Yes there's 10 SIDs. When the scripts starts for PVOut config, it asks me for SID. But do I haveto input the inverter serials as well?

I'm now live at the roof. System found all 10 inverters and I enabled MIS (script asked me). Now an error came up: Different numbers of SID serial...Any hint? thanks Ralf

Now changed the SID with serial in front and that worked. Now an error came up in Upload cfg. Syntax error line 34; but there's the list of all inverters with serial and SID - looks correct, like in the example - but 10 times, comma separated.

e.g. 2110415xxx:61927,21104119xxx:61936,... (xed the last 3). The serial numbers are 10 digits instead of 9 in the script example; SID is 5 digits instead of 4 in script example

Ah now; typed in all the data again and script runs successfully!

Log file says that it connects to the last BT device in the .cfg list: connecting to 00:88:xx (1/10). the cfg list looks like this:

You should only have one BTAddress in the config (only the last one is active like you noticed)
When MIS_enabled=1 and all inverters have same NetID > 1 SBFspot should be able to connect to all inverters.

OK, I tried it - but it seems not to work; always only the last in the list was read. Now I copied the cfg files 10 times and made them available in the cron job, starting every 1 minute a download of 1 inverter. periodically, after 10 min all the inverters are read and hopefully uploaded. I could not test this finally cause sunset stopped me out. Will check tomorrow. But thanks again for supporting!! bye Ralf

Sorry to hear that... a remote installation makes it extra difficult. It's also the first time I deal with such a big plant of 10 inverters.
You didn't give all of the details of the current config, but I think it's better (faster) if you create one scheduled script that calls 10 times SBFspot for each inverter instead of having 10 cron tasks for each inverter. One run of SBFspot takes less than 30 seconds, so a complete run for all 10 should be less than 5 minutes.

What is the NetID of your inverters? This is important when you have a multi-device plant

By default, the NetID is set to 1 for all SMA inverters and SMA communication products with BLUETOOTH. If your PV system consists of an inverter and a maximum of one further BLUETOOTH device (e.g. computer with BLUETOOTH interface or SMA communication product), you can leave the NetID set to 1.

You must change the NetID in the following cases:

If your PV system consists of one inverter and two other BLUETOOTH devices (e.g. computer with BLUETOOTH interface and SMA communication product) or of multiple inverters with BLUETOOTH, you must change the NetID of your PV system. This will enable communication with multiple BLUETOOTH devices.

If another PV system with BLUETOOTH is located within 500 m of your PV system, you must change the NetID of your PV system. This will help keep both PV systems separate.

If you do not wish to communicate via BLUETOOTH, deactivate the BLUETOOTH communication on your inverter. This will protect your PV system from unauthorized access.

All BLUETOOTH devices in a PV system must have the same NetID. You can set a new NetID in the inverter by means of the rotary switch C.

See also Setting the NetID

I have edited the cron job according to the installation guide. I added 9 lines to add all the inverter tasks:

*/5 6-22 * * * /usr/local/bin/sbfspot.3/daydata
55 05 * * * /usr/local/bin/sbfspot.3/monthdata

*/10 6-22 * * * /usr/local/bin/sbfspot.3/daydata2
50 05 * * * /usr/local/bin/sbfspot.3/monthdata2

*/11 6-22 * * * /usr/local/bin/sbfspot.3/daydata3
45 05 * * * /usr/local/bin/sbfspot.3/monthdata3

etc.
The daydataX files have corresponding SBFspotX files with the correct bluetooth address (MIS ist still 1; maybe I try 0 as well). I didn't adjust the upload.cfg (still only one).
It's little weired that only 3 inverters deliver data. But without log not easy to determine why. Is there any possibility to send the logs peridocally e.g. to an mail address?
Thanks Ralf

PS: Net ID is always 1; inverters are istalled at a higher wall - it's maybe a half day work with a ladder (in dangerous height) to detach all covers and change Net IDs. With the old Oxley app on the phone it all worked very well via bluetooth access to 10 inverters and sending out to PVOutput. So Net ID should not be a showstopper

Looks complicated to me...
Just edit daydata with one call for each inverter:

/usr/local/bin/sbfspot.3/SBFspot -v5 -ad1 -am0 -ae1 -cfg/usr/local/bin/sbfspot.3/SBFspot_1.cfg $1 $2 $3 $4 $5 &>>$log
/usr/local/bin/sbfspot.3/SBFspot -v5 -ad1 -am0 -ae1 -cfg/usr/local/bin/sbfspot.3/SBFspot_2.cfg $1 $2 $3 $4 $5 &>>$log
/usr/local/bin/sbfspot.3/SBFspot -v5 -ad1 -am0 -ae1 -cfg/usr/local/bin/sbfspot.3/SBFspot_3.cfg $1 $2 $3 $4 $5 &>>$log
...
/usr/local/bin/sbfspot.3/SBFspot -v5 -ad1 -am0 -ae1 -cfg/usr/local/bin/sbfspot.3/SBFspot_10.cfg $1 $2 $3 $4 $5 &>>$log

Each SBFspot_X.cfg has its own BTADDRESS=

MIS is still 1; maybe I try 0 as well

There won't be much difference but it should be set to 0 to make connection a bit faster.

Net ID is always 1
So Net ID should not be a showstopper

Fair enough, but it would be easier to set up. Changing the NetID should have been done by your installer.
So you never used SunnyExplorer?

Is there any possibility to send the logs periodically e.g. to an mail address?

Never tried it but it is possible

All the data comes from SBFspot.db
Serial number is key here, not MAC address

Look here for some things to check

OK, I check this. Next steps is that I gonna install a reverse ssh on a VPS here to work remotely.
But do you - as a developer - have some hints or assumptions what happened here? I mean, in the DB all the inverters serials are stored - that's what I checked before. The Upload service simply has to fetch this data (according to config information) and start the upload. Obviously that works with the first inverter MAC identified by the installation script. Why not with other inverter serials? Really do not get this..

Without logfiles it is difficult to say something useful. Most issues can be found by following the wiki, starting with second inverter.

inverter time is 1.1.2001. Could this be the issue?

Yes it is... (at least one of the issues)
Unfortunately SBFspot is not always able to set the time. The only workaround is to use the official SMA SunnyExplorer application (windows only)

But you said: "only 3 inverters send data to PVout" so there must be data in the daydata table

What is the output of select * from inverters;
Are all 10 inverters there with correct S/N?

I tried to set time today (daytime) but it didn't work. Seems that I have to go onsite again with SunnyExplorer. In DB there are only those inverters that are corresponding to BTt address in SBFconfig.cfg. Every time I change the BT there, a new entry is written in DB. And for every inverter that has not a correct system time no data is written in DB. Seems that SBFspot stops writing data to DB in case of time issues. That's weird, because the inverter data (power generation) is fully available in log. Question: Would it be possible to write the local's computer time (together with the power geneartion data) into the DB (and ignoring the incorrect inverter time) instead of stopping the data transfer to DB completly? I mean, the old oxley solar solution did exactly the same thing. It ignored any time data differences and wrote actual time to the log/DB. This would mean a patch to actual SBFspot release right?

I found out that when SBFspot connects the very first time after wake up in the morning - it is able to set the time at the inverter. After a first connect, the inverters refuse to adjust time. So I will try to set the other inverter tomorrow and keep you posted

This morning I checked the inverter time and it turned out that all the inverters I adjusted the time via -settime fell back into reset time 1.1.2001 again over night. To fix this there has to be a call of SBFspot -settime before the first run of the SBF reading scripts in the morning. cronjob is difficult because depending on sunrise the inverters are powered up at different times. Do you have any ideas to solve this?

interesting detail: after -settime the inverter keeps the "wakeup time" at 1.1.2001:
pi@raspberrypi:~ $ /usr/local/bin/sbfspot.3/SBFspot -v -nosql -nocsv -finq | grep -i time
Local Host Time: 13/01/2025 08:59:44
Plant Time : 13/01/2025 08:59:45 (+1 sec)
Last Time Set : 13/01/2025 08:52:18
Operation Time: 50536.41h
Feed-In Time : 49469.64h
Current Inverter Time: 13/01/2025 08:59:41
Inverter Wake-Up Time: 01/01/2001 01:00:13
Inverter Sleep Time : 13/01/2025 08:59:45
startTime: 13/01/2025 00:00:00
startTime: 01/01/2025 12:00:00

Is there an option/parameter in SBFspot to adjust wakeUp time as well?