J’ai deux batteries 280 AH pilotées par des BMS JK. Mes batteries sont connectées à un CERBO VICTRON et à un inverseur Victron Multiplus II.
Bien que les batteries soient identiques, leur SOC (State of Charge) respectif diverge au cours du temps. Ceci conduit à des situations où pour un même voltage une batterie considére qu’elle est chargée à 10% et l’autre à 40%.. Comme le BMS maitre n’envoie au VICTRON que la moyenne des deux SOC, cela provoque des situations où le Victron considère à tort que l’ensemble des deux batteries a atteint une limite basse de charge et arrête de les solliciter.
La solution que j’ai trouvé pour le moment est de recharger périodiquement et complètement les deux batteries. Mais c’est contraignant et cela ne permet pas de laisser le système en fonctionnement autonome pendant des longues périodes.
QUESTION : Y a-t-il un moyen d’éviter cela et de maintenir sur le long terme la cohérence des SOC ?
Bonjour johnnybug,
Comme le dit LX, Cela peut être à cause d’un mauvais étalonnage entre les deux batteries, aussi un câblage avec des longueurs et sections différentes, état de santé diffèrent, …
Dites nous en plus sur votre systeme. Si tout est strictement identique ?
Mes deux batteries sont à base de cellules 280 AH. La plus ancienne est gérée par un BMS JK_PB2A16S15P. La plus récente un JK_PB2A16S20P. J’ai un souci avec la plus ancienne qui ne se charge qu’à 99% quelle que soit la durée où je la laisse en charge seule sur le Victron (paramètre keep battery charged). Je suis en train de vérifier TOUS les paramètres des deux BMS mais en en principe ils sont identiques. Lorsque je mets les deux batteries en parallele en fonctionnement normal, au bout de deux ou trois jours des écarts considerables apparaissent sur leur SOC respectifs et je suis obligé de les recharger à nouveau. La section de cable est identique pour les deux batteries et leur longueur est egalement comparable. La seule difference est que les cellules de la batterie plus ancienne ont une année de plus mais cela n’explique pas pourquoi je n’arrive pas à charger cette batterie à 100% ni pourquoi les soc divergent autant. Je peux joindre des captures d’écran des settings des deux batteries si cela permet de faire avancer les choses. En tout cas merci pour toute aide.
Had You performed the current calibration of both BMS? At low currents there could be serious error, I suggest You check the current readings of each BMS with ampermeter or Victron shunt.
In my experience (have smaller BMS with eight 280 Ah cells) BMS has measurement granularity of 0.22 A and is mostly unusable at calculating SOC of system working at charge / discharge currents of several amps.
When the drift occurs between both batteries, their voltage is almost the same while their SoC is totally different. I suspect that my problem is linked somehow to the fact that I can’t charge the second battery to more tan 99%. Therefore I ‘m following three tracks of investigation : 1/ Check and re check the settings of the bms of both batteries. 2/ Trying to find a way to reset the second battery so that it charges at 100%. 3/ Looking for a possible control of the SoC by Victron and not by the JK BMS.
It’s mostly for this third option that I posted my question here.
CHATGPT suggest that the best way to fix my problem is haviing the SoC managed only by Victron. The JK BMS having then only the responsibility of safety (over voltage, under voltage, etc). Does this make sense? If yes how can I do it?
C’est un travail de longue haleine que d’ajuster deux batteries d’âge différentes et plus encore avec deux versions de BMS dont probablement deux variants d’algorithmes. Le travail en vaut la peine, c’est fastidieux et il faut bien comprendre le fonctionnement des LFP4.
Perso j’ai un smartshunt, il est bien plus précis.j’ai reglé une charge tous les mois a 100% pour qu’il synchronise le soc du JK. met nous tes réglages JK que l’on t’aide a régler le problème des 99%.
You get 100% SOC after the cells has been fully balanced, but 99% isn’t a problem, but do a fully cell balancing by limit the charge current if the cell voltage difference is more then 60 milivolts.
I have done a cell balancing recently with only 0.8A charge current, balancing took more then 12 hours (315ah 16 cell battery, jk bms), 2 cells were off by more then 150 milivolts.. maybe 200 mv I don’t remember exactly.
OK you will find attached a zip file containing all the settings for both bms. Both have the same settings. At the time I took the pictures bat 0 (master) is the new one and bat 1 is older. Bat 0 is master.
The new one reaches 100% SoC with a voltage of 52.04 and a delta between lower and higher cells of 0.22 volts
The old one remains at 99%SoC with a voltage of 51.87 and a delta of 0,046 volts
In what I have seen, in this old battery cells 5 and 8 never exceed a voltage 3.4 volts. I believe that this is why this battery can never reach 100% soc.
And since battery old is never fully charged this could explain why both batteries can’t keep a SoC coherent when discharging.
Your setting are good, you may start balancing at 3.40Volts (I have set this value at 3.35 Volts, balancing trigger only 0.004 Volts)
Watch the YouTube video from “ Off grid Garage” date 30 jan 2024… especially for what is meant by the settings, watch Andy’s video from 25 jan 2026, this video is great about balancing a real bad battery with an external power supply.
The problem with jk bms of not being able to charge the battery to 100% is widespread
Thank you. Good advice. I spend hours watching the videos of Off Grid Garage. They are very interesting and most of my settings are inpired by what is proposed there. But so far I have no explanation to the fact that 2 or 3 cells in my older battery can’t exceed now a voltage of 3,440 volts. Is this because the bms can’t balance them properly or is it because these cells are ageing?
To fix this I need to dismount the cells, do a complete charge/discharge, balance them and remount everything. But if after doing all of this the problem appears again just because the cells are aged, perhaps it will be more efficient to replace these cells by new ones before doing all this process. But of course this will have a cost. Any comments?
55.2v is 3.445 per cell.
Screenshot 55v so 0.2 v lower than there. It may be the thing making the difference.
Did you do the voltage calibration on the jk bms to make sure it is reading correctly?
It could be saying its at 55v but is not actually there (verify with a multimeter)
Below are the settings of the charger of the Victron. I don’t use Victron Connect very often. So these parameters may have been there unchanged for a while.
Sorry for this question but, how do I actually calibrate the JK BMS?
Your 100% full indication in the bms settings is triggered bij the Voltage of first cell in the battery, if this voltage hits 3.449Volts soc reads 100%.
But if the first cell has the highest voltage, the total battery voltage is way less then 55,18 Volts, if the cell voltage is way less, charging is stopped because total voltage reach 55.2 Volts but cell number one still not trigger 100% soc in the bms readings.
Therefore it is important that the cells are very well balanced.
In the Victron Smart shunt 100% soc is done by measure the total battery voltage.
You can calibrate the Voltage reading of the bms by using the Voltage meter of an Victron Smart Shunt wich reading is of very high accuracy, or use a digital multi meter with equally high dc voltage accuracy, my meter has 0.02% accuracy, 50000 counts.
In the app screenshot the second line from below reads “ calibrating volt (v)” you can tap on the value to adjust the Voltage reading.
You can also calibrate currents with an equally high accuracy current meter (smart shunt), but is more complex to set up.
OK let me try to understand. Sorry if I’m too slow. In my pictures
I have set the Victron to “keep batteries charged”.
CVL is set to 55,2 V
Cell N° 1 was at 3,439 in my picture. But let’s assume it may have reach 3,45 (3,45*16 = 55,2 volts)
Then SoC 100% is supposed to be reached even if the other cells have a value lower than 3.45. Right?
This is the situation of my battery in the picture I sent where cells 5 and 8 and 9 are at 3,395 3,399 and 3,400 while the others are around 3,440 and the average is 3,449.
My expectation was that the BMS will “push” these 3 low cells so that they reach the same value than the others. This is the “job” of the balancing of the bms. But this doesn’t happen. These cells never excess a voltage of 3,400. The battery remains at these values forever and reports 99% charge.
I can force 100% charge by just setting off “charging float mode”. But this doesn’t explain why the bms is unable to push these cells. Are they defective? Am I just looking for a possible problem where in reality it doesn’t matter?
