I have been trying a daily absorb these last 2 days but with 10 minute absorption time as that would be better when I am cruising with shorter engine run periods. Yesterday and today I had over 3 hour absorption times, c!early the rules are not working, or are too rigidly coded or the user settings are having no effect and the code only uses default values. I will pass my experience on.
Tried to upload some images but they will not upload.
Surely Victron could do something similar to what my REC-BMS does? That holds a float CVL of 54V (charging from solar here, about 99% SoC) until the SoC drops below 95%, then the CVL is upped to 57.3V. When it reaches this itâs held there until cell imbalance falls to 10mV, then SoC is reset to 100% and charging is stopped, CVL is reset to 54V, and the battery voltage drops back until it gets back to the 54V float voltage.
The attached shows a ânormalâ cell balancing/SoC reset cycle where it held at the balancing voltage for about an hour (cell imbalance dropped from 40mV to 10mV), this typically happens every couple of weeks. After a long time out cruising without a full charge (several weeks, maybe 8C throughput in and out), the maximum imbalance I saw was 150mV and it took about 14 hours to get back to 10mV and stop balancing, see second plot.
The relatively long times are because the balancing current is a bit less than 1A and the cells are 700Ah, so 14h is about 2% capacity imbalance (0.25% per 100% DoD). AFAIK the Victron NG batteries have considerably bigger balancing current compared to capacity, so should do all this more quickly.
Yes, that is what they are meant to do, but the end of absorption / balancing phase is not stopping correctly when cells are balanced or if the time limit is reached. The settings in the Lynx BMS are not user accessible.
@hennephuis finally uploaded the screenshots from an absorb set at 10 mins with balanced cells after current reduces.
Thanks for the graph. I can tell from it that the battery cells were balanced way before 12 oâclock but the voltage remains at 3.5V until 14 oâclockâish.
Just to be sure: Victron does not detect balanced battery cells and does not use that to terminate absorption. No surprise here. @iandâs suggestion is key here, although I would probably do it slightly differently.
It also seems like more than 10 minutes absorbing (duh). Not good. That is because it misses its timer as a result of the ripple (my hypothesis).
Do I understand correctly that you (@pwfarnell) are now also confronted with the same problem as @zedamoca and me, @hennephuis?
Here is a graph from today. It initiated the first charge cycle after I set the absorption voltage to 55V and the monitoring voltage to 54.9.
It raises the voltage to 55V around 13hâish and very soon it syncâed to 100% charge. This is what I hoped and expected.
Then it keeps the voltage at 55V forever, at least till 18h, also as expected, but not what it should do. I expect it to lower the voltage when the cells are balanced. They were balanced all the time as one can tell from the min/max cell voltages.
I think that both bugs are (again) clearly demonstrated. At the same time, there is a third bug, namely that I lowered the SoC level at which it must start a full charge cycle to 50%, but it still does so at 70%. Most likely because this setting is duplicated and not synchronized. Maybe that is already a topic; I will look for it.
Yes, I have had longer absorption times than set when on alternator charging. I have increased my SOC threshold to get an earlier absorption and that works and reset to 70% OK but I have not tried a lower SOC threshold.
Today was a funny day.
I took half of my batteries of service yesterday evening because I was relocating them, and didnât finish yesterday.
This is the new place 
So, because there was half of the capacity, this night (and morning) they went to 59,5% SoC when my setting to start a new absorpion cycle is 60%
The BMS here started a charge cycle (see my previous post) and I also have more to tell about it:
At some point the MPPT could no longer deliver all the power that was needed and the voltage dropped. However after a few minutes, the MPPT could again power everything and the voltage raised to absorption voltage again. So it seems quite persistent in pursuing absorption time.
At the end of the day the voltage finally dropped to float, but the next day, it rose to absorption voltage again, although the battery SoC did not drop under 70% (not even close). So it really seems very, very persistent.
Now this could very well be a side effect of setting the absorption voltage to only 55V. Victron warns for that and I agree with it. It does not come a a surprise to me. If-then-else âcontrollersâ with hard parameters are very fragile and, as I said before, are very hard to understand, if at all possible. Side effects are imminent.
I think I let it rest for now as I can no longer add meaningful information to my hypothesis that the controller looses track of its timers because of the DC ripple. That seems a bug to me (causes by ad hoc if-then-else rules instead of solid signal processing; but that is my opinion).
Then there is the problem that absorption time is not related to cell balancing and that installers just must continuously monitor usages patterns and adjust absorption time by hand. Making absorption time end when cells are balanced seems more like a feature request.
I do not regret buying these batteries yet (like @zedamoca), but I have a couple of systems lined up in the planning and Victronâs response to this matter is a very important factor in deciding how to continue with those.
If you have set a 55V limit in DVCC for managed battery charge voltage then absorption will never occur and it will.keep asking for it day after day continuously until it meets it. I have seen where the BMS CVL is at 14.0V for 36hours because solar was not enough on one day to recharge the overnight. The SOC threshold does not get reset daily, it is a one off trigger.
If the absorption timer and sync worked OK then you the Lynx BMS needs to be left to run at 14/28/56V for absorption. The way it works with fixed hidden voltages means that that adjusting charge voltage via DVCC will stop the BMS absorption logic from working (assuming correct functionality).
I said that I would feedback to Victron again and have been told that the absorption/synchronisation issue is being worked on and the next firmware update should address this.
yes today its still asking CVL at 56V
its limited at 54,9V
I had it days in a row at 56V, I donât want my cells long time at that voltage.
Whatever anyone might say, itâs my cells and I should be able to make a more conservative configuration than whoever engineer say.
I agree in not letting making them less conservative, but more conservative, I canât agree being denied to. This is why I regret these batteries and I wasnât informed of these limitations before buying.
I have this system for 9 months now, is this not enough time to fix it? An official position from Victron is very much needed.
All the rest of the Victron system Iâd recommend every single time.
I could agree with this, but if the lowest and the highest cell are within a certain range, in other words, the cells are balanced, absorption must stop!
Odd, but I can agree with daily absorption if the absorption is seen finished (cells balanced)
Absorption is only needed to balance. Instead of whatever complicated algorithms are used to find it, just look and max and min cell voltage. KISS
Yeah, thought as much. Perhaps some hardcoded 56âs here and there.
Thanks for the update from Victron! Weâll look forward to the next release.
Please note that new firmware has just been released for the Lynx BMS NG.
The chsngelog is here
The firmware can be either
In reference to the discussions we have had this firmware has the following changes (amongst others).
thank you very much for bringing this to our attention.
I have updated my personal BMS and reset the voltages back to defaults.
Iâll let you know how it works out the coming days.
Erik
Here was my charge from end June with 3 hrs absorption even though cells wre balanced.
Here is the performance with new firmware with the same settings, stops absorption quickly even though the voltage was moving around a bit.
This is encouraging. I need to find focus time to do this in the next few days.
Complete change log for the ones that canât read docx. I still donât understand why arenât they in pdf.
Thank you for your report. Lucky you!
Here I was unlucky as I forgot to set the âbattery chargedâ voltage back to 55.6V. So it remained 54.9 (my work-around).
But⌠I unknowingly performed an interesting test.
âcharged voltageâ a.k.n.a âmonitoring voltageâ: 54.9
âcharging voltageâ: 55
As can be seen below (with magnifiers 
) the voltage reached 55V and stayed there the whole day.
It did not sync.
Although I thougth I set the charging voltage back to 56V, apparently I also forgot that, as it did not go beyond 55V.
Regardless the wrong values, it still would have syncâed and balanced shortly because as per the spec, 54.9 and 55 are both valid values.
However its behaviour stays the same: unwanted long absorption. What did change however is that there is no syncing any more.
This canât be true, so I will be running the test again.
I suggest you remove any other limits and let the BMS do its thing to check it works for you. Then if you have alternative limits that do not work then the issue can be reported back.
Using CVL will prevent proper end of absorption. Aldo, note Victron are recommending charged boltsge setting 0.4V below charging voltage.
