The way the 2 ring detectors detect current is well documented and used by 100’s of devices including measuring tools.
In Homeline/QO panels in reality the 2 wires feeding the main lug are spaced very close to each other. (Split panel: residential). If one attaches the white probes their respective bodies actually end up touching each other or almost touching. I measured this using an advanced measuring devices and there is definitely spill-over of inductive current. I decided to temporarily put a non-conductive material between the two white clamps. Within 24 hrs the Sense discovered 2 additional devices!
I am strongly recommending Sense / SE develop and market an insulating product (2" to xx") which can be placed between the 2 clamps negating all spillover inductive current. I believe doing so will greatly increase identification of devices.
Hi @baivab.mitra - thanks for sharing this!
I spoke to our team here about a few of the tings you mention. In our installation instruction, we suggest that you keep the CTs apart from one another. If they are actually touching, the magnetic field from one can influence the other.
It does sound like the device detection right after spacing them apart was almost certainly a coincidence, just want to level-set expectations on something like this.
I read installation instructions. The point is: I have installed it in a brand new Homeline series 200A 30 space main breaker loadcenter. I also helped install this is my neighbor’s new loadcenter QO 200A also. Now, in both cases by virtue of design the incoming lines once clamps are put are really tight. All 3 products here are manufactured by same company Schneider Electric and not only that SE certifies it for Homeline and/or QO panels. Heck: there are now panels available with Wiser installed.
So, when you say what installation document states I get it but then since the underlying certified to work products have this issue i am suggesting you come out with something which can be pushed in the middle. That is all.
I do not think its a coincidence though. Even today a new device was found. Its as if everything woke up. BTW: I have had this now for 1.5 years. Only recently has 4 devices got identified after I put the insulating block.
Could you please post some pictures of your installation and maybe your neighbor’s installation.
I would like to see the raw data where you measured the ‘spill-over of inductive current’. What is the device you used? Pictures?
What material did you use to to separate the two coils? Can you provide more details and specs?
I’m really interested in learning more about your findings. Thanks.
This is what I’m interested in. I would guess you used some type or piece of plastic? I’m thinking about a plastic cap I could use to separate them as I think mine are touching.
Just use a zip tie to hold each CT where you want it on the service mains.
As per NEC codes specifically told by inspector not permitted. That’s what the electrician also said who installed my panel. But I still made him do it and then the inspector made me remove it. In fact, zip tieing neutrals and/or grounds on either side is also not recommended.
The front panel is closed and inspector put a sticker. Technically I can’t remove it only an electrician can (licensed). But to answer the question:
Initially thought of plastic but didn’t help. What helped as a small piece of chicken wire sandwiched between two narrow wooden pieces. As for the meter, it was my electrician’s but vaguely remember the name of the company - TRIFIELD. He also used another meter made by something called: Fluke? Flake?. Not sure. But that one was a whole gizmo. He said the chicken wire really helped. Mentioned something about Faraday but no idea. He’s actually not an electrician but an electrical engineer who works at a senior position in public utility company. As he knew my dad he did me a favor.
Can i give Data Science permission to look at your monitor data?
Hi there @baivab.mitra - i forgot to ask when exactly you installed the non-conductive material.
We recently put out a new firmware version that made some changes to our detection infrastructure and I’d like to be able to rule that out as the cause here:
I responded above. A 2” x 2” cut of something called chicken wire fence sandwich between 2 similar sized wooded pieces. The individual wood pieces are just few mm thick.
Of course. Anytime. BTW: I sent you a note separately earlier. You have my implicit permission. As you said sometime back I was one of your earliest adopters and you/your team had accessed my data earlier. You can access anytime.
I passed your monitor information along to our Data Science team and we did not see any substantive change to the wattage data over the past few weeks. Additionally, your house has been up for a while so the cadence of device detection has naturally slowed.
Even in a comparable situation where someone fixed an open CT, we would expect to see an influx of new devices within more like 1-2 weeks at least. So yes, on our end this seems like a coincidence and we have no hard evidence of this change impacting the detection process @baivab.mitra.
Not sure. After I did what I wrote four devices were found in quick succession:
Motor 2
Microwave 3
Fridge 3
And Heat (some #).
As you said device discovery slowed and was zero for quite sometime. After I added it within 2 weeks these were identified.
There are few other observations. For example quite a few times of late Sense was incorrectly identifying a device. Now that has got corrected. I mean: Sense already identifed a fan and xx. Then it started showing XX instead of usual: fan. After I did it, Sense started showing fan! Again.
Did they provide a code reference for the zip tie prohibition?
Guessing the inspector is basing the “no cable ties” on remembrances from many years years ago. Things change, but people often don’t. That said, whether they be right or wrong, you need the inspector’s approval.
UL listed cable (zip) ties ARE allowed in electrical panels and used quite frequently. In fact in some specific instances (irrelevant to this discussion) they are required 210.4(D) 200.4(B).
There is plenty of spacing between the main lugs in Modern Homeline or QO panels. Some of the older panels from the 60’s could be an issue fitting CT’s but even then they could be offset, with one higher/lower than the other.
That is the problem. Try figuring out as an inspector if a cable tie is UL listed. He said no joy, remove them. He also said you will thank me later. Many times you have to trace the wires, without cable ties: easy. With them: no.
As for NEC code he specifically mentioned incoming service line cant have grouping. Cable ties create grouping.
Coming back to the main tooic Sense guys seems to have not found anything. But my view is after the change Sense does behave differently. The devices, bubbles, etc. also what the heck was the chicken wire for? Can someone help explaining please?
210.4 Multiwire Branch Circuits
(D) Grouping.
The ungrounded and grounded circuit conductors of each multiwire branch circuit shall be grouped by cable ties or similar means in at least one location within the panelboard or other point of origination