> Just plug several units into a power strip ¯\_(ツ)_/¯.
Hehe, ok. I think that's your insurance company calling on the other line.
> are you exporting power
Exporting 12 MWh / year or thereabouts.
> or zero export with a current monitor upstream of the main panel?
There is a current monitor (a Shelly tri-phase one), right now it is still economically viable to do so (though the utility companies are trying what they can to dissuade you by changing the deal through politics). If it is no longer then I will just install a battery and disconnect from the grid for the summer months.
And I'm not using a transfer switch because I don't have a battery to stabilize the system.
Well, technically those breakers are a manual transfer switch, only it is broken up into two halves and I can just disconnect the mains feed and run in island mode but I would still need to install a battery and a charger. House mains breaker off, solar on would be the house running entirely off the local stuff, I just don't trust that inverter without a battery behind it to be able to react quickly enough to load changes and by default it is set up to disconnect if the grid goes off, so you'd have to manually override that. The main issue with it being two halves is that you can not guarantee that the house net is in-sync with the grid at the moment you make the switch and that's a bad idea with a system this powerful, so I'd definitely get a proper automated one if I intended to do this for real, otherwise you might cause a load spike which could trip breakers and annoy the neighbors.
Right now I can't switch that on or off under load anyway because the large inverter would simply disconnect as well.
Tri phase breaker of the right amperage was about 150 bucks.
If I were to do this I would probably get a complete set from Victron, their stuff is amazingly well engineered, but if these open source people are going to make an inverter/charger combo then I might go for that and add a another manufacturers automated transfer switch.
An inverter is, complexity wise, not that much harder than a large switching power supply, there is some more instrumentation and some more rules but it isn't super difficult. It is much harder to make one that is commercially viable because those guys all cut corners to stay competitive. Ironically a proper case is probably the hardest part, there are also some larger inductors that might be tricky to source. And if you were to design one you should probably make the low voltage stuff (UI, CPU) on a completely separate board from the line voltage stuff and go for tri-phase right away because it is so much cleaner. Bonus points for modularity of the output stage.
Perhaps I'm in the wrong thread but there is no part of me that wants to DIY my own power electronics design. I've designed a few bucks and boosts professionally and chasing down those failure modes was a headache. Victron = solid hardware, openly-documented protocols, no cloud? Fahgettaboudit.
(then again maybe someday I'll hit some wall with off the shelf MPPTs and find myself wanting to go down that rabbit hole lolol. but honestly AC coupling seems cleaner in terms of things like fault protection on longer runs - fault on a stiff mains circuit -> breaker will trip. Fault on a circuit where the current/power is intrinsically limited to what the solar panels can supply -> ???)
I asked about the transfer switch / monitoring because I've looked at the same problem here, first with a generator now with solar. Incoming power service is on the complete opposite side of the house from where I really want the power handling gear. The manual two breaker thing is practical (for a generator at least), but not code compliant here (no positive lockout).
I would think Victron would have an option for a remote transfer (/disconnect) switch, but I haven't really looked into it yet. It would still have to get the grid phase timing somehow to line them up before connecting, so something more than merely a dumb contactor.