I’ve been having very odd issues with my new lithium batteries / random charging issues.
Basically, what is happening is that, yes, I purchase batteries from Alibaba and have been in very close contact with them; they have been helpful
So, When charging it the batteries will go from 16% to 100% within a few seconds, they will do random things when charging, the Victron is reading the batteries at 53.2v when they are sitting at 1%, I put a multi meter on each battery and the batteries are reading 3.331v and out of the 16x 3.2v 314ah batteries I have 2 batteries that are reading 3.330v and 1 of them is reading 3.329v
When happens is that the batteries reports fully charged, I don’t know, the system is reporting its fully charged and telling the BMS it’s fully charged at 300ah.
When the battery hits 0% the house power turns off for a moment until the generator has fired up and its a hard stop from the BMS side there isn’t any smooth transitions. I don’t understand this as I’m very new to this and the system has had 20 cycles. The lowest point the batteries hit is 25% is what is set in the Victron ESS and 27% in the AC conditioning settings
Apart from one large voltage dip ( your disconnect?) The voltage seems to be reasonably constant.
This is probably a faulty design of BMS, or some problem with it. Try disconnecting the BMS for a few cycles, and running the battery without -if you can.
What BMS are you using and have you got it communicating with Victron inverter? For example I use a JK-BMS and it communicates directly with my Multiplus.
JK BMS, Its using the VE Can and the Inverter is picking it up
Oddly enough the Victron App is reporting the battery is being 54.03v and the VRM connected to the Cerbo is showing the battery as 50.01v It’s not frozen also because it’s flashing between 50.01, 50.00 and 49.99v
I’m confused that one to trust, As I’m connected directly into the local IP address of the cerbo
There are also things you have to set in the initial setup of your Victron for the battery even if you have a BMS. Basically it’s a failsafe to protect your battery in th event of a BMS failure.
Depends on what type of Cerbo you have. Informatin from the manual.
Key Differences
Cerbo GX (original):
1 × VE.Can port (non-isolated)
1 × BMS-Can port (non-configurable, fixed to 500 kbit/s)
Cerbo GX MK2:
2 × Fully configurable VE.Can ports (VE.Can 1 is isolated)
Both ports can be set to:
VE.Can (250 kbit/s, default)
BMS-Can (500 kbit/s)
CAN-bus BMS (250 kbit/s)
Other supported CAN profiles such as RV-C
Usage guideline
VE.Can (250 kbit/s, default)
For Victron devices like:
VE.Can MPPTs
Skylla-IP65
Lynx Shunt VE.Can
Lynx Smart BMS and Lynx Smart BMS NG
Terminate both ends using the included VE.Can terminators
BMS-Can (500 kbit/s)
For managed lithium batteries (e.g. BYD, Pylontech, Freedomwon)
Terminate at the Cerbo GX with the included terminator
Follow the battery manufacturer’s instructions for termination on the battery side
It looks like you have communication with the battery but now you have to determine if it is correct or getting garbled. Have you checked the settings on your BMS yet? It is very important that they are correct. PS have you terminated the CANBUS with the supplied terminator? In the port adjacent to where the BMS network cable is attached you should have a terminator unless you are looping through to another device on the same bus.
Yes, everything should be correctly isolated, I made up my own RJ45 terminators from the RS Multi to the Cerbo, not sure if it’s Mk2 or not. I made them 3 years ago when I used to have AGM batteries. Just note it has been working fine for about 20 cycles of the battery, and suddenly it just lost the plot. I’ve been in contact with the supplier who said that if I wanted to get the battery’s capacity tested, they would be happy for me to do that and they can provide the serial numbers matching the batteries’ certificates. I ended up dropping the total capacity from 314ah to 300ah and it’s become stable again (On the BMS end) and on the victron end I dropped the battery capacity to 280ah but the battery doesn’t drop below 30% until start of this week when it became very unstable.
Note: I also made up the CAN cable using brand new CAT6 and I followed a guide on how to correctly terminate the cables for CAN from memory it was Cerbo was pins 8 7 6 and the BMS side was pin 1,2,3 I forgot now what I did but it was the correct way of doing it and the Victron detects the BMS and the serial number and it also detects all the batteries and everything.
When battery voltage and system voltage are different in the case of JK inverter BMS than most likely this is either charging or discharge not allowed. The mosfet’s can be closed or open on direction. This happens a lot when you are near 100% full or 0% empty and one cell is hitting the configured limits in JK inverter BMS.
The system voltage is than floating…. on its capacitors… Quite dangerous situation. This should only happen while testing batteries to their limits etc. This should never happen during normal system operation. Spikes on the ACload and clouds over solarpanels can suddenly change system voltage and the system will shutdown on DC voltage too quickly fluctuating
OK, I may have found an issue. 2 of the BMS leads were dodgy and it was suddenly reading 2 batteries with 0v. I cut the end of the eyelet and redid it and the battery started reading 3.330v
The batteries seemed to have charged somewhat smoothly today - We had very very nice sun today and it was pulling a solid 5.5kw in from the panels, the one thing I never like is that when the batteries hit 100% charged the inverter shuts the solar panels off until the batteries soc hits 98% and then it stays 98% until sun down.
Great to hear that you found a clear battery problem. The charging stopping at 100% is because that battery is not really compatible with Victron or in the case of JK inverter BMS the configuration is not correct. Battery at 100% says in battery parameters menu CCL 0A… Thus Victron MPPT’s stop charging till it’s allowed again by the BMS. What the long list of supported batteries do is lower the CVL to a float voltage and keep CCL and DCL wide open. In the newer firmwares of Jk inverter BMS you can set the timer for bulk charge and float charge and set a float voltage. LFP batteries should not be kept at over 3.4V after equalisation time. When sudden load changes and clouds happen than there should be some “roam” left in the cells. Otherwise they spike at even small charges and you will get cell overvoltage alarms. To prevent cell over voltage alarms the MPPT will simply listen to the BMS and make sure it doesn’t even risk charging… only discharging till BMS opens up CCL…. Thus your BMS settings are causing this behavior.
