Daily reset/shutdown of my Powerwall+ and/or Inverter after coolant top-off?

[kairojya]

So I ended up having what I thought were 2 days of no solar production drop out. We’ve had a lot of bad weather, so it could be that it just wasn’t clearly evident in the daily solar graph what was happening. Now that we’ve had several days of normal sunshine, I can see that my dropout reset to the start of my typical solar production window.

I think yesterday morning at 1:50 AM EST my system restarted due to firmware upgrade from 21.24.1 to 21.24.2

 

[Bitslizer]

So I ended up having what I thought were 2 days of no solar production drop out. We’ve had a lot of bad weather, so it could be that it just wasn’t clearly evident in the daily solar graph what was happening. Now that we’ve had several days of normal sunshine, I can see that my dropout reset to the start of my typical solar production window.

I think yesterday morning at 1:50 AM EST my system restarted due to firmware upgrade from 21.24.1 to 21.24.2

Did you try the stop enable again in the evening?

 

[darryllee]

Ah, I forgot to post here that I also got the firmware upgrade on Monday night around 10:45PM PDT, as reported by api/system/update/status:

{
   "state":"/download",
   "info":{
      "expected_bytes":[ "220782656" ],
      "got_bytes":[ "88313088" ],
      "offset_bytes":[ "0" ],
      "pass":[ "0.0" ],
      "status":[ "continuing" ]
   },
   "start_time":1629178911372,
   "current_time":1629179104979,
   "last_status_time":1629179092782,
   "version":"21.24.1",
   "offline_updating":false,
   "offline_update_error":"",
   "estimated_bytes_per_second":457580,
   "last_download_status_time":1629179092782
}

At which point, my authentication token expired, and almost all of my logs just have errors until I noticed and reauthenticated on Tuesday night:

{
   "state":"/update_succeeded",
   "info":{
      "status":[
         "nonactionable"
      ]
   },

   "current_time":1629252816823,
   "last_status_time":1629246389384,
   "version":"21.24.2",
   "offline_updating":false,
   "offline_update_error":"",
   "estimated_bytes_per_second":null
}

Since api/status doesn't require auth, it kept chugging away. But it's funny that it reports the wrong time zone. (I save my own timestamp to my logs, and it definitely went down after my last check at 2021-08-16T22:45:01-07:00):

{
   "din": "XXXXXXXXXXXXXXXX",
   "start_time": "2021-08-16 23:26:28 +0800",
   "up_time_seconds": "3m33.532291411s",
   "is_new": false,
   "version": "21.24.2",
   "git_hash": "e8242aeb212fc7730c3f608acc07867ec63f27ab",
   "commission_count": 2,
   "device_type": "teg",
   "sync_type": "v2.1",
   "leader": "",
   "followers": null
}

The good news is that so far (knock on wood!), I've not seen any drop-outs, I guess owing to the fact that it restarted in the evening.

 

[kairojya]

Did you try the stop enable again in the evening?

I used the e-stop switch yesterday evening and did not have a production dropout today. I have no idea what reset my timer after my last attempt.

 

[abek]

Ah, I forgot to post here that I also got the firmware upgrade on Monday night around 10:45PM PDT, as reported by api/system/update/status:

{
   "state":"/download",
   "info":{
      "expected_bytes":[ "220782656" ],
      "got_bytes":[ "88313088" ],
      "offset_bytes":[ "0" ],
      "pass":[ "0.0" ],
      "status":[ "continuing" ]
   },
   "start_time":1629178911372,
   "current_time":1629179104979,
   "last_status_time":1629179092782,
   "version":"21.24.1",
   "offline_updating":false,
   "offline_update_error":"",
   "estimated_bytes_per_second":457580,
   "last_download_status_time":1629179092782
}

At which point, my authentication token expired, and almost all of my logs just have errors until I noticed and reauthenticated on Tuesday night:

{
   "state":"/update_succeeded",
   "info":{
      "status":[
         "nonactionable"
      ]
   },

   "current_time":1629252816823,
   "last_status_time":1629246389384,
   "version":"21.24.2",
   "offline_updating":false,
   "offline_update_error":"",
   "estimated_bytes_per_second":null
}

Since api/status doesn't require auth, it kept chugging away. But it's funny that it reports the wrong time zone. (I save my own timestamp to my logs, and it definitely went down after my last check at 2021-08-16T22:45:01-07:00):

{
   "din": "XXXXXXXXXXXXXXXX",
   "start_time": "2021-08-16 23:26:28 +0800",
   "up_time_seconds": "3m33.532291411s",
   "is_new": false,
   "version": "21.24.2",
   "git_hash": "e8242aeb212fc7730c3f608acc07867ec63f27ab",
   "commission_count": 2,
   "device_type": "teg",
   "sync_type": "v2.1",
   "leader": "",
   "followers": null
}

The good news is that so far (knock on wood!), I've not seen any drop-outs, I guess owing to the fact that it restarted in the evening.

How do you check the api status and get this info?

On a separate note, after speaking with the Tesla electrician that installed my system, I just noticed that I'm now running version 21.31.0. I look forward to seeing if the daily restart issue has been resolved tomorrow around this time. I'll report in to let everyone know.

 

[kairojya]

My system upgraded to 21.31.0 yesterday morning around 2:00 AM EST. I had not had a daily solar production dropout in a few days up until that point (after using my e-stop in the evening beforehand). Today I had a dropout around 8:30 AM EST which is a little later in the morning than is has happened in the past after this cycle has restarted. Notably today my solar production itself started later than usual due to cloud coverage.

I can also report that the solar production dropout is visible in the new iOS app version (4.0.0).

 

[abek]

My system upgraded to 21.31.0 yesterday morning around 2:00 AM EST. I had not had a daily solar production dropout in a few days up until that point (after using my e-stop in the evening beforehand). Today I had a dropout around 8:30 AM EST which is a little later in the morning than is has happened in the past after this cycle has restarted. Notably today my solar production itself started later than usual due to cloud coverage.

I can also report that the solar production dropout is visible in the new iOS app version (4.0.0).

Interesting… in my case I should have had a daily restart around 6:15 pm tonight but after the 21.31.0 update it didn't happen. I had a nice, uninterrupted day of solar.

 

[Bitslizer]

How do you check the api status and get this info?

On a separate note, after speaking with the Tesla electrician that installed my system, I just noticed that I'm now running version 21.31.0. I look forward to seeing if the daily restart issue has been resolved tomorrow around this time. I'll report in to let everyone know.

Interesting… in my case I should have had a daily restart around 6:15 pm tonight but after the 21.31.0 update it didn't happen. I had a nice, uninterrupted day of solar.

I think/guess the firmware update "reset" the timer to when the update finished. Was it updated overnight?

 

[kairojya]

No dropout for me this morning in either my own local API data or visible via the latest iOS app. I did not manually restart/reset the system. I’m still showing the same firmware that I’ve been on since Wednesday.

 

[abek]

I think/guess the firmware update "reset" the timer to when the update finished. Was it updated overnight?

Turns out the daily reset happened again this morning, and if you’ll notice yesterday’s post with the screenshot, it happened then too, early in the morning. I didn’t notice it because it happened early and I was expecting it in the evening. Nevertheless, it seems the daily reset was not resolved with the new software update.

 

[Bitslizer]

The mystery deepens..... Seen the latest firmware 21.31.0 may have forced a fixed 24 hours timer about 7:45 in the morning. Doing the disable/enable in the evening once didn't move the reset. Not exactly a time I normally can monitor the status like a hawk.... I just noticed sometime my string 4 was working when I wake up and not working after 8am or so.

 

[kairojya]

The mystery deepens..... Seen the latest firmware 21.31.0 may have forced a fixed 24 hours timer about 7:45 in the morning. Doing the disable/enable in the evening once didn't move the reset. Not exactly a time I normally can monitor the status like a hawk.... I just noticed sometime my string 4 was working when I wake up and not working after 8am or so.

I’m not seeing that with my system. I haven’t done a manual reset since the new firmware and I’ve had 2 days with no solar drop out so far.

 

[yves31]

Inverter issue:
I have tesla solar panel installed without powerwall. The system was activated two days ago and I connected it to the internet ~11:00 am yesterday. My panel produced ~5.2kw at peak time throughout yesterday as shown in picture 1. This morning the panel generated ~4.2kw only under full sunny daylight. I suspected something wrong with the system, then power-cycled the inverter. It took ~10 minutes for inverter to reboot. After booting up, the system started producing the same power as yesterday, ~5.3kw, throughout remaining hours of the day (pic 2). What could possibly happen to make the difference with inverter power recycle? Firmware update?

 

[kairojya]

My system is doing the daily 10 min solar production outage around 7:30 AM. I’m on version 21.31.1 and the night reset using the e-stop switch doesn’t seem to work anymore to move the daily reset out of solar production hours. I’ve tried on 2 different evenings so far.

 

[yves31]

It happened again today: rebooting inverter increases production.
I was not at home until 4pm when I found power production was only ~4.1kw. I immediately power cycled inverter and production went up. So I must do daily inverter reboot to keep high power production. This must be a firmware bug in inverter. Will tesla fix it?

 

[Bitslizer]

It happened again today: rebooting inverter increases production.
I was not at home until 4pm when I found power production was only ~4.1kw. I immediately power cycled inverter and production went up. So I must do daily inverter reboot to keep high power production. This must be a firmware bug in inverter. Will tesla fix it?

How many strings do you have. Do your inverter has a teslaPV or TEG wifi access SSID?

Maybe a string is disconnected and rebooting reconnected it

 

[yves31]

How many strings do you have. Do your inverter has a teslaPV or TEG wifi access SSID?

Maybe a string is disconnected and rebooting reconnected it

I can access inverter with wifi. There are two strings. One shows ~2x amperages than the other. I don't think one string was disconnected due to importance of both strings. I cannot compare amperage before and after reboot since inverter webpage does not allow login before rebooting. So, it's impossible to determine what have changed before and after power recycle.
I'll reboot inverter tomorrow to get information about firmware information.

 

[Bitslizer]

I can access inverter with wifi. There are two strings. One shows ~2x amperages than the other. I don't think one string was disconnected due to importance of both strings. I cannot compare amperage before and after reboot since inverter webpage does not allow login before rebooting. So, it's impossible to determine what have changed before and after power recycle.
I'll reboot inverter tomorrow to get information about firmware information.

Can you upload your panels layout?

The amp can be misleading unless you are constantly monitoring the amp value. As it's highly dependent on how direct us the sun shining on that particular string (ie a west facing string will read low amp in the morning when the sun is still in the east) but the voltage will be fairly high all day while the sun is out

 

[yves31]

Can you upload your panels layout?

The amp can be misleading unless you are constantly monitoring the amp value. As it's highly dependent on how direct us the sun shining on that particular string (ie a west facing string will read low amp in the morning when the sun is still in the east) but the voltage will be fairly high all day while the sun is out

I have 24 panels placed to two levels. 6 are on top level and others are on lower level. All panels have the same orientation towards southwest. I am not sure how the panels are connected, but the two strings do not show even output. At the moment when I logged into inverter web page two days ago, one string was showing ~310V x ~9A, the other ~220V x 4.5A.

The software version of my inverter is 20.51.2. It's different from other reports, possibly due to different equipment. I have panels without powerwall.

I rebooted the system twice today. The first time was at 8am when production at 0.4kw. It did not make any difference for output. I was expecting this reboot could affect production that should go up to 5kw with full sunshine at ~noon. However, output stuck at ~4.2kw at noon. I rebooted again the system at noon, and production jumped to 5.2kw immediately. It seems that reboot has to be at high power output in order to increase production.

 

[Bitslizer]

Can you upload your panels layout?

I have 24 panels placed to two levels. 6 are on top level and others are on lower level. All panels have the same orientation towards southwest. I am not sure how the panels are connected, but the two strings do not show even output. At the moment when I logged into inverter web page two days ago, one string was showing ~310V x ~9A, the other ~220V x 4.5A.

The software version of my inverter is 20.51.2. It's different from other reports, possibly due to different equipment. I have panels without powerwall.

I rebooted the system twice today. The first time was at 8am when production at 0.4kw. It did not make any difference for output. I was expecting this reboot could affect production that should go up to 5kw with full sunshine at ~noon. However, output stuck at ~4.2kw at noon. I rebooted again the system at noon, and production jumped to 5.2kw immediately. It seems that reboot has to be at high power output in order to increase production.

When you rebooted first time in the morning, was it showing the same number of string compared to when you rebooted in the afternoon? These are Tesla inverters right? Do you have the 340w panels or newer 425w panels (I can calculate the string base on the panel specs)

If you can take picture of the inverter status after reboot in the morning (where it didn't increase) vs reboot in the afternoon (where you get the productions increase). That will be helpful.

Facing the same direction, generally without shading issue the amp reading should be roughly the same for the different strings, where the voltage is determined by the number of panels in the string.

The fact that one string is double the amp of the other makes me suspect one string should have been parallel (3 strings total, 2x9 and 1x6 with the 2x9 parallel and split back into 2 string at the inverter).

If you have access to open the inverter taking a picture of the wiring (black out the password) will also help