MG ZS 44.5 kWh battery

[Zyga76]

Hi
I'm starting up 44.5 kWh MG batteries. The battery powers up and the contactors engage correctly.
Unfortunately, there's a problem with the display of the total voltage relative to the individual cells.
The program incorrectly displays the cell count.
I'm attaching photos of the battery and the CAN log readings.
Please analyze this battery so I can test it and share your observations.

Best regards,
Pawel

[Zyga76]

canlog_00-06-38.txt

[jonny5532]

There's a fix that should be in the next release, or you can try these images:

#2141 (reply in thread)

(Let me know if it works!)

[jonny5532]

Oh I misread your question - currently it is hardcoded to 96 cells (this is fairly easily fixed, will do).

[Zyga76]

Hi
I downloaded the version from the link and it still shows 96 cells.

[Zyga76]

Hi
When will the new version of BE be available?

[Zyga76]

Hi, I downloaded the version from the link provided and it still shows 96 cells. When will the new version be released? czw., 11 cze 2026, 11:46 użytkownik jonny5532 ***@***.***> napisał:

…

Oh I misread your question - currently it is hardcoded to 96 cells (this is fairly easily fixed, will do). — Reply to this email directly, view it on GitHub <#2444?email_source=notifications&email_token=BMNUKKMH2LITMUDMPNFOZCL47J5W3A5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZSGYYTSMBQUZZGKYLTN5XKMYLVORUG64VFMV3GK3TUVRTG633UMVZF6Y3MNFRWW#discussioncomment-17261900>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BMNUKKKW26GH57MJI5QYVOL47J5W3AVCNFSNUABIKJSXA33TNF2G64TZHM3DAMZYGYYDEOJVHNCGS43DOVZXG2LPNY5TCMBSGM2TQMRTUF3AE> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/BMNUKKP2QFIF5JEC4SU3RNL47J5W3A5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZSGYYTSMBQUZZGKYLTN5XKMYLVORUG64VFMV3GK3TUVJTG633UMVZF62LPOM> and Android <https://github.com/notifications/mobile/android/BMNUKKKL3Y3GJMLWO3WO7L347J5W3A5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZSGYYTSMBQUZZGKYLTN5XKMYLVORUG64VFMV3GK3TUVZTG633UMVZF6YLOMRZG62LE>. Download it today! You are receiving this because you authored the thread.Message ID: <dalathegreat/Battery-Emulator/repo-discussions/2444/comments/17261900@ github.com>

[Zyga76]

Hello,
In the new update, the battery still shows 96 cells, but it actually has 108 cells.

The capacity after powering off and restarting changes from 44.5kWh to 52.5kWh.

[Zyga76]

Hello,
The MG ZS battery parameters are not working correctly. Instead of 108 cells, it still shows 96 cells.
I'm attaching the CAN logger readings and a list of errors.
canlog_00-02-05 MG ZS 44.5kWh.txt

[Zyga76]

Hi.
Can I get some help?
Please.

[ibrahimshweiki1995-tech]

Try to up load 10.10.dep
#2141 (reply in thread)

[ibrahimshweiki1995-tech]

If it not work this man “jonny5532 “awesome he will solve it just wait your problem I think not big

[Zyga76]

Cześć
Podesłane linki z wersją 10.10.dev też nie działają.
Problem jest w tym że maja bateria ma 108 ogniw i 44.5kWh. Jest z auta MG ZS.
Bardzo proszę o pomoc.

[jonny5532]

I have updated this PR: #1941 which has a 'MG Gen1' setting (not the MG5 one) which should autodetect some batteries and adjust the cell count.

If you could:

• OTA to that PR image
• Choose 'MG Gen1' as battery
• Check it works (contactors close, voltage/current looks sensible)
• See if the autodetection worked (turn on Log to webserver, and look in the Log panel)
• Copy/paste the values on More battery info and add them to a comment here
• Also see if Read DTC works (under More battery info)

If it doesn't autodetect correctly I can fix that as long as I have the More battery info page data.

[Zyga76]

[Zyga76]

Hi,
It detects the cell count correctly.
It doesn't detect the capacity and I can't manually change it.
The charge and discharge currents are also too low.

[Zyga76]

This is version 10.10 dev

[jonny5532]

I've updated the PR here (#1941) or here's the 2CAN image:
mg_2can_1.bin.zip

It now doesn't overwrite the capacity you enter, so you can change it yourself.

I don't think you were using the firmware from that PR? It now has 11kW charge/discharge limits, whereas those screenshots are showing 4kW and 3kW.

[Zyga76]

Hi jonny5532,
There is an issue when discharging the battery down to the low end. At a voltage of 3.52V, the BE prevents further discharge.
What could be the cause?
I would really appreciate your help.

[jonny5532]

I've increased the power limit to 15kW for the ZS/MG5 packs.

The power limiting you are seeing is the 10% min-SoC limit, with tapering below 15%. I have changed this to 5% min-SoC, with tapering below 10%.

I'm not sure if we actually need a min-SoC limit, the car will let you drive it down to 0%. However it does avoid unattended systems draining down to 0% and then overdischarging.

mg_2can_2.bin.zip

[ibrahimshweiki1995-tech]

Hi jonny is this code work with mg marvel or ZS 69k discharge limit up to 15k ?
and is all edit about mg5 in last update 10.12.0 ?

[jonny5532]

Yes it should work with the MG5/Marvel/ZS, use the 'MG Gen1' setting.

[Zyga76]

Hello Jonny5532,
First of all, a huge thank you for your wonderful help.
In my opinion, the lower limits should be removed (set to 0%), as the lower limit can be set manually via voltage, which I think is a good idea.
If possible, please reduce the lower limit to 0%.

[Zyga76]

[Zyga76]

[Zyga76]

Hi,
I noticed that at the 10% threshold, the power draw is cut to around 200W.
The discharge/cutoff limit for 0% needs to be removed, though.
Ideally, it should be lowered to 3.1V per cell.
If possible, please address this quickly so I can run some tests.

[Zyga76]

Hi,
I noticed that the cell voltage drops to around 3.5V at the lower limit, and the BE actually started to throttle the discharge.
I believe the cell voltage can safely go down to 3.1V, and the 0% mark could be set at that level.
It could then safely start throttling at 2% to 0%
Please update the parameters so I can run tests and share my observations regarding these changes.
I am also attaching the logs.

canlog_00-03-13.txt

[Zyga76]

Hi Jonny 5532,
I have a quick question: is it possible to remove that lower threshold? I mentioned this issue earlier because the value is too high for me.
I’d really appreciate some help, as I’d like to share my observations regarding this battery.

[ibrahimshweiki1995-tech]

Hi you can’t discharge battery low then 10% it will effect to heath of battery and damage you need battery big then you have

[Zyga76]

Hi,
Please lower the cut-off threshold to at least 3.3V per cell, because at 3.5V, the battery still holds a significant amount of energy.
I would really appreciate it.

[Zyga76]

Hi,
I know you’re very busy, but I’d really appreciate it if you could help me by lowering the cell's low-voltage cutoff to at least 3.3V, as that is a very safe level.
Cells of this type can be safely discharged down to 2.9V without any issues.
Once the lower threshold is adjusted, I’ll test it immediately and share the results so they can be integrated into the BE project.

[Zyga76]

Cześć,
wiem, że jesteś bardzo zajęty, ale byłbym bardzo wdzięczny, gdybyś pomógł mi obniżyć dolny próg napięcia ogniwa do co najmniej 3,3 V, ponieważ jest to bardzo bezpieczny poziom.
Ogniwa tego typu można bezpiecznie rozładowywać do 2,9 V bez żadnych problemów.
Po ustawieniu dolnego progu natychmiast go przetestuję i podzielę się wynikami, aby można je było zintegrować z projektem BE.

[Zyga76]

Hi,
I would really appreciate some help lowering the cell's low-voltage cutoff to 3.3V, as the current threshold is quite high; this type of cell can be safely discharged down to 2.9V.
The manual low-voltage limit isn't working either, as the system automatically cuts off the cells at 3.52V regardless.
The battery's capacity doesn't matter; what matters is the voltage level to which it should discharge.
I believe a 3.3V cutoff would be perfect.

[jonny5532]

Currently, BE normally relies on BMS-reported SoC as an indication of battery state, stopping charging at 100%. This requires changing this (within this integration, anyway), so that:

• Charge/discharge power is calculated based on min/max cellvoltage rather than SoC
• SoC is capped at 98% until max cellvoltage is reached
• SoC is capped at 1% until min cellvoltage is reached

I've implemented this and am currently testing it - there are obviously potentially serious consequences if it goes wrong (although the BMS should intervene and open contactors).

There are also complications with behaviour around the limits - it can 'bounce off' the upper limit if the current limit drops too suddenly (and similarly at the bottom) which will also need careful control.

[Zyga76]

Zgadzam się z Tobą.
Jeśli możesz przesłać nowe rozwiązanie, chętnie je przetestuję – z dużą ostrożnością i na własne ryzyko – i podzielę się swoimi spostrzeżeniami.
Uważam, że progi 98% i 1% są bardzo dobre.

[Zyga76]

I agree with you.
If you can send the new solution, I’ll gladly test it—with great caution and at my own risk—and share my observations.
I think the 98% and 1% thresholds are very good.

[Zyga76]

I agree with you.
If you can send over the new solution, I would be very happy to test it—with great caution and at my own risk—and share my observations.
I think the 98% and 1% thresholds are very good.

[Zyga76]

Hello,
When will the new version be available?