Summer Solstice at 35ø latitude.
By 1:00 PM the bank had to be at 80% full to allow the array to achieve 100% fill* before array output decreased to 10% or less.
*Gravity readings.
So configure the size of your array accordingly.
Sure, the above is ambiguous. It has to be.
Minimizing conjecture and WA guessing.
It take math to translate this into usable information. But without kWh load estimates, it's all wild speculation and conjecture.
This is why insist customers acquire a means to charge their bank to gain instant vMax 14.8 vdc absorbsion. Now -that- is not conjecture.
Want a tip?
PLUG IN!
And measure kWh used by the converter to power calculated use of 12 volt accessories. Add 20% (120%) and the sum will be close enough for thermonuclear grade accuracy.
An AMP HOUR METER would be even better.
But utter conjecture, guesswork, and theory is utterly a waste of time.