[BUG] Battery Emulator reports "Battery-Undervoltage" at 360Vdc (SMA SBS2.5 connected to Think-Battery)

[Simply-Johann]

Describe the bug
A clear and concise description of what the encountered bug is

To Reproduce
Please try to explain the steps required to reproduce the issue

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Version and settings (please complete the following information):

• Please attach screenshot of configuration for easier debugging
• Software version: X.Y.Z
• Battery used: [Nissan LEAF]
• Inverter communication protocol: [BYD Modbus]
• Hardware used for Battery-Emulator: [LilyGo T-CAN/Stark CMR/LilyGo T-2CAN]
• GPIO controlled contactors: [yes/no]
• MQTT: [yes/no]

[dalathegreat]

Fix for the Undervoltage event on startup: #1960

[Simply-Johann]

I made several pairing attempts at a battery voltage of 324V. I navigate to the battery configuration in the SBS server form wizard, but nothing happens after that. The inverter does not recognize the battery. The battery emulator reports "Battery_Undervoltage" and "Inverter missing". I've currently even raised the battery voltage to 360V - BE is reporting "Undervoltage"...
Then I connected the SBS2.5 to the mains, and when the enable line went high, I started the BE and the battery (about 2-3 seconds after enabling). This time I had a pairing attempt.

[Simply-Johann]

The battery emulator sends the CAN messages to the inverter as soon as the enable line is set to HIGH.
The high voltage at the inverter input is low at this time. My main contactor only closes after 1200 ms. Now HV reach its value. Emhal reports that a delay for CAN Messages after Enable_High solve the pairing problem at SBS2.5. And, apart from the Battery_UnderVoltage problem, my inverter is behaving the same way as his. Could you implement the SBS2.5-inverter protocol of Emhal to check it? What else could be causing the BE error message "Battery_Undervoltage" at real 324V?

Hi Daniel, why can I not describe the BUG? I can not edit my report...

[dalathegreat]

Undervoltage is fixed with this fix: #1960

Pairing I cannot say much about, maybe someone else that has more SMA info can say!

[Simply-Johann]

Hi Daniel, the fix #1960 does not lost the Problem..., and I can see that the BE CAN transmission to the inverter begins before the contactor closes. I'm monitoring the CAN traffic with a PCAN-USB, and I have a LED connected via a resistor to the enable line.
It is clearly visible that communication begins first (directly with Enable high), and only later does the main relay click.

@emhal wrote in discussion #1599: "...and made sure the CAN messages will not be sent before the contactors are fully closed".
That is not happen with the new update.

P.S. When updating I disconnected Inverter- and Battery CAN-Bus. Battery and Inverter are switched off.
The BE event "Battery_Undervoltage" is coming very early, estimated in the first minute, my be after first contactors opening...
Wich logs/traces should I make to show the bug?

[Simply-Johann]

Hi Dala, after 2 days I downloaded again the fix #1960, the firmware.bin was 1h older, then my first download (but exactly the same size). I again updated OTA my LiLyGoT-2CAN. The BE reports "Battery_Undervoltage" direkt after Start (first start I get Battery_Overvoltage, my be because real Battery_Voltage were 360,3V -->I discharged the Battery to 359,5V). Now, despite "Bat_Undervolt" Inverter try to pairing... Three attempts with inverter protocol BYD Battery-Box H were unsuccessful.
I changed Inverter protocol to BYD Battery-Box HVS and made next 2 attempts --> seems Inverter try longer to pair...
On the third attempt, there was a loud bang and I thought the inverter had exploded... But don't worry, I'm a repair engineer for power electronics... - only the fast-acting semiconductor fuse 63A is blown (second time and with the second inverter).

Daniel, if you have time, please try to solve the "Bat Undervoltage" problem for my ThinkBattery (Enerdel A306).
@emhal : Frank, what kind of fuse do you have in your HV DC circuit? Have you also had problems with the fuse blowing?
Is your inverter now working stably with your new settings?

[Simply-Johann]

@dalathegreat
@emhal
After my 63A fuse blew, I bypassed it with a 24-ohm heater to test the inverter's function. I've never seen so much battery data on the SBS2.5 server as this time... (probably because I didn't immediately go into the battery configuration). The BE inverter protocol was BYD Battery-Box HVS, and the BYD Battery-Box Premium HVS was previously stored in the SBS2.5 server's battery configuration. The pictures prove that the inverter is still working…, only the wizard assistant needs to be completed

...and after 10 minutes again Klick/Klack of the contactors with about 20 s intervall

CAN traffic is disrupted; I can see this in the monitoring and also in the trace. Just as @emhal already reported.
But Frank also reported success...

HVS_Pairing_Start-4-CanStocktNachContaktorsClosing.trc.txt

[emhal]

Hey Johann,
sorry for not replying earlier! I'm quite busy with other stuff the last weeks. I know how to solve the SBS 2.5 pairing issue and I have a fully working BE version based on BE 9.3.0 I could send it to you via mail as it is not possible to upload it here. Maybe if I find the time I could also create a version based on the latest BE release. But of course the final goal would be to implement that new inverter protocol especially for the SBS 2.5 into the BE project.
BR
Frank

[Simply-Johann]

Oh, I'm so glad to hear that from you, Frank!!
I've already posted the dbc file for my battery (do you need it?). But I don't have an Arduino IDE to compile yet... - I'm a "Raspberry- Python" guy. But I could flash a pre-compiled bin file to the BE via OTA.
You have my PM as well - johann.eckert@freenet.de
Thank you very much!
Johann

[Simply-Johann]

@dalathegreat
Hi Daniel, I looked at the installation instructions for SMA SBS2.5 again and found the following:

That is wrong, because BYD Battery-Box HVS&H over SMA CAN does not exists.
Insteat of this exists "BYD Battery-Box Premium HVS over CAN Bus", but it didn't work for me...
I think you've confused the protocol with "SMA compatible BYD Battery-Box HVS&H" - both work (unfortunately not correct) with the SBS2.5
Just FYI

[emhal]

@Simply-Johann did you receive my mail with the modified 9.3.0 version and did it work for you?
today I had time to build a 9.3.5 version with my SBS2.5 compatible HVS protocol and it worked directly (just had to change the BE setting to SMA HVS as there was a previous change making the old tripower to the new HVS and the old HVS to the new H. Therefore my old setting now pointed to the BYD H instead of BYD HVS). I also didn't implement my contactor switch delay this time which I implemented before to get the pairing to work. I just haven't tested if the pairing is still working flawlesssly as the update was installed while the inverter was running and therefore also didn't pair again. I will test that the next days. Then still the question is how to contribute my SBS2.5 inverter protocol to the BE project. We could use the now free InverterProtocolType 13 as "SMA-SBS2.5-BYD-HVS-CAN" with the SBS2.5 compatible version. I will send the firmware file I've just built for the 2CAN board to you via mail. I've also sent you my mobile phone number in the last mail if you need support for the pairing. And I just saw, that #1960 was not implemented yet in 9.3.5 so I built another version with that fix in it and sent it to you via mail too. I've tested that one too and it is working. Of course not with a Think battery because I don't have one but at least the build is working fine with my setup.
BR
Frank

[Simply-Johann]

@dalathegreat
@emhal
I received firmware.bin version 9.3.5 from Frank, and pairing my SBS2.5 was trouble-free. Now my inverter is working (charging/discharging), but only for about 3-4 minutes at a time. After that, the battery BMS switches off its internal contactors with a service code in message ID 610h: "Isolation Fault With Contactors Off".

However, there is no isolation problem, because the BMS (controlled by BE via ID messages 310h, 311h) works without issues with all other devices. Only the SBS2.5 switches off Battery internal contactors (via BE??) after about 3 minutes. All devices (BMS, BE and SBS2.5) continue to operate without errors; only because the high voltage is now missing does the inverter report "Voltage Deviation"...
Do You have any idea for this behavior?
Wrong implementation of my (ENER1 A306) Think Battery?
Which BE event could always trigger the BMS error after 2-4 minutes?
I am grateful for any help.

[dalathegreat]

Hi @Simply-Johann

Yes this is a common problem on EV batteries. When put into stationary storage, the solar inverter performs isolation measurement. This test can interfere with the EV battery internal testing of the same thing, and make the EV battery think there is an isolation leak.

On the Stellantis ECMP platform, we have to modify the BMS to be unable to detect isolation fault, otherwise the battery will open contactors after a few minutes. https://github.com/dalathegreat/Battery-Emulator/wiki/Battery:-Stellantis-eCMP-(Citroen,-DS,-Opel,-Peugeot)#disabling-isolation-monitoring-via-hw-modification

I suggest closing this issue, and opening a new one dedicated to Think BMS isolation issues.