you elaborate?

 

[pmknowles]

Can you elaborate?

If the charger can read the SoC it can pass it over OCPP to a control system (or the charger app). I do some complicated maths to estimate 80% to stop charging but, once a month, I need to do a 200 mile trip so I need to charge to 100%. I forgot to change the in-car 80% limit once and (due to 3 defective chargers en route) had a 7 hour journey instead of 4.

 

[MickeySw]

I do some complicated maths to estimate 80% to stop charging but, once a month, I need to do a 200 mile trip so I need to charge to 100%.

What kind of ‘complicated maths’ do you have to do if you have a 80% charge limit set in the car?

We have all been there forgetting to change settings one way or another, so I fail to see the logic how transfer of SoC will prevent that?

 

[Geriatric Joe]

I decided to keep the charger and battery in simple mode so that I just plug it in let it charge and then unplug it. I have the MG four with a Tesla wall charger and decided to give up on Settings using the car or the app on the iPhone.

After all it’s just a battery and a charger.

 

[pmknowles]

What kind of ‘complicated maths’ do you have to do if you have a 80% charge limit set in the car?

We have all been there forgetting to change settings one way or another, so I fail to see the logic how transfer of SoC will prevent that?

Because I no longer rely on the car to stop at 80%. EVCC has an SoC limit but doesn't give an accurate figure from the charger (overstates by about 7.5%) so it's estimate of SoC is out. Because I'm also not able to charge from mains (looped supply) matters are further complicated. My calendar tells the controller when my long trips are and automatically allows it to charge to 100%

 

[swampgal]

I'm having the same issues but I have an EVnex charger with solar diversion. The MG iSmart App won't let me change the scheduled charge settings ("The request timed out. Please try again later" error) and I have to manually restart the charging after each time I use an appliance or the sun goes behind a cloud. It's the MG4/app not the charger that's the problem. But what's the solution? I just sent feedback via the app - maybe if we all do that?

 

[MickeySw]

I'm having the same issues but I have an EVnex charger with solar diversion. The MG iSmart App won't let me change the scheduled charge settings ("The request timed out. Please try again later" error) and I have to manually restart the charging after each time I use an appliance or the sun goes behind a cloud. It's the MG4/app not the charger that's the just sent feedback via the app - maybe if we all do that?

It seems to be an issue with the car itself, OP reported all was fine before the car went to the dealer for an unrelated problem that required some FW updates.

A hard reset, see post #35 might help with the scheduling issue, best to set it on the car itself, don’t use the app.

 

[Dafc1fan]

Equipment:

MG4 Trophy LR
3.8 kW solar panel array
Latest firmware, infotainment, and MGiSmart iPhone app versions*
MGiSmart Charging settings: Target Battery Level 100%. Scheduled Charging: Off. Scheduled Battery Heating: Off. All settings Saved.
Zappi tethered home charger with latest versions of firmware and MyEnergi iPhone app*
Tesla Powerwall v1 domestic battery with latest firmware*. 100% charged.
Octopus "Intelligent Octopus Go" tariff (cheap rate electricity 23:30 - 05:30)

* as of today (1/6/24)

When charging from solar panels using the Zappi's Eco+ mode*, the MG4 Trophy initially accepts the charge, supplemented by top-up from the Powerwall as required.
If there is inadequate current available (eg. sun gone behind a cloud, or increased domestic usage), charging would pause with the Zappi and MyEnergi app displaying the message "Waiting for Surplus". Then, when there was more output from the panels available, charging would resume automatically, and continue until the preset Target Battery level (eg. 100%) was reached.

*(The MyEnergi app is set to Eco+ mode with the "leaf" slider setting on 1% - ie. draw current from Powerwall or grid as required if solar panel output is insufficient),

This is the desired behaviour, and this is how my system worked until I got the most recent upgrade of MG firmware and MGiSmart iPhone app last week.

Since then, if the output from the solar panels dips transiently (eg. clouds moving across the sun; using electricity for domestic appliances such as boiling a kettle or using the electric stove), the MGiSmart app reports "Charging Completed", and the MyEnergi app reports "Charging Delayed" indicating that the MG is no longer accepting charge.

When the sun comes out again or the kettle has finished boiling, the MG subsequently refuses to resume charging, even though the charge in the battery is nowhere near the Target Battery Level, and any surplus electricity from the solar panels is diverted to be exported to the grid.

When I look at the Charging Settings on the MGiSmart App, it seems that they have spontaneously reset to: Scheduled Charging: On (Start time 22:00 End time: 06:00). Scheduled Battery Heating: On.
This is why the MG is not accepting charge, even when surplus is available.
I correct these settings and press Save again.
I then try to Start charging from the MGiSmart App.
The app sometimes requires me to re-enter my 4 digit PIN; sometimes not.
However it is unable to Start charging, and the message "Please Insert The Charging Plug" is displayed (despite the charging plug remaining inserted throughout).

The only way to resume charging from the solar panels is to physically get into the car and power it up.
I touch the "battery" graphic on the home screen, and go to Charging Settings where I find that these have all been reset to the same settings as the iSmart App (Scheduled Charging: On, Scheduled Battery Heating: On).
Again, I have to reset these to "Off" and Save them.
However, unlike the MGiSmart app, the MG firmware recognises that the charging cable is still connected.
I then have to press "Start charging" on the Charge Settings Screen, and charging resumes.

This sequence of events happens repeatedly when I am trying to charge the MG from my solar panels: basically every time the sun goes behind a cloud or a domestic appliance draws current for more than a minute or two.
Each time, I have to reset the charge settings on my iPhone app, then get into the car and boot up the software and reset the charge settings before charging will resume.

The fact that charging from the solar panels all worked fine until the latest MGiSmart App update; and that the MGiSmart Charging Settings ignore the saved values and keep resetting to "Scheduled Charging: On" every time there is a break in flow of current to the car during the charging cycle all seems to suggest that the cause is buggy software in the latest update of the MGiSmart App.

I'd be grateful for any shared experiences, advice or workarounds.

And if anyone from MG is reading this:
You have an potentially excellent EV which is being seriously let down by ongoing buggy and badly written software to the point which makes for an dreadful user experience.
I would currently hesitate to buy another MG or recommend MG to friends and colleagues.
A software fix is simple and cheap to implement. Get it sorted!

Wow - I was reading your problem & the title was solution, so mistakenly thought this would be an answer to my prayers, but no, just a description of my exact same problem with A Zappi. I just stop using eco++ I could sometimes/ mostly get the charge restarted by using the Zappi Ap change to Eco then insert the pin to start charging on my Ap for my MG4 SR, but it’s not as awful as you described in my view. You are right though, another bug needs fixing.

 

[North Wales Nick]

I get the same problem with the Zappi switching to 'Charge Completed' on a day when the sun keeps going in, instead of 'Waiting for surplus'. BUT note that I am NOT using the iSmart App at all (that's another story - I couldn't get it to bind and nor could my Dealer - so I just gave up). It's very frustrating. If the problem is with the software in the car, as the above discussion suggests, do we just wait until MG sort out their software???

I get the same problem, with the Zappi switching to 'Charging complete' on a day when the sun keeps going in. BUT I am NOT using the MGiSmart app. So do we just have to wait until MG sort out their software??!

 

[MickeySw]

I get the same problem, with the Zappi switching to 'Charging complete' on a day when the sun keeps going in. BUT I am NOT using the MGiSmart app. So do we just have to wait until MG sort out their software??!

It appears that the MG EV FW in general (not just MG4s) doesn’t like too many (the figure 7 was mentioned) stop/start interruptions during a PV charging session therefore the ultimate fix will be a SW update.

But that means MG will have to acknowledge the issue first, come up with a fix and patch the cars. Given the (relatively) low number of owners affected, I doubt this will be a priority.

Maybe best to look at workarounds:

1. if your setup allows, use grid or battery to supplement min charge rate

2. don’t use the ‘wait for surplus’ setting on sunny days, just plug in the car with the granny charger or set the EVSE to it’s lowest current setting

3. export all your PV yield to the grid and charge the car at low tariffs at night.

4. any other ideas?

 

[North Wales Nick]

Thanks for the suggestions MickySw. In our case, we are trying to maximize the use of surplus PV power during the day to charge the car because a) we don't have a battery and b) we don't get paid for power exported to the grid. We could use our granny cable I suppose, but that is rather frustrating when we have an expensive charger (Zappi) and a nearly new EV - which between them should cope! A comment from someone earlier in this thread who said he/she is resigned to not bothering on sunny/cloudy days and only trying to use surplus PV on a good sunny day makes sense to me - but it is frustrating once again that the technology can't cope. How can we be sure that MG are aware of the issue (in the hope that they will get around to fixing it)?

 

[Coulomb]

4. any other ideas?

For the technically minded: perhaps try open circuiting the control pilot signal for a second or two. It may be that the car has a low power circuit to detect pulling out and in of the charging plug, via the low voltage control pilot, rather than the 240VAC. This could be done via a small relay with normally closed contacts. For extra points, detect the pilot changing state. Though it could well be that the EVSE (AC "charger") already does this, in which case it won't do any good. It depends if the EVSE signals "no charging" via +12V or by open circuiting the control pilot pin. Maybe open circuit the proximity pilot pin instead, or as well as the control pilot pin.

The other crazy thought is a solenoid to automatically "press" the close door button on a spare key fob when you need to wake the car. Or maybe use your main key fob, and make a habit of storing in the "electronic thumb" gadget when charging.

It's a shame we have to get so creative to work around these glitches.

 

[MickeySw]

How can we be sure that MG are aware of the issue (in the hope that they will get around to fixing it)?

Start by contacting your dealer, not that I’m optimistic they do much about it but if everyone in the same boat does that MG can’t use the ‘nobody told us’ excuse..

It depends if the EVSE signals "no charging" via +12V or by open circuiting the control pilot pin. Maybe open circuit the proximity pilot pin instead, or as well as the control pilot pin.

Good thinking, if an EVSE treats a ‘pause charge’ due to low PV differently to ‘finished for the day’ i.e opening CP instead of applying +12V to it (or the opposite) then it may well be possible to fix this on the EVSE side.

Almost afraid to ask, what is Zappi technical support like?

 

[North Wales Nick]

I remember I had a satisfactory on-line exchange with Zappi tech support early on.

Anyway, apart from the baffling discussion above (for the non 'technically minded'!) I wanted to double check, before I contact the dealer - how sure are we that the problem is with the car and not with the Zappi charger, or indeed the charger / car interface?

 

[MickeySw]

I remember I had a satisfactory on-line exchange with Zappi tech support early on.

Anyway, apart from the baffling discussion above (for the non 'technically minded'!) I wanted to double check, before I contact the dealer - how sure are we that the problem is with the car and not with the Zappi charger, or indeed the charger / car interface?

It’s more circumstantial evidence than clear proof:

- the same setup seems to work fine with other EV brands

- OP’s MG4 worked fine until it came back from the dealer after FW update

- at least one MG5 owner has reported the same problem

- a Zappi is the go to EVSE for PV charging, must be for a reason

- many MG users have abandoned PV charging due to problems and use night time tariffs for this purpose instead (financial incentive)

If you know someone with an EV, invite them to your house for a ‘free charging session’ and video it, then do the same with your car.

Show the difference to the dealer (hope there is one…)