question

nuno-m-g-santos avatar image
nuno-m-g-santos asked

Pylontech SoC jumping Up and Down

Dear Community


Till now, at least for me, the expected behaviour of the Pylontech SoC on the Dasboard Overview graph was a point over time, with little variance within the same hour.


That would result in a graph like this.

1624857172924.png



However, in latter installations, done with multiple US2000C (and using a LV HUB), the behavior of the systems has a wide range of SoC, that does not correspond to reality.

1624857345705.png


For sure there is a communication problem, as current SoC moves from 100% to 50% within minutes and then comes back again, and that results in the chart above with the "undefined" SoC.


Another example (another installation also using 18 Pylontech US2000C and the LV Hub):

1624857607218.png


This one is even stranger, as the consumption adheres to the production, but I will ask for it in a different topic.


Note: it was hard to put the 18 Pylontech working right with the LV Hub, as the communication cabling is different from Pylontech US2000B.


Moreover:

Pylontech type B, use the Victron CAN cable type B.

Pylontech type C, use the Victron CAN cable type A

Which cable is the more accurate for the Pylontech LV Hub, using the new Pylontech type C?


Can you get me a updated manual on this type of installation with multiple Pylontech C batteries?


Best regards and thanks in advance.


Nuno M. G. Santos

Pylontech
1624857172924.png (160.3 KiB)
1624857345705.png (167.1 KiB)
1624857607218.png (166.4 KiB)
1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

wkirby avatar image wkirby ♦♦ commented ·
The Type-B cable should be used between LV-Hub and Victron GX device, regardless of battery model. The battery model does not change the pinout of the LV-Hub CAN-OUT. Type B cable has GND on pin 2 and LV-Hub CAN-OUT also has GND on pin 2.


Maybe @Ned Yu (Pylontech) can confirm?
0 Likes 0 ·
4 Answers
Ned Yu (Pylontech) avatar image
Ned Yu (Pylontech) answered ·

@Nuno M. G. Santos hi there, Could you provide your email address? Our Technical engineer will contact with you for further assistance.

or you can directly send relevant information to below email address:

jeff.cheaung@pylontech.com.cn

Thank you very much !

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

shelldor avatar image
shelldor answered ·

Has this problem been solved?

I have a Victron Easysolar II 48/3000 with a Pylontech Force L1 and sometimes also see the SOC jumping in an unrealistic way, but far less often then in the pictures of the thread author.

Is this a malfunction of the Pylontech Battery?

Best regards

shelldor

screenshot-20220904-081233-vrm.jpg



screenshot-20220904-081233-vrm.jpg (93.8 KiB)
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

nhuillard avatar image
nhuillard answered ·

I have a similar problem, and checking the "Advanced" page on the VRM shows that some values read from the Pylontech batteries (4×US3000C) are sometimes wildly wrong. It can be temperature=-1516°C (impossible, at least) or +456°C, DC current=-755.2A or +2882.7A. It is usually a single reading; everything is OK the rest of the day or week.
I use an official Victron cable, which is unshielded. The installation is not old (6 months).

I guess there are transmission errors along that cable, and no error dectction mecanism or no checksum at all that could enable error detection.

I'd humbly suggest that some sort of error control be implemented on the receiving side (GX OS): reject the whole packet of data when some reading are obviously wrong (like below 0K temperature, or a current that would blow any fuse or wire, etc.)

On the original poster screenshots, the problem is very obvious and annoying. I guess that double-checking the cable or adding a shielded one could improve the situation.

As asked above: was the problem solved, and how?

2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Alexandra avatar image Alexandra ♦ commented ·

@nhuillard

How would the system know it is wrong?

With hub use, sometimes it just needs a good old reboot. Are you using the hub?

0 Likes 0 ·
nhuillard avatar image nhuillard Alexandra ♦ commented ·
  1. these are transitory readings
  2. a temperature below absolute 0 can't be correct
  3. we may argue about +2882A though...
  4. same about values above sensor specs, say +456°C...

See screenshots for my setup:

capture-decran-de-2022-09-16-13-07-00.png

capture-decran-de-2022-09-16-13-05-56.png

Here the Pylontech BMS reports a wrong temperature and current a single time, but cell voltage is correct that same time. Sounds like a transmission error that could be detected.

A reboot won't solved this (I incidentally did reboot the whole system recently). It's not a real problem for me and does not have apparent consequences. It's just annoying (in my case) to get flattened graphs because a single wildly inconsistent value.

Seeing the graph from @Nuno M. G. Santos is a bit more annoying though...

0 Likes 0 ·
capture-decran-de-2022-09-16-13-07-00.png (96.6 KiB)
capture-decran-de-2022-09-16-13-05-56.png (91.7 KiB)
rodenberger avatar image
rodenberger answered ·

Das ist ein Softwareproblem der Pylontechs.
Neue US2000C mit neuem Chip und Version 1.3 ???
Updaten auf 1.4, aber vorher unbedinght den Verkäufer fragen !!!!! Garantieverlust möglich !
Auslesen mit BatteryView:
INFOS unter Effekta.com .
Update auf falsche Version --> Akku defekt !

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.