light level vs inverter output

[davidwf15]

my system comprises 2 strings of 7 x 475W Aiko panels and a KH7 inverter, all new in April this year
I have noticed on many occasions that even with a "reasonable amount" of daylight the inverter is producing no solar
This morning for instance at 8am the it was bright cloudy with a light level of 13,000 lux, the inverter said the panels were at 205V but ther current (and power) was zero
A short while later the light level had increased to 17,000 lux and the panels were producing around 500W on each string
The panel cottage always seems to be around 220V and never really increase although their current and power does (presumably under mppt control of the inverter)
Does that sound right ?????
Surely the inverter should produce SOME output with a string voltage down to around 80V

[MaterialBarracuda48]

Last year, on earlier firwmare versions, it was showing me generating some Wattage very early and late into the evening.
Now it seems to start late, end early.

I am sure that the power is being self-used by the actual inverter, but don't quote me.
The other thing, is the daily PV totals remain unchanged despite the generation changes. I suspect the annual figures for PV production add up the same (less a little PV degradation)

[daveg]

Asking Google AI about updates for my H1G2 it also threw out this regarding KH
While your H1-G2 system got a relatively quiet rollback, the KH series rollout has been a much bigger headache for the community.
The situation with the KH series mirrors yours, but with higher stakes:
The Force Discharge Bug Hit Them Harder: On the KH series, the recent firmware updates (specifically around the Master 1.55 / 1.58 versions) severely broke the "Force Discharge" logic. Instead of reverting to "Self-Use" battery power when a timed discharge slot ended or hit its limit, the inverters got stuck pulling full power straight from the grid—often right at the peak, most expensive time of day for the user.
Low-Voltage Solar Drops: Users also reported a bug where the inverter stopped tracking and logging low-voltage PV production (anything under roughly 200W–250W) during heavy cloud cover or early morning. It essentially "blinded" the inverter to trickle solar generation
Might be a hallucination, but it seems to chime in with MBs report.

[MaterialBarracuda48]

Low-Voltage Solar Drops: Users also reported a bug where the inverter stopped tracking and logging low-voltage PV production (anything under roughly 200W–250W) during heavy cloud cover or early morning. It essentially "blinded" the inverter to trickle solar generation
Might be a hallucination, but it seems to chime in with MBs report.

Sometimes AI results are a bit screwy, but if I run my inverter in Feed-in mode, then the battery will trickle charge, but won't show up as if power if going to the battery.

But as I mentioned earlier, I do think comparing last year to this year that my kWh daily figures are around the same (best day for example)

EDIT: here is a link to a previous thread, same mystery increase (2nd post from me has picture of SoC gain)
viewtopic.php?p=13812#p13812

[daveg]

Could that SoC increase be the Feed In mode bug that Tony M refers to here?
https://github.com/TonyM1958/HA-FoxESS- ... -pages-box

[MaterialBarracuda48]

Yeah noticed the bug myself, and see Tony's site at a later date.

[daveg]

That's one I think they've managed to fix then (at least for H1G2). I use feed in mode routinely in summer and never see it.

[davidwf15]

are we perhaps getting away from my original question......

[daveg]

Sorry!
I may be mistaken, but as I read it MB has seen the same issue as you, and if Google AI is to be believed so have other users.

[MaterialBarracuda48]

To be fair to your both, I was suggesting that reported events and reality are different.
The system won't show you actually charging the battery in 'feed-in' mode (where the SoC line goes green) but you will see the SoC increase.

To expand upon my earlier post, last year I could see a few Watts being made after the Sun went down in the Summer, same period this year, on the new firmware versions and now it 'stops' maybe 90 mins before sunset.

22nd June 2026 - Reported PV output stopped at 19:00
22nd June 2025 - Reported PV output stopped at 21:34

[davidwf15]

am seeing an mere 31w now

[davidwf15]

.....but this morning nothing at all then suddenly 1kw .....that can't be right.......but why are the panels still showing 200V now at sunset with no generation

[davidwf15]

...and literally a few minutes later nothing despite the light level being 1700 lux which is still quite bright

[davidwf15]

Earlier this evening.....

Now the battery is on high discharge to grid, the panels are showing a higher (presumably open cct) voltage with zero power

Octopus are presently managing my system as a Virtual Power Plant

[davidwf15]

Hmmm....never seen this before....

...28W...

[davidwf15]

This is what i was mainly querying.....the panels are still showing well over 100V but not providing any power as it is now dusk .......but the inverter spec says it starts generating at 80V input .....