question

Not much to add as i am not affected as much as others.

Below all I noticed is now the shade power is closer to cloud power (graph below) and can see where the sun pops out. On beta4 i note that the start point was earlier and maybe a more stable voltage? but too early to tell. (after 12 i was @ 100% SOC so the arrays where only need for extra loads)

Looking good for me, good job guys!!

Update:

Since we have today relatively cloudy weather:

I have two strings oriented south.

The yield on both a relatively close together now, even due to the fact that atm one is more shaded than the other.

When looking closer on the remaining 5 Strings:
Those are more or less laying flat with 15° toward north-west.
Those are also close together now from peak power and yield perspektive.

Looks like the update does its thing. Yield compared to the last comparable day is a bit better.

L1 + L2 = South

R4 = not connected yet - needs to be build when the weather gets better.

Rest = 15° north-west.

PS: All Strings are the same and containing 6x Trina Vertex 425.

Generally quite well behaving, but got another surprising sudden behavior this afternoon between 14:50 and 15:12 on first unit MPPT1 tracker, when there was a cloud going over, leading to a bit erratic power output (at low power but still no reason for only 14W and 0W): Something still seems slightly fuzzy with global MPPT at low power at least:

Zoom here on that unit at those times which shows the obviously wrong gmppt decisions on mppt1 (no such sudden partial shadings at that time):

@Alex Pescaru Thanks for answering, and asking for type and diodes:

Edit: btw: On your graphs, the voltage seems always at maximum until 13:15, so it looks like you didn't have any partial shadows then ?

I double-check once more:

My solar panels are 120 cells Trina Vertex dual glass monocrystalline type TSM-NEG9.28 of 415Wp each Datasheet: https://static.trinasolar.com/sites/default/files/Datasheet_Vertex_S+_NEG9.28_EN_2022_PA3.pdf . They have bypass diodes, according to their manual https://static.trinasolar.com/sites/default/files/Vertex_User_Manual_July2022.PDF like any modern PV panel (are there any without bypass diodes nowadays?) and have corresponding visible junction boxes at their back for the diodes. They are new, came to the roof in original manufacturer palette packaging, have been handled and connected properly by trained professional installers, and strings have been measured for compliance to specs by an auditor. All frames are properly earthed to a professionally installed and audited/measured lightning protection system. And we didn't have lightning strokes nearby. So any damage to diodes, although never impossible, are highly unlikely, especially on 4 different strings.

I could try to verify those, by applying a very small current source and measuring the voltage, I don't see how to test all diodes of a string.

Now, what makes you think on my graph that diodes would be damaged ? Except for the odd behavior at low power, the rest of the day seem to work properly for a not great PV day ?

The odd thing is that today the weather was with a lot of clouds, and very little sun, and a few weeks per year we have lots of medium-far shadows, which is the reason to have tried to choose great MPPTs in conjunction with modern panels, not too large strings and a strings cabling layout on the roof taking into account shadings and the solar panels architecture. We excluded optimizers to avoid electronics up there and also due to documented incompatibility, potentially damaging Victro-MPPTs.

If Victron MPPTs could show us the current-voltage or power-voltage result of their global MPPT scans in VRM or at least Victron Connect apps, like some other inverter manufacturers, installers could diagnose much better diode failures, and many many other PV issues, like e.g. bad contacts..... Hint Hint @mvader (Victron Energy) ;-)

Hi @beat

My mistake about the diodes...

Now, looking with more attention to your graphs, I see that you've left the autoscaling option ON in contrast with me where I've manually set the graphs vertical axis limits and therefore I had the impression that the vertical axis starts from 0 on your case, but no, it's not the case.

I believed that you had an excursion of voltage greater than 3/4 of panel total voltage for the same current that it's not normal and characteristic for strings with problematic diodes.

My bad, sorry...

But maybe it's not a bad idea to set manually the vertical axis from 0 to max, because in that way you have a more ample and correct vision about all parameters and their variance and a big variance like you see now on your graphs could be, in fact, just a small one.

As for me, the partial shading starts after 14:00, but unfortunately, not so evident on yesterday graph.

Only from 15:40 to 16:00, where there was a need for increased power and you can see the partial shading and the diodes working.

Alex

Hi @Andrii Podanenko

Can you create a service report (https://ve3.nl/VictronConnect_How-to-create-a-VictronConnect-Service-Report) after connecting to the Multi RS and it has this weird value and post that service report here in a comment?

Kind regards,

Thiemo van Engelen

Shows 0 after some time ( 5+ minutes ) ( this was always such )