Hi all,
I have just configured 2x multiplus II as a parallel system and I believe they are working. However, the remote console readout is not correct: the AC Loads reading appears to be approximately twice the actual value and for some reason the battery is being consumed even though there is a scheduled charge in place which should stop the battery from being discharged.
Any ideas what might be wrong ?
Would need more details here.
Are you meaning it’s 2000W rather than 1000W, or are you talking about 25 vs 50W?
Where do you see this difference? And what is your configuration?
In my case I have a seplos multi battery box setup, and with “has dc system” turned on, it shows charging as 1 box charging and the other boxes as consumption, as it reported the kWh of only one box.
So the “has a dc system” should only be used if your battery kWh metering is correct and you have a real DC (positive or negative) load not covered by sensors that is worth reporting.
I am asking because you use the term consumed. We need more information. (Battery types/configuration, MP2 hookup, and exactly what do you see what you think is wrong).
As for parallel system: do make sure you’ve connected all AC-IN, OUT and OUT2 parallel. The second MPII is like a “dumb brother” of the master, as the master dictates all signaling and sensoring and the dumb brother just applies the signalling (like the PWM of the AC-OUT) the master dictates.
Its 2x 5000w multiplus with 4x Pylontech total 9.6kwh. No DC loads.
For instance “AC Loads” on the remote console and VRM were last night 7000 watts and “Grid” was showing 3500 approximately and battery was discharging at 1500w.
This morning it’s a bit different as “Ac Loads” is around 1500 and Grid around 1000 with battery charging currently at about 200w which corresponds with the amount of solar.
I have had a thought that I should untick “has external current sensor” in the slave config. Could that be it ?
I have unticked the “external current sensor” box on VEconfig for the slave and it has changed the behaviour although still not quite right.
First of all it looked good as the grid and solar/battery values added up to the AC loads value then a large load came on and it stopped making sense. The Grid value went to 4kw with the Ac loads showing only 2.5kw.
It’s bizarre…the AC loads value seems to be 50% higher than the Grid when battery is completely off so just doesn’t add up.
So, on the slave Multiplus I removed the shunt for external/internal current measurement and the readings are pretty much correct, only the Grid is lower than the AC loads sometimes by a few tens of watts so not bad really.
So, config only the master “external current sensor” and connect current sensor to master and put nothing at all in the slave connection.
The VRM/Remote console readout is fixed but I realised I still have the scheduled charge issue as in the system still draws from the battery when a scheduled charge is active which it didn’t do before. Basically it seems to be trying to split the draw equally between battery and grid when on a scheduled charge.
Perhaps a separate post would be better.
Slight error in my last post, the Remote console readout isn’t quite fixed as the current under AC loads is approximately twice the current under Grid. The watts and volts on both those parts of the readout seem fine though. This seems very likely linked to my other problem of half power being drawn from the battery when it is on a “scheduled charge”.
I’m thinking that if a CT is in use it needs to be somehow connected to both Multis or maybe a CT per Multi is necessary. I have (a second) one on order and for now have reconfigured to be operating without external current sensor but it is extremely slow to react to loads as it was before I originally fitted the CT.
@Ardje One of the problems is that in fact the two Multis do not behave the same. They see massively differing currents passing through them but only sometimes, so it is unlikely to be wiring plus the wiring has been checked and checked again. I have current-clamped the AC connections and the VE.bus info that I have been monitoring seems to agree that they are behaving differently.
