Sure, I’m not in any particular rush. I’d rather properly understand what’s going on. If you need any additional data from my side, feel free to ask — I’ll provide whatever I can.
This morning I worked with a colleague to see what is going on.
There is a step in the logic of the Multi where it will not raise the charge voltage (for the solar chargers) while it is still charging from AC-in.
In a system with a PV-inverter on the input side, like yours, it will be importing that energy and charging the battery. So the system is charging from AC-in.
Once the battery gets full enough, so that it stops accepting charge current, the Multi will then raise the charge voltage. The effect is that the DC-coupled solar charger can stop charging for a period of time while the system charges with only the AC-coupled PV.
Once the battery is fully charged, the situation will usually resolve. The larger the battery, the longer that will take.
It also appears to depend on minute calibration differences (beyond what we can synchronise), which makes the solar charger back off at the charge voltage, while the Multi continues to charge from AC.
Once you get past this hurdle, the system will feed DC-coupled PV into the grid for the rest of the day, unless a large load or loss of production causes the charge state to return to bulk.
So we understand now why this happens, but I don’t yet know what to do about that.
What you can do right now: You could move (some of) the AC-coupled PV to AC-out. Or you can move more of the DC-coupled PV to AC-coupled, as you suggested.
These are workarounds, but I didn’t want to leave you hanging while we work this out.
OK, so I understand that we’ve confirmed this is not correct system behavior. For now, we know there is an issue, but we don’t know where exactly it is or how to fix it yet — is that right? Did I understand correctly?
In general, my idea was actually to move all PV over to MPPT charge controllers — not the other way around. The question is: if I disable charging on the Multi side, would that solve the problem? As far as I’m concerned, energy from AC-coupled PV can be fully exported.
The only question is whether what I currently have on the MPPT side will be enough to recharge the battery.
Do we have any kind of timeline for fixing this issue?
In the meantime, I’ll reconfigure the system so that all AC-side PV operates on the AC-Out side.
EDIT: All PV AC Inverters are operating on AC-Out from now on.
Yes.
That’s also a good idea. For a short period of time, the PV connected on the DC side waits for the PV connected on AC-in to finish charging the battery. If you move everything to the DC side, this immediately goes away.
DC charging also becomes a lot more efficient.
Yes, I know that charging on the DC side is much more efficient. The issue is actually using that energy — I have to go through the Multi for that, and they’re not particularly efficient on their own, especially under higher loads.
The Hoymiles microinverters I use on the AC side of my installation are much, much more efficient in that regard.
On top of that, in the event of a power outage, microinverters connected to AC-Out effectively increase the available AC power of the system. Not to mention that exporting 10 kW of power via the microinverters is completely silent, whereas my Multi setup already becomes quite noisy at around 4 kW.
Indeed. There are tradeoffs for both.
The benefit of a full victron based charger config is better control.