I am running a Victron Multiplus 2 48/5000 in an off-grid configuration.
Battery pack 16x EVE MB31, in a Yixiang battery box, with JK Inverter BMS in closed loop setting (200A, 2A active balancer, firmware 15.35). 314 Ah rated, each cell was tested as 330+ Ah as per excel sheet I received from supplier. I have no way of accurately measuring actual capacity, but performance so far is spot on.
Charged by Victron MPPT 150/35 and AC coupled (on AC out) SunnyBoy 3000TL.
I took most of my hints from Andy (Off-Grid-Garage). Goal was a simple set & forget solution.
At the same time I am a bit of a statistics nerd and have integrated everything with Home Assistant. That's where the following chart comes from. Typical charge session on a sunny summer day with no big loads.
Most of the settings are as per Andy. Charge to 3.45V per cell (Cell RCV). Stay there for 1 hour, with balance also starting at 3.45V. Then drop to 3.35 for float (Cell RFV). That drop reduces SOC from 100% to 98%.
Changes compared to Andy:
1) I increased float time to 12 hours (up from 6 in Andy's settings) to stay in float for the rest of the day and avoid a second 'afternoon' peak. Battery capacity is plenty for my use, I don't need 100% for the night.
2) I decreased balance trigger voltage difference to 5mV (from 10mV in Andy's settings) since cell voltages were very close already after a couple of days of operation. Now typical cell voltage difference at float is 2 to 3mV.
3) I limit charge current to 30A (0.1C) since that will very comfortably fill my battery (at least now, in summer). And I drop the charge current limit to 15A once battery voltage gets over 54V, a minimal slow-down for the the last ~1% of SOC.
Observations:
4) With the current 15.35 firmware, the BMS will set SOC to 100% at the end of the absorption time some days, and not on others. Maybe loads pull down the voltage just below the cut off sometimes (SOC-100% Volt is set at 3.449, as per Andy). Not a big deal for me.
5) There doesn't seem to be much a of tail current once the 55.2V target is reached. Certainly within 15 minutes current is basically zero.
Questions to the audience:
6) Since nothing much happens after the first 15 minutes at 55.2V (no tail current, no balancing), any reason to keep voltage there for a whole hour? I am considering a reduction to 0.25h for RCV time.
7) How often does a BMS break and needs replacing? Anyone had a JK Inverter BMS hardware fault? Would it be worthwhile to get a spare BMS to have on hand for when the current one dies?
8) Victron offers a warranty extension from 5 years to 10 years for 10% of the purchase price. Worthwhile investment or waste of money? Has anyone had Victron components fail between 5 and 10 years of age?
Battery pack 16x EVE MB31, in a Yixiang battery box, with JK Inverter BMS in closed loop setting (200A, 2A active balancer, firmware 15.35). 314 Ah rated, each cell was tested as 330+ Ah as per excel sheet I received from supplier. I have no way of accurately measuring actual capacity, but performance so far is spot on.
Charged by Victron MPPT 150/35 and AC coupled (on AC out) SunnyBoy 3000TL.
I took most of my hints from Andy (Off-Grid-Garage). Goal was a simple set & forget solution.
At the same time I am a bit of a statistics nerd and have integrated everything with Home Assistant. That's where the following chart comes from. Typical charge session on a sunny summer day with no big loads.
Most of the settings are as per Andy. Charge to 3.45V per cell (Cell RCV). Stay there for 1 hour, with balance also starting at 3.45V. Then drop to 3.35 for float (Cell RFV). That drop reduces SOC from 100% to 98%.
Changes compared to Andy:
1) I increased float time to 12 hours (up from 6 in Andy's settings) to stay in float for the rest of the day and avoid a second 'afternoon' peak. Battery capacity is plenty for my use, I don't need 100% for the night.
2) I decreased balance trigger voltage difference to 5mV (from 10mV in Andy's settings) since cell voltages were very close already after a couple of days of operation. Now typical cell voltage difference at float is 2 to 3mV.
3) I limit charge current to 30A (0.1C) since that will very comfortably fill my battery (at least now, in summer). And I drop the charge current limit to 15A once battery voltage gets over 54V, a minimal slow-down for the the last ~1% of SOC.
Observations:
4) With the current 15.35 firmware, the BMS will set SOC to 100% at the end of the absorption time some days, and not on others. Maybe loads pull down the voltage just below the cut off sometimes (SOC-100% Volt is set at 3.449, as per Andy). Not a big deal for me.
5) There doesn't seem to be much a of tail current once the 55.2V target is reached. Certainly within 15 minutes current is basically zero.
Questions to the audience:
6) Since nothing much happens after the first 15 minutes at 55.2V (no tail current, no balancing), any reason to keep voltage there for a whole hour? I am considering a reduction to 0.25h for RCV time.
7) How often does a BMS break and needs replacing? Anyone had a JK Inverter BMS hardware fault? Would it be worthwhile to get a spare BMS to have on hand for when the current one dies?
8) Victron offers a warranty extension from 5 years to 10 years for 10% of the purchase price. Worthwhile investment or waste of money? Has anyone had Victron components fail between 5 and 10 years of age?
