Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

1.25.0

This site may earn commission on affiliate links.
Yesterday, on 1.26.0, was the first completely normal operation on TBC Balanced that I've seen since October 8, approximately when I got 1.25.0. By normal, I mean that there was no interruption to Powerwall charging during Part-Peak and discharge did not start before the Peak period started. The Powerwalls also charged more than they discharged, so I had a net SOC gain yesterday.
Well, it was only two good days. Back to the strange pauses in charging and early discharge that was happening for most of October.

2018-10-26_14-54-49_000.jpg
2018-10-26_14-44-19_000.jpg
 
Last edited:
  • Informative
Reactions: Bigtanuki
Well, it was only two good days. Back to the strange pauses in charging and early discharge that was happening for most of October.
Same here. I bumped up our reserve to 65% and got the Powerwalls to charge up to about 80%, which seems perfect. For now, I'll probably just play games with the reserve percentage to basically get the behavior I want. This entails bumping up the reserve around mid-morning when our solar really kicks in, thus forcing more charging, then lowering the reserve later in the day to ensure all of our peak usage will be supplied by the Powerwalls.

As I've said before, I wish the app gave us access to a "power user" configuration screen to allow us to specify the desired behavior in more detail. I'd prefer this over a system that attempts to "learn" our usage patterns, which in our case are not very regular.
 
I've had a web-server lockup on 1.25.0, so this bug is still not fixed. I wonder if they've given up trying.
Mine is on 1.25.0, and locked up back on the 11th. I have an open case with Tesla support to diagnose it, and have resisted the urge to reset the gateway myself, hoping they can get to the bottom of it.

I can PING the IP address connected to my Wifi, and the 192.168.91.1 IP address when connected to the TEG-XXX WifI, and all that low-layer networking works fine - but the builtin webserver API just timesout.

Working through Tesla support is like trying to swim through honey. Any question or request for update results in a 'Lets give level 2 48 hours to monitor whats going on', and with timezones not lining up and no level 2 on weekends, the best I've managed is a question from Level 2 left on an answering machine about once per month. That, and every second request being to make sure I'm using Chrome browser, and have cleared the cache.

The last useful bit of info, was that 'Level 2' were able to SSH into the gateway remotely, and can access the webserver pages and API remotely, so they can't see the problem over there in Tesla HQ. To Tesla support the gateway web API seems to be working OK.

Really really wishing I had a gateway with a magic reset button. At this rate it will be 2019 before I can get the gateway rebooted remotely.
 
Not sure if we have the same issue, but you can take a look at my recent thread inquiring about how to reboot the gateway.

In my case, rebooting the gateway did not help but login to the gateway and press that big "Start Powerwall" button fixed my problem. I even got to run diagnostic a couple of times just for fun. :)

Mine is on 1.25.0, and locked up back on the 11th. I have an open case with Tesla support to diagnose it, and have resisted the urge to reset the gateway myself, hoping they can get to the bottom of it.

I can PING the IP address connected to my Wifi, and the 192.168.91.1 IP address when connected to the TEG-XXX WifI, and all that low-layer networking works fine - but the builtin webserver API just timesout.

Working through Tesla support is like trying to swim through honey. Any question or request for update results in a 'Lets give level 2 48 hours to monitor whats going on', and with timezones not lining up and no level 2 on weekends, the best I've managed is a question from Level 2 left on an answering machine about once per month. That, and every second request being to make sure I'm using Chrome browser, and have cleared the cache.

The last useful bit of info, was that 'Level 2' were able to SSH into the gateway remotely, and can access the webserver pages and API remotely, so they can't see the problem over there in Tesla HQ. To Tesla support the gateway web API seems to be working OK.

Really really wishing I had a gateway with a magic reset button. At this rate it will be 2019 before I can get the gateway rebooted remotely.
 
Not sure if we have the same issue, but you can take a look at my recent thread inquiring about how to reboot the gateway.

In my case, rebooting the gateway did not help but login to the gateway and press that big "Start Powerwall" button fixed my problem. I even got to run diagnostic a couple of times just for fun. :)

If you could see the initial power-flow graphics and press the big 'Start Powerwall' button, we don't have the same problem!
My problem is that all web-browser attempts to connect result in a big blank timeout. No button, no graphic, no diagnostics, no response.

This was confirmed by the Tesla tech who made a housecall and was sent around to 'help me connect to the TEG gateway' - and he couldn't connect to the web interface either. Now at least I know which plugs to unplug inside the gateway box to reset the internal server without dropping power to the house!
 
So many different ways to fail... And I got it to prepare for Zombie apocalypse :D

So how to reset the gateway without powering down the house? It may come it handy. Did anyone have experience with losing all internet connection and if the backup gateway will still function. (see above but seriously.)

If you could see the initial power-flow graphics and press the big 'Start Powerwall' button, we don't have the same problem!
My problem is that all web-browser attempts to connect result in a big blank timeout. No button, no graphic, no diagnostics, no response.

This was confirmed by the Tesla tech who made a housecall and was sent around to 'help me connect to the TEG gateway' - and he couldn't connect to the web interface either. Now at least I know which plugs to unplug inside the gateway box to reset the internal server without dropping power to the house!
 
I finally talked to someone in Powerwall support about my system's strange charge and discharge behavior. They were unwilling to acknowledge that anything was wrong. They said that the system was responding to and relearning after I adjusted the Reserve "8 times" in the period since it started "misbehaving". The way I remember it, I raised the Reserve BECAUSE it was behaving strangely and the average SOC over the week was dropping because it wasn't charging as much as it should. It has been behaving more normally this week, so I will continue to watch it. However, if it starts to act up again, I will ask if there is a way to turn off the "Learning" feature and have it follow basic rules for charge and discharge like it used to before 1.25. Basically, I want it to continue charging until full in Part-Peak and only discharge during Peak or during evening Part-Peak (9pm-11pm) if there is sufficient SOC.

They did acknowledge that my solar should not have shut down during my last outage and have escalated the issue because the Powerwalls raised the operating frequency during an outage when they shouldn't have.
 
I finally talked to someone in Powerwall support about my system's strange charge and discharge behavior. They were unwilling to acknowledge that anything was wrong. They said that the system was responding to and relearning after I adjusted the Reserve "8 times" in the period since it started "misbehaving". The way I remember it, I raised the Reserve BECAUSE it was behaving strangely and the average SOC over the week was dropping because it wasn't charging as much as it should. It has been behaving more normally this week, so I will continue to watch it. However, if it starts to act up again, I will ask if there is a way to turn off the "Learning" feature and have it follow basic rules for charge and discharge like it used to before 1.25. Basically, I want it to continue charging until full in Part-Peak and only discharge during Peak or during evening Part-Peak (9pm-11pm) if there is sufficient SOC.
I really wish there was better documentation or indication of why they function as they do so we could have a better idea of what's happening. As it is, we see things that aren't what we expect/want so we change the reserve, change the price schedules, change it back and forth between Self-powered and TBC...and whenever we do that, it confuses the "learning" process. :confused:

If they had a few more options, I wouldn't have to keep changing settings in order to manually charge fully before a storm or charge my car during the day. Argh...
 
  • Like
Reactions: Bigtanuki and eml2
I really wish there was better documentation or indication of why they function as they do so we could have a better idea of what's happening. As it is, we see things that aren't what we expect/want so we change the reserve, change the price schedules, change it back and forth between Self-powered and TBC...and whenever we do that, it confuses the "learning" process. :confused:

If they had a few more options, I wouldn't have to keep changing settings in order to manually charge fully before a storm or charge my car during the day. Argh...
They should have an option to disable the learning process, esp. if it doesn't include weather forecast as one of the conditions.
 
They should have an option to disable the learning process, esp. if it doesn't include weather forecast as one of the conditions.
My impression (software 1.26), over the past few weeks, is that the battery assumes that tomorrow will be like today in orer to determine the extent of off-peak charging. And it looks likely that if I switch it back to Balanced when I see a sunny day coming then it restart the learning process when I switch to Cost Saving again. Not clever! Give more control to the users and let the battery learn by observation.
 
Basically, I want it to continue charging until full in Part-Peak and only discharge during Peak or during evening Part-Peak (9pm-11pm) if there is sufficient SOC.
.

I've configured peak & off-peak (no part-peak) to get that behavior. I've extended peak from SCE's 2-8pm an hour to help with the duck curve. My two powerwalls discharge about 20‰ during that time, sufficient for SGIP charge/discharge cycles.
 
  • Like
Reactions: abasile
I've configured peak & off-peak (no part-peak) to get that behavior. I've extended peak from SCE's 2-8pm an hour to help with the duck curve. My two powerwalls discharge about 20‰ during that time, sufficient for SGIP charge/discharge cycles.
During the Winter that might make sense for me because I doubt my Powerwalls will get to 100% until I go on vacation (drastically reducing peak consumption) or the solar production picks up again in March. However, I am philosophically opposed to this approach. Tesla should work out the firmware so that it works correctly without us forcing the issue with artificially skewed settings.
 
We had a very, very brief power outage a couple of nights ago, and for whatever reason our Powerwalls didn't kick in seamlessly. A few of our circuit breakers tripped, on the backed-up loads subpanel. This was actually our first power outage since having the Powerwalls installed this past March. For others on 1.25.0+ who've experienced outages, have your Powerwalls behaved as expected?
 
We had a very, very brief power outage a couple of nights ago, and for whatever reason our Powerwalls didn't kick in seamlessly. A few of our circuit breakers tripped, on the backed-up loads subpanel. This was actually our first power outage since having the Powerwalls installed this past March. For others on 1.25.0+ who've experienced outages, have your Powerwalls behaved as expected?
Yes, mine have powered loads as expected during recent outages. Here is my history:
October 31: 22 minutes
October 16: 1 hour, 23 minutes
September 27: 1 hour, 0 minutes

I was home during the October 16 outage and the lights flickered briefly and I think I heard a UPS click over to battery and back, but it did not beep. However, after about 20 minutes my solar stopped generating because the Powerwalls pushed the frequency high and out of range for my micro-inverters. As far as I can tell, my UPS units did not transfer to battery during this frequency change. Tesla is still looking into this issue.
 
  • Informative
Reactions: PBBear and abasile
I'm not actually sure that it's always expected that the transition be seamless. For the one real power outage I've experienced, some of my devices (e.g. microwave oven clock) reset. Looking at pvoutput, though, it appears like the power was flaky for some time before it went completely out. I think the Gateway didn't decide that the power was out despite some fluctuations that were large enough to cause issues inside the house. My solar inverter had been kicking out several times a day for a few days leading up to the outage too.
 
  • Informative
Reactions: abasile
My power is out again. Gotta love PG&E. The solar shut down again while running on the Powerwalls. I called Tesla Energy Support and they tried to tell me that the solar should continue to run at less than 1 Hz deviation and I should contact Enphase. I mentioned that the power was out at the moment and the solar shut down again. I asked them to remotely check the frequency and they did. The rep acknowledged that the system should not be running at 63 Hz when the batteries are at 55% SOC and my case would be escalated again. I hope they get to the bottom of this because charging from solar while the grid is down is pretty important.
 
  • Informative
  • Like
Reactions: abasile and PBBear
My power is out again. Gotta love PG&E. The solar shut down again while running on the Powerwalls. I called Tesla Energy Support and they tried to tell me that the solar should continue to run at less than 1 Hz deviation and I should contact Enphase. I mentioned that the power was out at the moment and the solar shut down again. I asked them to remotely check the frequency and they did. The rep acknowledged that the system should not be running at 63 Hz when the batteries are at 55% SOC and my case would be escalated again. I hope they get to the bottom of this because charging from solar while the grid is down is pretty important.
My solar came back during the outage. I assume Tesla was able to remotely force it back to running at 60Hz after it got escalated to an engineer. I hope that Tesla gets this figured out because there is really no reason it should be blocking my tiny 1.5kW of winter solar generation from going into the two Powerwall batteries while the grid is down. Altogether we had three outages yesterday: 2 minutes, 5 hours 39 minutes and 2 hours 33 minutes. The longest one was because the middle of the 3 overhead distribution wires broke and fell down to the roadway about 100 yards from my house. I suppose the good thing is that I found that I can get the house down to ~300W total consumption by turning off non-essential stuff while still keeping the Internet, DVR, and a LCD TV going. The Powerwalls only got down to 42% a little after midnight when the power was finally restored the last time.