At the ESS I can set the lower SoC (20%), but I am not able to set the higher SoC. Chat GPT tells me that I have to lower the Target SoC to 80% to enhance battery life. So I want to limit the Target SoC. I also tried it with DVCC (maximum charge voltage is 53.4V), but the ESS ignores this setting. The Target SoC determined by ESS is 100%. How can I limit the the target SoC to 80%?
Best regards,
Mark
I noticed that 54V is not 100% charged, so maybe I have to change that voltage again with the MK3.
@MarkBng âŚI know is is quite trendy to ask Chat GPT for information but I would like to caution anyone who believes this is a source of sound engineering information to see the information I have received from ChatGPT about the settings for Power Boost factor used for the Power Assist function in a Quattro or Multiplus.
The information provided by Chat GPT was given with great confidence with no sign of doubt and was absolutely and positively totally incorrect. Please use Chat GPT with caution when asking engineering questions. I know enough about the subject to know its answer was complete rubbish but it would be very easy to have the completely wrong impression of how it works.
This is a snippet of the response showing just how wrong it can be.
Donât believe all that you read online.
The BMS is already keeping the battery at a conservative level when 100%, bringing it down will only hinder the battery efficiency.
Battery soc is usually âcalculatedâ.
Two triggers are usually
1.above a certian voltage and
2.not drawing current.
So depending on how clever the one is in your bms, it could be wrong
And FYI a battery does not stay healthy if it is not taken to the level where is balances. The bms is already designed to manage the battery, donât mess with it.
For your 16S battery, a healthy voltage will be about 3.5V per cell, so about 56V for 100%.
This above considering that 3.65V is the maximum allowed.
So you see, your 54V is already way to conservative, but accepted for active BMSes that are balancing pretty constant.
I have a basen green battery housing and the BMS is connected to CAN BMS. both are connected to the 2 ports on the cerbo GX. This basen green battery has the setting Victron protocol (over CAN) and is connected with an ethernet cable. I donât know if there is any good communication with the BMS. Where can I see that everything is working? A third battery is underway. How can I connect this BMS to the BMS CAN ports since there are only 2 BMS CAN ports?
This is possible, but not directly
Youâd need to use Dynamic ESS Trade mode and allow battery to discharge to grid
Then itâll stay on lower Target SOC most of the time
If you need to keep it without DESS game - just set a lower float voltage, something like 52V for 16s or 50-51 for 15S - and itâll work just fine.
Do not forget periodical Absorbtion for balancing
The Basen green is a 16S Eve 314Ah configuration. It has an active balancer inside and a BMS that is Victron compatible. Itâs only a little bit less than 1900EUR, so thatâs a good price for a 16kWh battery. The system is working well now. This is my first system. Had no training and figured everything out by myself. I am a hardware design engineer, so I know something about (designing) electronics. But it was not easy to get it working for the first time. Now it works like a charm. I used 95mm2 cable to connect the cables. Those are wires used for electric welding. If you hesitate which cable to use, than take the bigger one. I know, maybe itâs overkill. But 300A is also a lot of amperes.
We use 95mm for 300A. Its all good there.
CAN is designed for and works with daisy chaining connections.
Maybe try a splitter if yours doesnât have two ports for the chaining.
Your battery needs to be charged to 100% to balance cells you will gradually loose efficiency if you donât charge fully charge and allow absorption at least once a week
I use Pylontech the do recommend full charging , all the batteries
do have reserve capacity above and below standard capacity
I continue to be amazed at the almost obsession with top balancing an LFP bank.
It is my actual experience that a smallish 720Ahr bank of 180Ahr LFP large prismatic cells used as a house bank on a live-aboard boat that actually cruised regularly, (i.e., not a dock queen) stayed in balance to 0.002VDC between the for over a year. The BMS did not have a balancing capability so the owner finally top balanced them with a small power supply just for practice.
Repeatedly driving an LFP into the charging knee will damage the battery. Period.
I agree. According to all Iâve read the âstressâ increases close to the saturation limit and at voltages close to cut-off, and this sounds reasonable to me. A small issue might be that you clearly see unbalance only in or beyond the upper or lower knee.
Iâm doing balancing (and SoC calibration) after the winter and during summer every 1 or 2 months. In the meanwhile my common top SoC is 85% (or perhaps 90âŚ95 on a sunny day if Iâm sure there will be two âdark daysâ following).
@MarkBng this can be done by a very small and simple script limiting the charge voltage to float voltage level when SoC comes to 85%. Iâm using MQTT to control â/settings/0/Settings/SystemSetup/MaxChargeVoltageâ to do this. You can do via Modbus or dbus as well (dbus is perfect if the script runs directly on the GX).
But do so only if you have enough buffer - deep discharging is worse than charging to the top And do so only if you have a quite reliable SoC info, e.g. a SmartShunt or other battery computer. With SoC provided by e.g. a JK BMS or even Multiplus you may fail sooner or later.
CAN is designed to. But the current GX-to-BMS âarchitectureâ is point-to-point, isnât it? how does the GX seperate between the BMSâs on one bus? How do the BMSâs get different node IDs / using different COB-IDs? (but this is another topicâŚ)
There is a difference between active and passive balancing BMSes. The later need to reach almost 100% in order to balance properly while the former can have it already balanced even at 90%. Although you can see for sure if itâs balanced on the steeper part of the voltage curve, where the resolution is better.
When the voltage for 100% SOC is reached and after balancing is finished, even if you keep there at 3.5V/cell constantly, the current will drop to 20-30mA, the big part of it being the current needed for the BMS to function.
If someone can tell me how you can hurt a 200A battery by feeding it only 0.02A, even permanently, then I will take my hat offâŚ
At this current the battery is stone cold and the dendrite and SEI formation is inexistent.
Hi, there is two options / information about LifePo4
The right way to limit your max SOC is to lower the voltage settings for 100% SOC. Battery has no indicator to reach 80% and will drift away soon or later! So at the end it need always reach 100% even when the max voltage is limited to âaboutâ 80%.
if you have a 16S battery your 54V is 90% of SOC already or lower. Max charging voltage is about 58.4V for lipo
If you have 15S lipo 54.75V is max charging voltage.
53.6V is nearly 100% SOC
Sadly no comment from Victron as this is something I would dearly like. Currently I manage my SOC "manually " so in summer its around 70% Max SOC using 53.8V DVCC and winter around 85% at 54V. Battery is a DIY 1300AH. I balance it every 2 months at 54.8V top voltage and this is achieved in around one hourâs time using JK BMS. SOC is determined by Victron BMV 702 and its super accurate ( charge efficiency 0.98 , Peukert 1.00 ) - error at balancing is around 2-3 % . If you look at the scientific literature - for maximum life LFP batteries should be operated way below 100 %
Best advice for everyone - buy a trusted manufactured battery for your safety and sake of free time, and let the BMS deal with all battery related issues. Who has all the time to balance batteries? I envy you guys for having so much free time.
And do so only if you have a quite reliable SoC info, e.g. a SmartShunt or other battery computer. With SoC provided by e.g. a JK BMS or even Multiplus you may fail sooner or later.
I envy you guys for having so much free time.