I recently installed a Multiplus II 5000 with two Pytes 48100R batteries and configured it as specified in Victron & Pytes [Victron Energy].
Everything is running fine so far, but when I reduce the minimum SoC to 5% or 0% (from the default 10%) the Multiplus starts charging the batteries from the grid with full power as soon as the SoC drops below ~10%.
Then for sure the developers though that less than 10% is not recommended at all and for sure is not good for any battery, regardless of chemistry.
I, for one, never went so low and didn’t know that also the “without” is behaving the same way.
It’s not recommended to go below 10% regularly.
The voltage of a LiFePO4 battery starts to drop faster below 15-10%.
Maybe the low voltage triggers the charging?
What battery do you have? What are the VE.config settings?
I decreased the minimum SoC below 10% at 20:30 and it immediately starts discharging the battery for ~15 minutes. Then re-charge at full-current kicks in. The voltage at that point is 49.9 V which is above all thresholds.
Same Problem here.
I have a Multiplus II 48/3000 with two Pytes 48100R
When using ESS it shows the same behavior as above.
Only way to go below 10 % is to set ESS to external control and use MQTT/DBus to force the Multiplus to keep putting out energy to the grid.
Very weird problem. When i set the SOC limit to 5 % or 0 % i actually expect it to got there.
also it seems to ignore the setting made in the veconfigure tab called “inverter”
there it set the input low shut-down to 49V but it ignored it.
it seems as if the ESS assistant completely overwrites some settings.
But despite beeing not recommended, its very strange that it is not possible.
I think there is a difference between the BMS signaling a charge would be good and the BMS blocking discharge.
Because the current limitation to 10% triggers a cycle of discharge, charge and discharge which cycles the battery. so either disabling min SOC to go below 10% in the settings slider or a just idling would be much better.
but the current behavior might even be worse than just going down to the user wanted 0% at the moment it will cycle around 10% and also throw off the dynamic ESS because that thinks it has 10% left.
i am manually discharging via external control because the ESS will cycle at 10%, thats my problem
i have set the SOC limit to 0% (regardless if its recommended or not) and the dynamic ESS thinks it has more energy to play with than it actually gets. this will then start cycling at 10%.
and i am interested in what blocks going below 10% as voltage of the battery was well above the limits at 10% (the screenshot is well below 10%)
Disconnect the battery bms control and you will work out fairly quickly if it is battery comms or not.
Also will find out quickly if you should be doing that.
The cycling is really a problem and it looks to me like it actually is not connected to the SoC, but to the voltage. For me it always happens at 50V and this is a voltage which I did not put in any Victron settings.
Depending on the load, the voltage will be reached sometimes already at 13% for me.
It will be voltage, since voltage drops under load.
And voltage is fairly important at a cell level as well.
49v is the setting if the battery allows it.
The first one shows the SOC over time and the charge requested FLAG from the BMS.
In the red boxes, the SOC Limit was set to 5%.
In the green box the SOC Limit was set to 10%
Between the red boxes i did manual discharge to almost empty.
As you can see, the charge request FLAG is set when going from 10 % to 9 % SOC
The next one shows the corresponding pack voltage.
The fact that it is near 50V everytime can also well be because the BMS Software has this weird waiting phase at 13 % SOC when it then jumps to 10 %
Maybe its doing some recalibration there and this just happens to be triggered at 50 V because this might be known as 10 % SOC
The last one shows the actual current together with the allowed charge and discharge currents.
===========
So what would be nice is to know if the ESS Assistant “listens” to the charge request. At least its what i suspect from the data. I might testing it by disconnecting the BMS CAN as suggested.
Also it would be nice if this can be disabled. Or at least it might be usefull for other users in the future to be notified or prevented from setting the SOC limit to low. But i really dont see a reason for this restriction. LiFePo4 should be fine with a lower voltage limit of 3V which would be 48V on a pytes 48100R with 16S configuration. Which is around 1-2 % SOC when i look at my plots.
It cannin DVCC. Remove controlling bms. You can still leave it physically connected and see data.
Be aware though that mppts and inverter may switch off or need a reset.
Plot the cell min and max. Overall pack voltage may be ok but you have a weak cell. This may be the other difference in the behaviour.
