Sense solar setup instructions seem wrong


#41

Seems strange to say it is the calculation as mine is accurate and so are most other people. Is it possible that yours did not initialize correctly and has some sort of off-set that is causing the issue. I think the only way to fix that is to unplug the solar and go through a full initialization but that may loose all data. Suggest a new support ticket and a new post here as well. I think you might get more input on the FB forum as there are more people there and some really good tech people as well. IT IS A SHAME THAT YOU HAVE NOT GOTTEN A RESPONSE FROM SUPPORT FROM THIS FORUM!!!


#42

I can try a recalibration, can’t hurt anything really. I do feel like people are probably tired of reading about it so I give up posting about it.

You can see the effect in the picture below pretty well, where the usage graph dips during the day as the solar rises. My solar is underreporting 3-5% at any given time, representing 1-200w during peak production. That 1-200w is the exact figure that’s missing from my daytime data, and the reason why we’re typically humming along at 2-300w at night with nothing on, vs. <50w during the day with nothing on. However the solar values are being calculated, they’re being miscalculated and not only are they resulting in deficient solar graphs (here’s that assembly of data from a while back: http://imgur.com/GgTBOwF), it’s incorrectly factoring into the use graphs during the day (below), creating the “Daylight dip”:

We’ve got some banner days coming this week, I’ll recalibrate and see where that gets me on those days.


#43

Don’t know what to tell you, but those graphs vis-a-vis solar prompted me to take a look at mine and the shape of your graph. At first I thought mine looked very different but then I saw you had a portrait graph rather than landscape. Where are you located and what is the size of your system. Mine is 10kw in Morris County NJ, this is the graph from yesterday using a scale just about the same as what you used:


#44

Would you agree there’s a daylight dip in your graph? :slight_smile: Your home is obviously more busy than ours, but I see the trend. It’s especially obvious to me on days where no one is home and the only thing running is the fridge and whatever always on items we have.

We’re in Phoenix, 8kw system about 6 months old. SolarEdge inverter sending power through a large solar breaker with a meter and feeding the panel from there with what I’d guess is 8 or 10 gauge wire. Not quite the diameter of a pencil.

Here’s Monday:


#45

And what I think I’m going to try to do is recalibrate starting with the solar CTs off of the lines - in case there’s some sort of backflow going to the inverter or solar meter or something during the zero out portion.


#46

That is a good idea, if you take the solar out of the equation that would let you really prove the problem. You probably have better sun than we do at this point in the year. I won’t see anywhere close to bull output till May. Yesterday was actually very good day with 45kwh produced. At peak output mine will yield 60-63 kwh for a day. Not bad considering it is 12 years old now.


#47

I feel your pain. I’m still waiting for a response to tech support from 2/17/2017, and know many people have waited what seems like or is, forever.

Tech support software to manage the support is readily available, and for every tech support email going to the same e-mail address, without a different subject line with the support # in it, makes it real difficult to find the emails associated with a problem.

Perhaps letting @Jonny, @Mary, @jonah, @Maarja-Liis, @cui-lyn, @MattAtSense know that even if they are short-staffed, that customer service is important to the health of their startup.Some general information spread on their forum can go a long way to reducing their tech support questions.

Also, answering questions here in the forum allow other forum members to spread the knowledge to newcomers, and shows that somebody is listening, reducing the anxiety of the end-users as they wait for improvements.

Automatic detections of polarity of solar and the mains (as explained above here), as well as a manual configuration reset in the apps might go a long way.


#48

Hi everyone -

Sorry we have been super busy lately and we clearly need to work on increasing our attention to the community. My name is Hilario Coimbra and I just joined Sense recently as a product manager - I have a lot I’m getting up to speed on (and learning a lot from reading all of your posts!). In fact, the Sense team is still growing and we are close to hiring a new community liaison to better (and more frequently) participate in a dialog with all of you! And I myself hope to respond more often here from time to time, and taking all your great feedback & thoughts & frustrations into decisions of how we can improve the product experience.

Now, for the specific questions and issues I’ve parsed from this thread:

  • Confusing installation instructions for Solar CTs: Although the instructions recommend installing the CTs in a certain way, Sense does try to figure out the polarity and position of the CTs so it is not absolutely critical to get this right (no matter the wording in the manual we know lots of people won’t). But, if you do it the way we have in the manual the initial signal checks can be faster. Basically Sense goes through an initial time period of figuring out if they were installed the way we expect, and if not, after a few hours it should auto-adjust. This is for the INITIAL installation.
  • Moving CTs after install and initial signal checks have been done: If you move around the CTs after Sense starts to auto-detect then it does interfere with the automatic signal checks! It seems like we haven’t been clear enough about this in our documentation – sorry (we’ll work on it)! Once things do get moved around, sometimes we do detect problems ourselves and correct behind scenes, but not always. For your specific cases, you’ll need to work with our customer service team so they can reset things for you to correct it, so please reach out to them if so (you can reference this thread and/or mention my name). We are working on a new “data reset” function in the application to let you do this yourselves in the future… look for that in the coming weeks.
  • Solar under-reporting (@NJHaley’s reported issue): Based in your reporting about this (thanks for all the details!) we’ve been looking into it and have found that we did have a solar calibration issue with some monitors (it was related to the voltage issue you guys found a while ago – but in a subtle way!). We are working on a fix now – it is easy to do but we’re trying to make sure we know exactly which monitors had the problem.

Hope this helps! Let me know if you have any questions.


Solar customers: Recognizing the "Daytime dip" and how it might impact your Sense numbers
#49

[quote=“HilarioAtSense, post:48, topic:380”]Hi everyone -

Sorry we have been super busy lately and we clearly need to work on increasing our attention to the community. My name is Hilario Coimbra and I just joined Sense recently as a product manager - I have a lot I’m getting up to speed on (and learning a lot from reading all of your posts!). In fact, the Sense team is still growing and we are close to hiring a new community liaison to better (and more frequently) participate in a dialog with all of you![/quote]

Glad to have you join us. It’s nice that Sense is hiring a community liaison person, but Sense really needs the engineers to be responding to the beta test feedback, not some community liaison person who doesn’t actually write any of the code.

If both ways work, then the solarinstallguide.pdf should give consistent instructions. Install both pairs of current sensors with the labels facing towards the power source, or both facing away from the power source, but not utility sensors one way and solar sensors the opposite.

And the current solar instructions are even more ambiguous than that. It says the solar sensors have the labels facing the breaker. Which breaker? As you will have read, some installations have more than one breaker, with the solar sensors installed between two breakers. Which breaker are the instructions talking about? The instructions should say, “towards the inverter,” or, “away from the inverter.”

By the way, the current solar instructions document still has blank placeholders where images are supposed to go, and hasn’t been updated for months.

You say that installing the solar sensors in both directions will work, but that appears to be untrue. I ran the solar setup multiple times, and every time it identified the solar sensors’ phases backwards. I swapped the solar sensors, and ran the solar setup again, multiple times, and still it identified the solar sensors’ phases backwards every time. Then I flipped the orientation of the labels to be facing towards the power source, ran the solar setup one more time, and then it immediately identified the phrases correctly.

And other people here have reported that their installations have misidentified the solar phases too, with the current sensors installed as specified in the instructions.

I’m only wasting time on this at all because Signals Setup still says, “0% Complete”. If the Sense monitor were operational, I’d be happly using it, instead of still struggling to get it to work.

Sense customer support has moved on from taking a week to respond with a suggestion that I try logging out and logging back in again, to now taking a week to respond with a suggestion that I try deleting the app and reinstalling it (which, unsurprisingly, performs the function of wasting my time, but otherwise has no effect whatsoever on the operation of the Sense monitor).


#50

Glad to have you here. Please also check out the Sense user group on Facebook, over 360 users on that group so please check in from time to time


#51

Got my latest response from Sense customer support today:

A new iOS update will be released within the week with new fixes. If you still experience this issue after updating the app, I would have to reach out to our Engineering team.

After six weeks of the Sense monitor not working, they now think that maybe it might be time to start thinking about whether to share that information with the engineering team.

I think someone is not quite getting the concept of a beta testing program.


#52

Hey guys -

So first off, you are totally right on the Solar install guide needing to be updated. In fact, our team has been working on that, and we just updated it today: https://sense.com/help/solarinstallguide.pdf. It is now consistent styling wise with the non-Solar guide, has some nice new visuals and the instructions have been re-written for the solar sensors: “Clamp the solar sensors around the wires to the incoming solar feed so that both labels face the same direction. The direction of the sensors does not matter, as long as they are the same.”

@Stuart - specifically for your 0% Signal Check problem, we passed your issue on to our engineering team and they indeed were able to hunt down an issue. Someone from Sense support should be reaching out to you shortly on that!

@Howard - I’ll be sure to check in on the Facebook group as well. Thanks!


#53

I think the new guide is fantastic. Solid improvement over the old one, visuals look great and are sure to be an asset to either electricians doing an install or customers who are electrical professionals themselves :wink:


#54

It’s nice that the instructions say that, but I really don’t think it’s true.

If the two solar phases are almost perfectly balanced at the time of installation (as they generally are) there is absolutely no way the Sense monitor can tell which phase is which without knowing the orientation of the current sensors. And if the Sense monitor does get the phases backwards there is no way you would ever know, because, by definition, if you could tell by remote observation, then the Sense monitor wouldn’t have got the phases backwards in the first place.

I think it’s highly likely that every single Sense solar installation so far has the phases identified backwards, because the only way you could tell if the phases are identified correctly is by human observation, and since of of the sense of monitor is not doing any human observation, but instead is just assuming that it’s correct, then it’s probably not.

Everyone so far who has actually checked visually (including myself) has found that the solar phases are backwards.


#55

The solar install guide does look better, but it says “Connect the solar sensors, labeled with the “sun image” to the middle port on the monitor.” I don’t believe there is an image of the sun on my solar CT sensors.

In fact, the part numbers on both the mains and the solar sensors of mine have identical part numbers (I will double check tomorrow just as a sanity check). It would have been ideal if you used a keying pin or different socket for each CT pair on the hardware.

Also, I believe, that having the left side (as displayed on the Sense Monitor) of the Mains CT correspond to Solar CT that feeds the Left side of the Solar is important, and similarly for the right side. This can be done experimently, or by proper labeling of the Left/Right CTs for the Mains and Solar. The Solar should generate almost the identical values on both sides, but it is not precise, so it is better to get it correct.


#56

My guess is that, if you buy it now, the supplied current clamps are labeled to identify which are for the main supply and which are for the solar.


#57

Mine appeared correct - once I understood what you were referencing I was able to say that my right column was on the red lead and left was on black. But then again I installed mine facing the “incorrect” direction at the start :slight_smile:

I agree on the updated clamps, too. Although the ones I got this week were not yet labeled that way.


#58

Assuming that is correct, how would one identify the correct pair if it doesn’t have a difference and has the same part number? And why would they be different? They are both measuring the same order of magnitude currents, so they would have similar windings on the CTs I suspect.


#59

The instructions tell you to be careful not to mix up the sensors packaged with the monitor unit, and the sensors packaged separately in the solar box.

I’m guessing that some people weren’t careful, which is why they are now labeled.

Because the sensors packaged with the monitor unit were calibrated with the monitor unit before shipping.

The sensors packaged separately in the solar box have not been calibrated in advance, but the Sense monitor is supposed to calibrate them by comparison with the (known calibrated) sensors that were shipped with it.

To get the accuracy Sense needs you can’t just take a batch of current sensors and assume they are all identical. Each one needs to be individually calibrated.


#60

Sadly I did keep them separate, but the electrician did not :frowning: as he installed them and temporarily moved them around.