Hi folks, I’m running an off-grid system for a cabin and after a all-in-one setup for a competitor failed, I moved to the Victron side of things. It’s gotten a whole more complicated though, and I need some guidance. In addition to an incentive to clean up my connections.
My setup:
3x4 panels combined in a breaker box to 99V - 38 A
SmartSolar MPPT 150/70-Tr VE.Can
Cerbo GX
Discover AES 7.4kWh 48VDC Lithium Ion LiFePO4 Battery
MultiPlus-II 48/3000/35-50 120V
My problems:
Minor problem: I haven’t been able to get the calculation for battery charge level to work. I set it once, it worked, then changed a setting and it no longer goes below 95%. It charges up to 100%, but does not show any depletion past 95%.
Less minor problem: I’m getting near daily overload warnings and regular High DC Ripple warning, usually at a time when the battery gets to full. The Ripple warnings turn off my system.
Having a SmartShunt to monitor that battery will give you accurate State of Charge numbers.
Unless you can figure out if you can get the Cerbo GX communicate with the battery closed loop. This link may help (it is for their rack mount batteries, so maybe, maybe not).
Check your Discover AES documentation to see if it has Victron communication.
Based on the charts you’ve shown, I’m guessing that the BMS is disconnecting the battery at 56.7 or 8 volts, and when it does that, the Multiplus is effectively disconnected and you get ripple and overload. I would adjust the absorption voltage way down to like 55.2V and see what happens, then gradually increase it.
Sorry, but are you sure about that?
It seems, from the first picture that the overload, as time when happening, it has nothing to do with the full charge and battery disconnecting…
The same for the second event…
See below what I mean…
For the same picture, may be a coincidence for the first event, but consequent ones…
I have the feeling that another could be the reason…
Maybe some AC-IN and OUT events.
Definitely not sure! We need to see the battery full voltage/current zoomed in to the minute compare to the VRM alarm log with the time index to know for sure.
OP: If the ripple and overload are an AC load problem, then this should happen whether the battery is full or not. What is the largest load you have switching on? hvac with high inrush current? What is the length, size, and condition of the battery cables you have going to the multiplus? They look pretty short in the photo. Lugs torqued down?
Here are the details for the last two ripple incidents, let me know if I should include any other graphs. Neither of those had a large load connecting at that time, but I do run a portable dehumidifier these days.
Cables are in good shape, three years old, I think 2AWG, and 2 feet / 60cm long. I do have them stacked at the battery lugs though (see pic)
I assume you do not have closed-loop comms between the battery and Multiplus.
If you did, you could show a chart of CCL and CVL at the same time and see the BMS disconnecting.
If you had access to the BMS via a laptop you could also see what it is doing at that time.
On these charts, it looks like the low battery and high ripple alarms go off when the current goes to 0 on the multiplus. If the BMS is disconnecting the battery, this could happen.
I’ll go with my original suggestion: Change Absorption to 55.2V and float to 54.0 V and see if you still have the problem.
Since you do not have a SmartShunt, and you do not have comms with the battery, I assume that the Multiplus is acting as the battery monitor. A SmartShunt is highly recommended in this situation, because the multiplus cannot see the voltage closer to the battery after cable voltage drop.
I don’t think the DVCC settings will matter, SVS and SCS should be on so your MPPT and multiplus do not over-current the battery.
You have no way to know how much current your battery is getting during charging without a SmartShunt or the BMS, so beware.
Without BMS comms DVCC won’t work properly, as this depends on the BMS for limit data.
Best to go through the devices and ensure that the ABS voltages are 200 - 400mV BELOW the battery manufacturer’s recommended value. (ie 56 -56.6V) Float should be another 500mV less than this. Smart shunt is also a good idea, multiplus is not too accurate at voltage measurement.
I changed the absorption and float parameters to 55.2 and 54.0V respectively and no overload today, for the first time in a while. I’ll report back after the week-end to let you know if it holds.
Great!
If you think you are not getting enough charge into the battery at the top end, or if balancing is insufficient, raise the absorption 0.1v per day until the overloads return, then backoff. Or just leave it where it is and move on.
Looks like you have a load that turns on regularly which is causing these warnings.
Perhaps a compressor. This is not uncommon. Many people ignore these warnings.
If you have a clamp meter that can measure inrush, you can find out how much instantaneous current is needed by the device. VRM does not normally sample at the right instant in time to catch these transitory surges.
Sometimes the temperature of the multi is causing derating. Is it mounted vertically in a cool location?
It might be a spike in the dehumidifier. I’ll turn it off for the middle of the day and see if it’s the culprit.
The multi is mounted vertically against a north facing wall in a spacious enclosed cabinet in the shade. It is Hawaii warm (80F +) but it’s not in an extreme temperature.
