.
Battery 3 (37%) is discharging while battery 4 (58%) is charging. (There was no PV nor grid input.)
.
As far as I learned in this forum, the SOC values don’t correspond with voltage but
are calculated by current flow. Is it right? Then actually battery 3 must have a higher voltage,
must be on a higher energy level than battery 4.
.
.
The following hours batteries 3 and 5 (all others shut off) were loaded by PV and - the illuminating effect ! - SOC suddenly jumps from 65 to 100%.
.
That means the calibration made ca. one week ago - with all six batteries together -
failed for battery 3 and 5.
.
I’m now about to drain these two to 0% (no PV or grid charging),
then, according to instructions, after cooling down, load them to 100%
(without AC output), switch on the other four batteries and - well, we’ll see.
.
Or is there a better idea?
Here are the results of the re-calibration of battery 3 and 5:
.
Charged (both batteries together) with 5-6 A.
Full (100%) after ca. 6 1/4 h.
Energy input: 5,88 kW
Ratio energy input / design capicity: 5,88/(2*3,072) = 95,7%
.
Then switched on all the other four batteries 1, 2, 4, 6.
AC, grid, PV all set OFF
.
Comparing SOC at the beginning and controlled after 9 hours
.
Battery 3 and 5 started discharging, but 3 soon changed to standby.
All other batteries charged.
After 9 hours (or even earlier) all batteries in standby.
SOC greatest difference: 55% vs. 93%
Battery 5 dropped from 100% to 55%, battery 3 remained at 100%
Overall SOC dropped from 75 to 74%. Roughly estimated auto consumption to be ca. 1% of 18kWh = 180 W,
or ca. 20 W / h (inverter + 6 batteries).
.
.
My conclusion:
If all batteries are in standby, voltage should be the same for all.
Then, SOC values are not reliable, calibrations (first one for all 6 batteries, second for battery 3 and 5) were not (very) successful.
.
.
Well, how to get better results?
.
I read in another thread, that batteries should be calibrated one by one, not all together.
.
Would be very time consuming.
Ok, charging could be done by plugging off the single battery from the whole system
and charge it with the specific charger.
But draining? The 12V DC allows only draining with max. 120W. If you add USB-C output, another 100W.
Would last ca. 30h (15h) to drain. And what 12V consumers with 100-120W?
I don’t have any and I know only expensive camping freezers, heaters etc.
.
.
What do you think?
the problem has been known for a long time.
Bluetti knows it too.
The more batteries are connected, the bigger the differences.
Ultimately, the total SOC is roughly accurate.
You just have to make sure that it doesn’t drop below about 30% and that it doesn’t suddenly jump to 0% if it continues to discharge, this switching the system off.
I’m not worried about the problem anymore and Bluetti won’t be able to help us with that either.
By the way, it’s wrong to calibrate each battery individually. Only if all the batteries are connected in parallel can they adapt to each other.
Don’t be sad.
Hi @digait, Thank you for getting back to us so quickly.
For this issue, we recommend changing the connection positions of the B300S. Try moving the originally closer one to the farther position and the farther one to the closer position, then test again. Does this affect the results?
I changed the positions of the batteries in the following way (see picture):
.
1 ↔ 3 (1 always with highest SOC, 3 with lowest)
2 ↔ 5 (2 second highest SOC, 5 with second lowest)
4 ↔ 6
.
There was no effect: Battery 1 remained the one with the highest SOC, 3 the one with the lowest SOC, and so on.
.
Then I tried changing the plugs for the two cables which are connected to the inverter (green double arrow).
Again: no effect.
It depends on your particular case how much energy (or money) is wasted by calibration.
.
Draining to 0% does not mean all energy will be wasted.
In two rooms, I used two electric heaters and turned off the central heating (in addition to the continous consumers).
.
Then charging to 100%. One part came from PV, the other from the grid.
But in my case ca. half of the current from grid was produced by another PV plant.
Thus, I can say, total calibration costs were about 0,50 - 1,00 EUR.
Hi @digait, We apologize for the delayed response due to the holiday.
We would like to confirm one more thing. When a single battery pack is out of power, but the entire system still has power, will that battery pack automatically turn off? We will then further investigate this matter.
I have an AC300, and have had it for about 3 years running various critical loads in my home. I have done a full calibration twice over those years, and it was due to both times I purchased another B300 battery to add to the system. So I started with 2 batteries, and on two occasions I added another battery. The method of doing the calibration is very important if you want the BMS of each battery to learn the cell state of charge, so I will detail how I did it, and I will state that it did improve the SoC readings, but on days where there is not enough solar to charge to 100% the next day(s) that SoC will vary more until another 100% charged state occurs.
1. Do a full charge to 100%. This particular charge cycle is not a critical one, that will happen later.
2. Discharge cycle. Disconnect all normal loads. This discharge should be done at a steady rate so that the BMS learns how the cells discharges. Since my AC300’s max output is 3000 watts, I use a 1500 watt electric heater for this. I plug that it and let it run until the whole system shuts off. Having the system turn itself off is a critical step. If the BMS “thinks” the cells are empty, but they are not, it must learn what true empty is. Note that the cells are never fully drained, however, it keeps a minimum charge to prevent damage.
3. Allow the system to cool down from the discharge.
4. Charge cycle. Disconnect all loads, turn AC and DC outputs off. This charge should be done at a steady rate so that the BMS learns how the cells charge. For this case I configure my system to “Normal UPS” mode so it fully charges. I set the maximum input current to 15 amps (I have 20 amp circuits, if yours are 15 amps set to between 10 to 12 amps). Let the system charge until it actually says that the input power is 0 watts. This allows the BMS of each battery to learn what true full charge is.
It was not clear to me if that was how you did the recalibration, so I thought it would be worth sharing, and for others reading this post. This has helped my system to be better aligned with individual SoC levels. I will also note that the two original batteries are always very similar in SoC, and the other two I purchased at separate times are the two showing different SoC levels. Those last two batteries are both different in their hardware from each other and the original two, and are using different firmware even though all are at the latest available versions.
Appreciate your calibration details and I am wondering if I need to do the same? Have had an AC500 + two B300S connected for a number of months. The %SOC shown on the AC500’s home screen closely matched the average of the two B300S SOCs as shown on the battery product screen. Also, both of the individual battery SOCs always remained very close to each other, normally within 2% during charging & discharging.
Just added two additional B300S units to my existing ones, so now I have the AC500 + four B300S connected. The two new batteries were both showing about 62% SOC when first connected (as received), the two original B300S units were both at 95%, and the %SOC on the home screen was 47% (lower than I expected). Connected the AC500 to the grid and starting charging the entire system at a rate of about 825 watts. No loads were connected to the AC500 and AC and DC outputs were both OFF. After two or three hours, the two original B300S hit 100% and went into standby mode while the two new B300S units continued charging for another 3 or 4 hours before they both got to 100% and standby. At this point, the battery product screen shows all 4 batteries at 100% and in standby, and grid charging current is zero. The five LEDs on each of the B300S units are all lighted and not blinking, so it seems clear that the batteries are all fully charged. However, the %SOC display on the home screen is showing exactly 50%.
Correct me if I’m wrong, but I’m assuming I now need to perform a recalibration in order to get the home screen’s %SOC to 100% where it should be, right? I need to keep all four batteries connected as I discharge and then recharge them using the AC500, correct? In case it matters, both my original B300S units’ software are at 1022.08 while the two new ones are both 1022.13. The AC500’s DSP=4045.04, ARM=4044.04, and HMI=6051.04.
And one last question - how long did you let the system cool down after completing the discharge? You were using B300 units, but I suspect the cool down rate for the B300S is likely the same. Thank you for your time and any suggestions you may have.
@TMan, I do think it would be worthwhile to do the calibration since you now have the two extra batteries as they will need to learn their own charge/discharge process.
I am a bit baffled about the odd SoC reading from your AC500 as it appears to show about half the expected percentage. I have not had my AC300 do that. So, if it were my system, I would shut it fully down. Confirm all the battery cables are fully seated with the slide switches fully closed, and then power it up. If it continued to show an erroneous SoC percentage I would go ahead and do the full calibration cycle and see how the system is working then.
If it continues to show the wrong SoC, then I would put the question to Bluetti support.
Hi @TMan, Based on your description, we believe your original AC500 + 2 B300S system was working fine.
The issue may be with the two newly purchased B300S batteries. Generally, we recommend customers charge new batteries separately upon receipt. This is because the batteries may have been in long sea transit and stored in warehouses, which can cause them to discharge slowly, making the SOC (State of Charge) potentially inaccurate.
Could you try recalibrating by connecting the two new B300S batteries? If the AC500 displays the correct battery levels for the new batteries, then you can reconnect the two old batteries.
Typically, we recommend that customers fully discharge the batteries, then power off and let the system rest for 3 hours to allow for cooling. However, if time permits, letting it rest for 12 hours overnight may yield even better results.
@hnymann’s kind advice is very helpful! Yes, please make sure locked the cables correctly.
Yesterday I did connect the two newly purchased batteries to the AC500 by themselves after disconnecting my two older ones and the Home SOC display matched the average of the two new batteries’ SOCs as shown on the battery product screen. As soon as I connected a third battery (whether the two new ones + one older one, or the two older ones + one new one), the Home screen SOC display became inaccurate. I didn’t think to try one older battery + one new one together; don’t know what that would have shown.
I’ve started recalibrating the two new batteries one at a time and I’ve already noticed that the rate the SOC is decreasing on the first battery is occurring faster than it should be with my 630W electric heater load, so I’m hoping the calibrations will help.
One question about your reply above. I’ve seen cautionary notes not to let discharged batteries sit around too long (more than 3 hours) after they shut down from a re-calibration discharge before beginning the recharge cycle. Are you saying that it’s ok to wait longer (up to 12 hours) before commencing the recharge? I plan to disconnect the discharged battery from the AC500 and use my T500 to charge it since that’s about as stable a charging rate as I can do. Thanks.
Hi @TMan, Regarding this issue, we recommend upgrading all B300S batteries to the latest BMS version. This may help maintain SOC accuracy and consistency. Do you agree?
Please ensure that only one B300S battery is connected during each upgrade.
