Solar install screwed things up!


I installed Sense in March - I knew at that time that I would be installing Solar in late April/early May so I ordered the solar package. I installed Sense without the Solar CT’s and fired up in March - and I was quite pleased with the results. Over the first 5 weeks or so I had about 16 devices detected including big users like dryer, oven, hot tub, toaster oven, refrigerator, and several small devices as well. Still had a long way to go - about 60% unknown.

In late April my solar system was installed. I was excited to use Sense with solar so I quickly installed the Solar CT’s - and in the process adjusted the main CT’s a bit to make everything work in the panel with no chance of pressure on the clamps.

After installing solar all hell broke lose in my setup. The calibration failed a few times, saying “swapped mains” and some long numerical error message, but did eventually go through. Afterwards, the main voltage read negative instead of positive. After a few signal checks the Sense monitor did seem to correct that - and things seemed to be working. The wattage usage during the day was quite a bit off from before, however. All of a sudden it seemed to be underreporting wattage usage - I saw values of 7-10 watts at times from one of the mains, which makes no sense. And the device recognition was suddenly way off - was recognizing incorrect devices and saying a device was on when it wasn’t.

I was a bit premature in setting up my solar as the system hadn’t “technically” been signed off by the utility company - so I had to turn the solar back off to wait for the final “go.” Since everything was fairly screwed up by that point, I decided to do a factory reset.

I set up the Sense system from scratch, and this time said to disable the solar setup (even though the solar CT’s were still installed - the system was powered off). For whatever reason, this time the monitor would not go through signal checks - I spent 6 days waiting, and counted 19 separate error messages regarding a problem with installation, deleting data and re-starting signal checks.

I spoke with support a few times, who suggested I wait things out - but nothing improved. Then this morning they said the problem was from a weak WiFi signal - a strength of -77 at the monitor, and inconsistent data was causing the failed signal test.

This is weird to me - granted the signal is not perfect outside where the panel is but never had any drop outs of data, and never had any issues with the signal check before the solar snafu.

Anyway - solar system is all good to go now, so did another factory reset this morning with solar calibration, and everything seemed to go through well. Still waiting on signal check.

We shall see how things go. Anyone else go through anything like this?


I think you had a bunch of things going on at once. When you say one of the mains went negative, that is normally an indication that one of the solar CTs is not fully closed. What happens is you are producing more power than Sense is recording thus when it does the calculation that power shows up as less power on the main. As far as the long time going through signal check I had that on my replacement unit, after two days I contacted them and they had to reset something at their end. Should not take anywhere near as long as what you have waited.


Could very well be (the issue with mains going negative) - although it was both mains, not just one. And when the Sense went through it’s signal check process It seemed to reverse that on it’s own - without any change on the CT. Regardless, I am essentially 100% sure now that the CT’s are within the panel and no pressure - should be fully closed.

I have read about the issue with solar production not reporting correctly on the Sense - I assume this is the “daytime dip” I have read about? The posts seemed to imply a fix was imminent but that was a few weeks ago - any news on this?

I am suspicious that the Sense does not like having the solar CT’s connected but having solar disabled - and perhaps that is why the signal check failed, but the tech support folks seem to think it’s related to WiFi strength. We’ll see how things go this time, now that the solar system is calibrated and producing.

In the meantime - should I expect device detection to not go well until the daytime dip is fixed?


Here is another question - during the solar calibration the app asks for the inverter to be turned off and back on again. No problem. However - in my system I can either turn the inverter off with a switch on the inverter itself, or with a disconnect between the panel and the inverter. When I used the switch on the inverter - the app still read a few watts - when I used the disconnect it showed 0 watts. Which do you all think I should use to turn the inverter off during the calibration - or does it matter? I used the disconnect, since it seemed to make more sense to have a reading of 0, I felt, to avoid calibrating out the slight negative draw at night from the inverter itself.


There is potentially some news on this, perhaps @BenAtSense can chime in. The answer is still “imminent.” :wink:

I also agree that this error is likely leading to device detection issues. The earlier sunrise and sunset get to the times my water heater is set to turn on, the more frequently my water heater isn’t detected as on (among other anecdotal observations) - as now part of the power is coming from solar, whereas in the winter all of the power was coming from the utilities at 5am/5pm. .

When setting mine up, I used the breaker betwixt the panel and the inverter, not the switch on the inverter itself. I don’t know if either is “more correct” than the other.


You should probably use which ever method gives you zero output from solar to accurately calibrate. That said I don’t think in the scope of things it really matters as if it shows just a few watts that is very little error overall.


Count me in the Solar Sense users with similar issues.
I have a small negative reading on one Mains leg, just at night. Not convinced that is a CT closure issue.
Slow to detect even what would appear to be easy Devices. Hasn’t found the majority of the high power devices in my all electric home after 5 weeks.
I’ll wait some more. Don’t think the problem is an easy one to solve.


These issues are a bummer - hopefully fix is coming soon. I felt my Sense was working well before enabling solar, so perhaps if this can’t be straightened out soon the fix is to return the solar CT’s and go back to a non-solar install, relying on the inverter reporting software to determine solar production.


I can confirm that Sense sailed through its signal check after calibrating solar. I do not think wifi strength had anything to do with it failing, but do think that having the solar CTs connected but not enabled in the app may have caused the failures before?


Watch that wifi because many times it’s not obvious. Sense can backfill 7ish hours of data so you can have intermittent loss that looks like nothing is going wrong - Sense is very quickly filling the gaps back in.

Once in a while, though, you’ll see your unit is “offline,” occasionally long enough to see the gaps and sometimes so long you have to reboot your router and then try to come up with a permanent solution involving a closer access point.


If you have a negative reading on one main at night, even if small, you have a calibration issue. Assuming your home is not totally unbalanced like none other the two mains should not always be significantly different. I am assuming what you are saying is you are seeing something like -50 and perhaps 1,200? Maybe you could provide a sample of what you see day versus night.


I just wanted to confirm my experience with the solar daytime dip. I have seen several times where the usage wattage drops as the solar generation increases. Here is a screen shot from this afternoon when I shut down my solar inverter. You can see at the exact second the solar generation goes to 0, the wattage used jumps up. It’s a slight but relevant jump. Immediately before hitting the breaker I was pulling 831 watts, per the app. Immediately afterwards was 1038 watts, with no other changes in the house. This is a change of 207 watts. My solar was creating 8507 watts just before it shut down. So since I believe this to be an issue with mismeasurement of solar wattage and then incorrect calculations (right?) - seems to be a 207/8507 or 2.4% variance from correct solar production. Does this jive with everyone else’s understanding as well?


Here is another example where the dips in production during clouds correspond exactly with a slight increase in reported usage.


Yep - that sounds about right.

On cloudy days, it makes for some really choppy waters :slight_smile:

I’ll be glad when we can put this issue behind us and maybe get device discovery back on track :wink:


Sorry that you had that experience, @garywitt29!

The day time dip solar calibration fix is still imminent. It unfortunately took longer than initially anticipated but it should be right around the corner.


HI @BenAtSense: any updates on the solar calibration fix?




Expect movement on this next week! A monitor firmware update should be pushed in the next few (business) days that will get the infrastructure in place to apply necessary correction factors. I’ll provide another update when this firmware update goes out.

closed #18