My Multiplus II 5000 broadly does a good job, however 3x in 4 months it has cut out, and with different problems:
Once I got an Error 24 and everything seemed to cut out but did come back after a restart with no ill effects except for the error message being in the log.
The second time it strangely stopped inverting and everything cut out, but I think this was due to a cell voltage (couldnt say for sure, but it had been a long time since a 100% calibration so possibly one cell caused it to cutout early.
Tonight at 22:37 BST it randomly turned to off and then power assist, before automatically restarting again. I didnt notice any power cut on the AC-In connected circuits, only the AC-Out ones.
How do I go about diagnosing what actually caused this? Is there a victron problem, is there a comms problem, is it the battery again, was there a temporary grid problem…etc. For the sake of ease I’d focus on the one from tonight, but this feels like a place that it would be helpful to have some Victron support to diagnose with VRM? A simple summary such as ‘A happened on the grid, triggering B, that then caused C after D minutes due to E rule’ would be incredibly useful.
Some background:
Half my house circuits are on AC-In and half on AC-Out, and total house consumption was about 400W at the time (see the logging below from my Iotawatt system that stopped recording at the time as it lost power). The jaggedy power consumption in green comes from my washing maching I believe, as green is my kitchen circuit, and oscillations are far too frequent for it to be my fridge.
I am running Venus OS 3.73 and Multiplus v558 firmware. I am using ESS but not DESS. I will run an update of firmware now on both, just in case.
The Multiplus was in ‘charger only’ mode (digital setting not with the switch) at the time as grid electricity is cheap at this time. That was the easiest way I found to ensure it forces all circuits to draw from the grid during cheap electricity hours, whilst still absorbing any excess solar that might be generated at another time of day. I appreciate this could be the ‘problem’ here, but I thought a ‘fault condition’ would always trigger it to invert, whatever mode it is set to?
There is nothing in ‘Alarm logs’ or ‘Event logs’
I am a DIY user, but consider myself quite advanced (electronic engineer).
Can you share a bit more about your system? Is the IotaWatt interfaced to your GX device somehow eg providing current sensor data? Do you have a Victron or third party current sensor?
Have you had a look at Alarm and Event logs, and perhaps looked into the NS logs using your PC connected to the MP II ? It might give you a hint as to what is going on.
Sorry, I was realising that I needed to give more details this morning:
Iotawatt is totally independent, but I find it useful for understanding the various current draws in the house (and AC In/Out on the Victron)
Multiplus has the CT sensor installed on grid tails (didnt have space for an external meter in the consumer unit)
Battery is a Yixiang and JKBMS based device talking via CAN (but with a failover to RS485, which I could now remove as CAN seems much more reliable, and avoids the complexity of SerialBattery hiding battery behaviour underneath)
Nothing in the VRM logs (Alarm or Event) or Venus OS notifications, but I havent connected the Multiplus to a computer since installation (always managed by Venus OS device), are there more logs available that don’t get read by default? Or might Venus OS store them locally and I can access by SSH or something?
This feels like a perfect example where a simple AI tool could very quickly analyse what happened during a period of time and explain what had happened to users to make debugging much easier…
Thank you, I did see that, but by the time I found it I had already done the reset. As I understood it there was no point in making the resistance measurements after the unit has been reset? Or is there a way to put it in fault mode and then measure?
Then you have to expect power cuts. Charger only means just that, it wont invert. In the event of the grid being out of spec, the MP will disconnect from mains and drop the load (since its not allowed to invert).
If you leave it in “On” then in the event of the grid not being in spec, it will disconnect from grid and supply the loads from battery. This will usually go without interruption.
Due to the system being setup as an ESS, there has to be a gridcode set, which narrows the allowed voltage and frequency range. So the grid does not have to drop out completely for the MP to disconnect.
You may be able to change LOM mode, from A to B. But that depends on your gridcode and grid supplier, if you are allowed to change that. Depending on gridcode its also possible that only one type is avaiable/allowed. Type B is slower/softer.
Im not sure on the behaviour in the event of a BMS error, i would expect the system to switch to pass-through even if the MP is set to “Charge only”.
If you want to force-charge, then rather use a charge shedule, or set the ESS mode to “Keep batteries charged”.
Thank you @lxonline - Presumably measuring the Neutral In-Out resistance (as advised in the Error 24 troubleshooting) needs me to somehow trigger fault conditions? By default they will be connected but I need the relay to ‘try and disconnect’? Or were you referring to the fault last night?
Also thank you @chrigu - I can understand that being in Charger only mode when a grid fault happens may trigger slightly unexpected behaviour, even if there was no fault visible to me, however I would expect some kind of fault to be logged by the Multiplus?
The reason I dont leave it in ‘On’ mode is that it means the only control is charging current? When I limit charging current to 0 (to effectively leave the battery idle) it then feeds excess solar back to the grid, whereas I basically want a mode where is charges using excess solar only, and not charge from the grid.
Thats what ESS with “Optimized with/without BatteryLife” does. You set a gridsetpoint, 0W for example, and the MP will keep the grid input at zero Watts, by using excess PV energy to charge the battery
I played with that but couldnt get it to do what I wanted all the time, though it does stay in that mode most of the time. The situation I failed to get working was when I didnt have solar, I want it to draw household load power from the grid rather than discharging battery. Having a grid setpoint of -500W would result in it charging at 500W even if there isnt 500W of solar, and a grid setpoint of 0W would mean it doesnt draw from the grid when the electricity is cheap. Setting it to charge was the only way I found to prevent discharge.
Im not sure if my explanation is good enough, but my objective effectively simplifies into two cases:
Cheap power:
Draw from grid for all loads, except if there is excess solar (in which case store it and grid draw should be zero). I.e. Only ever charge
Expensive power:
Draw from battery for all loads and keep grid at zero, absorbing solar if excess
Pretty sure that all you need to adjust is the inverter power limit (System, System Setup, ESS, Limit inverter power). If you set it to zero, it will only charge but not feed AC loads. And due to the ESS mode “Optimized” it will only charge on excess PV.
During high prices you set the limit up, so it will do both charging on excess PV and feed the loads, keeping the grid at the gridsetpoint.
During gridfailure this inverter limit is not active since its an ESS featur
Interesting, thank you, will test how things go when playing with max inverter power instead of using charge only. I can see the benefits of using an ESS setting rather than an inverter level setting! Thank you!
@chrigu Just to close the gap on one potential cause of problems. Your suggestion to use Max inverter power rather than changing mode to charge only does indeed work great. Hopefully that will partially solve the issue that happened last night with it changing to ‘off’ mode in grid failure situation… Though I guess I now need to wait again and see if I still have other problems (and/or a dodgy back feed relay…). Thank you!!
