Single MPII 2x 120v with Victron shunt and Cerbo
Gateway V3.51
Bus v552
712 v4.16
Running on inverter with virtual swtich
I saw a negative power “spike” 39kw on L2. I’ve seen this before. I only caught it today, because tonight I was working on another problem. This was right about the same time as my Kuerig was finishing making coffee this morning, but not absolutely positive on the time. The other issue was trying to figure what triggered my system to go to grid… which I figured out.
EDIT: L1 load is likely the electric water heater.
I would suspect a measurement glitch or a bit flip on the data connection to the Cerbo, as a real 40kW spike should trip/explode components (fuses and such) in most systems quite instantly…
Do you run the communication cables to the cerbo in parallel and near power cables?
Comm cables run in parallel with lines from SCC to buss bar, and from buss bar to 5th wheel 12v power inlet. The interesting thing, I’ve seen this before, around the same time.
Interesting is, that it only affects L2, which could speak against something impacting communication. Unless… does the communication cable end in the inverter providing L2?
Some pictures and a schematic of the setup might help someone to get an idea where the problem originates from…
MPII 2x, so it’s only one inverter. My plan changed during assembly, so the drawing is no longer valid. The networking cables are running somewhat parallel to the SCC output cables. Which is an interesting point, the problem seems to creep in both in the morning when solar is waking up, and occasionally in the evening, when it’s going to sleep. I haven’t taken the time to look hard at the times yet.
Only one Inverter invalidates the idea of bit flips in the communication line, as it would be extremely unlikely to only affect the value for L2, while leaving L1 fully untouched…
Forget about bad communication because of power events or cable routing…
The communication packets are checked for validity. If a packet is invalid/modified is discarded.
That is either a bad sampling from the start or an error on the VRM software.
But I tend to believe that it’s the former, because otherwise there will be many users to complain about it.
Look at your load/consumption around the time it happens.
It doesn’t have to be a long, permanent, big load. Maybe a big (maybe inductive) load stopping, from the looks of the spikes.
Or it can be a big induction spike on the area of the sampling point, the current transformer and/or shunt.
No inductive 120v loads are running (except microwave, is that inductive?). Fridge and separate freezer are running on 12v. AC is not running. The only high energy items might be the water heater on electric, microwave, or Kuerig. I am running into an issue with water heater coming on once every two hours, and possibly tripping when one of the other high load appliances running and putting Inverter into overload, but I’d see that on the graphs too.
I’ll put together a spread sheet of the data. But a quick glance over the last 30 days, almost every single negative spike is -39354, with a couple at -39174. Sometimes there are are L1 load in the roughly 1500w, sometimes nothing. Sometimes there’s an L2 load of roughly 1400w, sometimes none. The almost consistent number got my attention. I converted it to Oct and Binary, and neither jumped out at me.
If this is during inverting then there is either a problem with the inverter, or an aplliance is messing with the sine wave a lot and the inverter cant maintain a steady output.
Both ways are tested with an oscilloscope.
Hasn’t done it in the last 8 days… until this afternoon. Going into the hour view, it coincided with the electric water heater on L1 shutting down. Water heater seems to come on for a few minutes every couple of hours… I’ll watch this closer, but it’s baffling. It was about the same time we returned, but I can’t think of anything that might have been done.
