I’ve had a small system running with a 24V 1200VA inverter for a couple of seasons, but was having issues with the battery array getting unbalanced (2 parallel, 2 series, 100Ah) such that either the upper or lower batteries would fully charge and go into protect mode and the other end batteries wouldn’t fully charge. My bright idea was to convert this to a 48V, 100Ah system and bought a 500VA, 48V inverter. No drama getting it set up and adjusted (have a Victron charge controller and battery monitor as well). Next day, when the load gets switched on, the related computers (RPi’s) reboot, indicating that the AC output drops enough to cause the issue. The actual load is an aeration pump that draws 220W when running. The computers draw maybe another 40W max. The surge rating on the inverter seems plenty high to handle the motor start-up, but it seems not to be the case. One of the RPi’s is running Venus OS, and there aren’t any events logged, but given it’s losing power maybe not surprising. I can try to find a bigger inverter if that’s the solution, but open to suggestions.
Thx, Chris
Motor start current can easily be 6 times rated current or higher, so 1.5hp = 1125W. A 500VA inverter is nominally 400W so it is too small for the surge. A 1200VA inverter is nominally 960W so could handle it. You need a larger inverter or a soft start.
You should also look into battery balancers, a string of basic non communicating 12V lithium batteries in series can go out of balance.
Thanks. I can put it back to 24V and use the inverter I had, then add the balancer and see how it goes.
OK, so put the back to 24V with the 1200VA inverter, and added a Victron battery balancer. One thing I missed is that this balancer only has 1A of correction. I’ve got 400W panels, and control when the load turns on/off with an RPi. Full sun + 60 SoC starts the pump. By the afternoon all the batteries are charged to >98%. Once the PV power drops, the batteries provide the power. The issue is still that the batteries is protect mode may stay that way and the charge delta goes off the rails. I shut off the pump when the SoC delta is >20%. I’m trying to come up with ways to get all 4 batteries to turn on. At the moment, the pairs are connected at the negative and positive ends, a well as in the middle with heavy copper bars. Would removing the middle bar and getting a second balancer help or hurt? I’d hate to start adding relays and series loads to “encourage” the disabled batteries to turn on. Any other ideas are welcome.
Thx, Chris
So you got a 24V 200Ah battery bank that cant support 250W? Sounds like the main issue to me. And the one you need to figure out first, why do the BMS switch off.
I personally would not interconnect the middle points. Leaving that in can lead to imbalanced charge/discharge. The two batteries in series act as one 24V battery, and should be treated as such.
To your first comment; I’m confident the 24V 200Ah bank can handle the 250W load. Not quite 24x7, but close enough. The BMS switches off to; prevent over-charging, charging when cold, when the batteries are drained. For me, it’s the shut-off when charged. If just using a single series pair, probably not an issue. Pulling a load turns them on just fine. In this 2sx2p array, one or another may not turn on when needed because the load isn’t high enough to do so (the other parallel battery handles the load). The batteries individually provide detailed state data over Bluetooth, and I’ve got a Victron battery monitor, and the charge controller and inverter provide details, so there’s plenty of data to mess with. As to your last point, I understand that we well, hence the question about the mid connection. I suppose it forces the same mid-voltage on both sides, but given the issues seems reasonable to remove to see what happens.
Thx, Chris
But thats an issue. That is not normal behaviour. The BMS switching off should be the last resort, not a common thing. You need to check your charge parameters
Granted. I’ll check with the battery specs and the vendor, as well as the charge controller settings.
Thx, Chris
The usual advice is to reduce absorption voltage on the chargers to avoid the high voltage cut off from triggering. There is very little extra energy stored between 28.0V and 29.0V. The only reason for pushing higher voltage is if balancing is required.
So, I confirmed that the charge (absorption) voltage on the charge controller matches the low end of the charge range specified by the battery vendor (14.2V for 12V system, 28.4V in this case). The battery vendor was adamant that the mid-connection remain: “The mid-point connection between the two series pairs is crucial for maintaining the correct voltage reference and ensuring balanced charging and discharging across the series string.” No mention related to a second battery balancer (one for each pair). I’m logging and graphing plenty of data from all components, so I can see where things go sideways with the batteries. Today, we have clouds, and the load has yet to be enabled, and the charge disparities seem to have settled. I can only assume the balancer is doing it’s thing. That device doesn’t have comms. I’ll quit updating here and open a new thread if the need arrises.
Thx, Chris