Thank you for taking the pictures and daring to try it !
Looking forward to the full sun day report!
1 Like
As long as both Inverters are generating 240 volt, the current through L1 (red) and L2 (black) will be exactly the same.
Since Sense can differentiate L1 & L2 for loads and probably also for PV generation, it doesnāt matter if the one CT is on L1 (red) and the parallel CT is on L2 but with the arrow reversed (180 degrees out of phase )
As soon as it would be something 120 volt related, that would be making things go out of balance.
But with only 240 volt inverters, you should be good!
1 Like
Good news, and New issue. The CTās in parallel seem to accurately adding both solar system power output. So that issue is fixed (thank you, thank you again!!!) , but I noticed that when the solar systems are producing power, my consumption data goes to zero (0). When the sun is down or if I begin to use more power than I am producing I see my consumption go up from zero. Looking back in my Meter history on the Sense App, I can see that this started on the day they installed and turned on my new secondary solar system which is connected to a garage sub-panel. Because it started then, I can confidently say, this has nothing to do with me connecting the solar CTs in parallel, since I did that about a week after they did the original install. Could it have something to do with where on the 240 breaker they placed the Red/Black wires? I know it normally should make no difference, but if Sense saw a certain setup before and then another power source was added (secondary solar system) maybe it confused it?
2 Likes
Can you post a picture/screenshot of the sense reading where that shows ?
Says a 1000 words.
Are your 2 inverters both connected to breakers at your main panel ?
Or do you have a so called supply-side hookup ?
You mentioned the new inverter is hooked up to sub-panel, so that is load side for sure.
How about your initial inverter ? does it have a dedicated breaker or a fused disconnect ?
reason for asking: Sense tries to automatically determine if you have a supply-side or load-side hookup.
For unknown reasons, for years, it sometimes switches between the 2 modes.
And then you get what you are describing.
Search on this forum āsupply sideā and you will find many articles where things suddenly changed.
eg: Bug showing zero usage? - #4 by dannyterhaar
It only needs 1 email to support to fix it (if that is what has changed)
Looking forward to more information from your setup.
My original inverter has a 30amp double breaker within my main panel in the basement. The secondary (new) system has a 20 amp double breaker connected to the garage sub panel. The garage sub panel is fed from a 50 amp breaker on my main panel and has a 50 amp breaker in the sub panel itself. This way I can shut off the garage sub panel from my main panel in the basement or from the garage within the sub-pane itself. I will include pics as well.
- Pic 1 and 2 show data from the sense App. When Solar stops being produced, the value goes above Zero and I see Positive numbers vs negative numbers on the MAINS.
- Pic 3 - is of my garage sub panel. The feed from my main panel in the basement is on the top left. The solar breaker from the secondary new system is on the bottom right.
- Pic 4 - shows the Sense Consumption CTs in my main basement panel. They havenāt been touched since sense was installed several years ago
- Pic 5 - The Bottom double breaker on the LEFT is 50 amp and heads to my garage sub panel. The bottom breaker on the RIGHT is 30 amp and is the solar breaker from my original system.
1 Like
Okay, so totally a load side configuration.
In that case, when the second inverter got installed, my gut feeling tells me that the increased PV power triggered the (software) switch to supply side.
Mail support@sense.com and ask them to check if by any chance the software did change it and ask if they can reconfigure it manually to load-side setup (again).
You will not get your information back since you installed the 2nd inverter but from the moment they switch it back, it will look normal (== as it should) again.
Best of luck.
Will do and thanks again. I am ok with ābad dataā in the past as long as I get good data in the future 
.
Iāll keep you posted.
2 Likes
I updated my 2nd last posting with a link to a thread on this forum where the same happened.
Hope you have seen it, if not, reload the page in your browser
1 Like
Just to give you an update: I emailed Sense to check if something switched when I added more panels. They emailed back and said that it was normal to NOT see usage when solar was producing a lot of power, but they noticed āa setting issueā which they claimed to fix. All it did was mess everything up. It now appears as though the usage data is taking one leg of the solar production, multiplying it by 2 and then adding actual usage. But when the sun is set for the evening the usage data goes in the negative. haha. I have been updating the ticket with all this info and they said they will work on it and to wait 48 hours.
1 Like
What a balony!
You can always see if you use power and yet PV produces more than is being used.
I just got home and started charging my EV while clouds sometimes interfer with the PV production.
But you can clearly see both.
The answer you got is from the L1 helpdesk people and my guess they are limited/schooled in what they should respond and never NEVER admit any fault.
DDD (deny deflect diffuse) is what you saw here.
deny: āit is normalā
deflect: they noticed āa setting issueā
diffuse: we fixed it for you.
picture please!
Yeah. I agree. Heās obviously wrong. I can see my āusageā track solar production. Totally messed up now. Iām gonna give him the 48 hours he asked and then ask to escalate the issue. I have always seen usage before I added 9 panels. Ugh. I know it will get fixed eventually but this is unnecessarily annoying. lol.
This first pic is crazy . Check out that usage when no one is home and all appliances are off.
The second pic clearly shows the āusage ā track the solar production. Again on a cloudy day when no one is home.
Kenneth R. Herbert, Psy.D.
(mobile correspondence)
1 Like
Hi,
To chime in, we certainly donāt encourage our agents to never admit fault if they make a mistake some styles may vary, but we are focused on getting your issue resolved, hopefully quickly.
Also, signals can frequently have issues or get scrambled due to install errors or changes in a userās setup.
Frequently, we need to gather more data by waiting a few days to see if the day/night signals for several days look basically as we expect.
Finally, we can make some adjustments on our end to correct issues like reversed polarity or signals appearing in channels they should not.
The issue here is that we canāt always tell if the issue is resolved for the next few days until we get a fresh set of data. Just looking at how solar and mains are producing instantly after we make a change can give us an idea if something is correct or not, but it rarely confirms if everything is working without a few days of data as normal.
Iām willing to bet it was a misunderstanding if they said it was normal to not see mains usage running when solar was running. If this is the case, send me a private message with the text and a link to the
ticket so I can make sure the agents understand the question.
We want to make sure you have a good experience.
Thanks again.
1 Like
Thank you so much. Your explanation is very helpful. It is good to know they are working on it. Sometimes the emails are brief and when I dont see anything happening on the monitor end I begin to wonder if they were or werenāt still trouble shooting. Sense is a great product and I watch it more than I probably should. Haha. They escalated my case to a different Tech this AM, and he said as well that he needs up to 48 hours to see more data. I will be happy to send you the open ticket.
Thank you again!
Ken
1 Like
A search on this forum reveals that the change is suddenly by the software of Sense and has been noted for many many mannnnnyy years now.
I assume user errors are made, but this particular event has happened to me multiple times without any change on my side.
Every time I asked "can you lock it in so the software can not change it again "?
Answer is always āthere is currently no way to do thatā
And it happened again with me.
Do you see an IRL scenario where a supply side ā load side switch is done ?
Why is the software even checking for a possible switch after initial setup/config ?
That is so weird.
A customer opens support ticket, says āsomething has changedā
and when changes are made you only look forward instead of how it was okay 1 hr/day/week/month before the ticket was opened ?
Easy to spot if the setup went from supply-side/load-side.
You see the differences in the graphs, you check the setting. change if confirmed it changed.
You apologize to customer and tell it to keep an eye on new graphs to show it is working again
You open internal ticket that the software once more switched a basic setting of a working customer setup and ask for update in the detection algorithm.
How hard can that me ?
Probably not the way it works at Sense.
But when I had a company with 85 employees, that is how my company was run.
Total transparency.
You keep approaching it from the Sense point.
Look at this case (and many others over the last years) where customers get no notification/warning just suddenly graphs that are non-sensible because āsoftwareā made a change.
And the response āit is normalā was an error ?
No it is a pattern.
And from experience I can you how frustrating it is so have to convince the support people you are communicating with, that it is nearly impossible to convince them of the facts and that āthe softwareā changed something. As end user you are not able to see what changed, you only see the graphs.
I knew what happened but still had to plea with them to look into the supply/load side setup and change it for me.
Feedback is the breakfast of champions.
No matter how much feedback the sense users who shared their experience of the sudden switch from load /supply side have surfaced here, IT STILL HAPPENS as seen in this case.
And in my view Sense has no intention what so ever to tackle this issue.
To my knowledge, switching polarity, at least automatically, is not done, though it does happen sometimes for whatever reason. Usually, this is due to someone changing something on their end.
We can also flip the poalirty so an agent could make this change as well.
The reason that the system checks for them and āflagsā them is information that hints at an incorrect or bad install.
That is so weird.
A customer opens support ticket, says āsomething has changedā
and when changes are made you only look forward instead of how it was okay 1 hr/day/week/month before the ticket was opened ?
Easy to spot if the setup went from supply-side/load-side.
You see the differences in the graphs, you check the setting. change if confirmed it changed.
You apologize to customer and tell it to keep an eye on new graphs to show it is working again
You open internal ticket that the software once more switched a basic setting of a working customer setup and ask for update in the detection algorithm.
How hard can that me ?
Probably not the way it works at Sense.
But when I had a company with 85 employees, that is how my company was run.
Total transparency.
This community is more for other users to talk to other users. I am here to help clarify, provide answers, and see what I can help with. My whole day is spent explaining and fixing issues for people. I personally would not characterize Sense like that, but then, Iām an employee, so itās a matter of opinion. 
The way Sense analyzes signals is not just up and down. Even if we make a prediction based on past
data, only current and future data can confirm it.
Itās often a simple fix, but not always. I want to see future data after I make a change to absolutely confirm that things are as they should be. Sometimes, making adjustments can also show unforeseen setup issues.
The alternative to not monitoring install issues is to have bad installs that customers may never flag and thus unless a customer reached out will never be fixed.
As of now, all potential bad installs should be flagged eventually and are checked (anonymously for the customer) by an agent.
If anyone sees anything wrong with how their Senseās signals are acting, send me a PM.
https://www.reddit.com/r/Sense/comments/1bfjie1/sense_suddenly_not_showing_the_house_using_any/
One guess ā¦
ā¦
Likely is easy to fix if the customer contacts us.
1 Like
Oh, manā¦ I am glad I found this thread!!
I have two feeds coming off my main, as well as solar, and so Iāve only been able to monitor one of the feeds until I decided to try to just run two CTs in parallel.
I bought two new āaddonā CTs (I didnāt want to āeditā my original ones) and spliced them together but then was trying to figure out why it was reading exactly half!
This would explain it!
Iāll crack open one set of the CTs and removing the resistor in each - altho I am guessing this will solve my problem
Thanks for the pics @herbertk1 !
-Taner
Err, hmm.
The newer CTs (which have oblong openings instead of circular) donāt appear to have a non-destructive way to open them up. 
Anyone know if thereās either a compatible off-the-shelf CT I can combine w/ an official Sense CT (so that one set will have a resistor) ā or does anyone know if Sense still sells the old model CTs?
I suppose I could cannibalize one of my old sets, but I was trying to not destroy them, since I did that to the new ones, instead. But perhaps I donāt have many options.
Does anyone know the winding ratio for the Sense CTs?
-Taner
1 Like
You are correct. There is no way to remove the resistor without a lot of careful destruction. Haha. The Newer models are much harder to crack into than the older ones, but I have found by experience that the 2nd set of CTās in parallel need to be the exact same model. Originally I used an older Solar CT with a newer Flex CT and the values were inflated in the solar production but also messed up my usage data where the usage would significantly inflate and mimic solar production. When I put 2 identical Flex CTās ij parallel, the system works really well now. No more usage mimicking the solar production and the solar totals is very very accurate. If you find an aftermarket one that is identical to the Sense ones, then I am very interested. Although my set up is now working I made a mess of the second CT flex pair. I used a combo of a fine Dremel tool along with a soldierly iron to melt open some parts.
2 Likes