Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Samsung battery fails test #63

Closed
jezza61 opened this issue Sep 22, 2024 · 23 comments
Closed

Samsung battery fails test #63

jezza61 opened this issue Sep 22, 2024 · 23 comments

Comments

@jezza61
Copy link

jezza61 commented Sep 22, 2024

Hi,
I now have the Deye inverter connected to the Samung battery.
The Raspberry PI sends the following data:
can1 00000351 [8] 45 02 D0 02 B0 0E C0 01
can1 00000355 [8] 5F 00 64 00 00 00 00 00
can1 00000356 [8] 26 16 EF FF 00 00 00 00
can1 0000035E [8] 00 00 00 00 00 00 00 00
can1 0000035E [8] 00 00 00 00 00 00 00 00
can1 00000359 [8] 00 00 00 00 00 00 00 00
However Deye will not even look at it.
Also solar assistant fails to see it as valid data.
Is anything missing from it?
Best Regards,

Jerry
@jezza61
Copy link
Author

jezza61 commented Sep 23, 2024

I have done some more testing.
The Raspberry is sending data to can IDs with id padded with 0 as seen above. This is apparently not acceptable to Deye.
When I send the same data but with canid only 351, 355, 356 and so on, it works and Deye sends a message 305, with all zero data.
Second issue in the web page the voltage is shown with % and not V.
Also the batter current seems to be dived by extra 10, as when the battery reports 31A the web page shows 3.1A.
And the 35C is not present, right?
This is charge enable, discharge enable, force charge values, this I don't believe is reported by Samsung BMS.
Is there any way of dropping the charge voltage of 58.1 down to 56.6 to reduce the risk of overcharge ?

Best Regards,

Jerry

@jezza61
Copy link
Author

jezza61 commented Sep 23, 2024

Also I used this document:
https://www.studocu.com/row/document/%E5%8D%8E%E5%8D%97%E7%90%86%E5%B7%A5%E5%A4%A7%E5%AD%A6/physical-chemistry/can-bus-protocol-%E5%BE%B7%E4%B8%9Adeye-low-voltage/46210190

And this data appears to match the Deye inverter stuff.

@jezza61
Copy link
Author

jezza61 commented Sep 23, 2024

And I believe that the system current is FFFF - 2 byte data from ID 500 from Samsung.
First 2 bytes show system voltage in 0.01 units, second 2 bytes show system current, but it appears that you have to deduct these bytes from FFFF and then it gives you the correct value, or at least it appears to do that.

@jezza61
Copy link
Author

jezza61 commented Sep 24, 2024

Correction, CAN messages do have the correct System current value. RS485 also has the same value.
Somehow the software does not reflect the correct number.

@ai-republic
Copy link
Owner

Hi @jezza61,
sorry for the late reply I'm currently very busy - also at the weekends 😒
What I gather from your findings is that maybe the Deye inverter uses CAN 2.0A with standard length instead of CAN2.0B with extended can-id length. I will see what I can find out.
Thanks for the web-page issues, I will have a look at the values and also the %.

@ai-republic
Copy link
Owner

Right, I fixed all your found issues :)
It was correct that Deye uses t he standard CAN 2.0A can-id and not the extended. I also checked the system current and fixed that too as well as the voltage being shown with the percent sign.
Please do a Clean install with the Configurator application and let me know if its all working now 😄

@ai-republic
Copy link
Owner

@jezza61 did you get a chance to test the fixes? Please let me know if everything worked out

@jezza61
Copy link
Author

jezza61 commented Oct 6, 2024 via email

@jezza61
Copy link
Author

jezza61 commented Oct 6, 2024 via email

@ai-republic
Copy link
Owner

Hi @jezza61,
great to hear its working now 😄
For your problem with the max charge I can write a small plugin to set some parameters like max and min charge voltages/currents. I'll see that I get that done next weekend.
For your grid charging problem: that you need to set in the Deye inverter settings either via Deye cloud or on the inverter itself.
In Deye cloud it looks like this:
image
There you can check the Grid and set the times and voltage levels to charge via grid. Hope that helps 😄

@jezza61
Copy link
Author

jezza61 commented Oct 6, 2024 via email

@ai-republic
Copy link
Owner

Hi @jezza61,
I implemented the ManipulateBatteryPackDataPlugin plugin that should serve your purpose to manipulate max charge/discharge voltages and currents. You'll need to download the latest Configurator and do a clean install.
Please let me know if it works for you 😄

@jezza61
Copy link
Author

jezza61 commented Oct 13, 2024 via email •
edited by ai-republic
Loading

@ai-republic
Copy link
Owner

@jezza61 found the issue. It was not overwritting the defaults. Fixed it. Please can you do a Clean install and retest?

@jezza61
Copy link
Author

jezza61 commented Oct 14, 2024 via email

@jezza61
Copy link
Author

jezza61 commented Oct 14, 2024 via email

@jezza61
Copy link
Author

jezza61 commented Oct 14, 2024 via email •
edited by ai-republic
Loading

@ai-republic
Copy link
Owner

@jezza61 yes, bit 6/7 (discharge/charge) tells the Deye inverver the state of the dis/charging breaker. The Samsung CAN specification actually doesn't reflect this. There is no data for that.
What would you like to do?

@jezza61
Copy link
Author

jezza61 commented Oct 14, 2024 via email •
edited by ai-republic
Loading

@ai-republic
Copy link
Owner

@jezza61 any updates?

@jezza61
Copy link
Author

jezza61 commented Oct 31, 2024 via email •
edited by ai-republic
Loading

@ai-republic
Copy link
Owner

So I will close this ticket for now. If you encounter any problems please let me know

@jezza61
Copy link
Author

jezza61 commented Nov 3, 2024 via email

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@ai-republic @jezza61