I’m in the middle of building a new home, running low voltage wiring now. I am looking for advice and input on where to wire for PoE cameras, sensors, etc. Has anyone come up with good rules to go by, sweet spots to position occupancy or motion sensors?
More than one corner of the room for bigger rooms? How high off the floor? What am I missing? I am starting to realize that wall mounted occupancy sensors are better than ceiling mounted (field of view).
Doors get 2 lines, windows get 1 line, each corner of the roofline, peak of the roof for starters. Anywhere a TV, fridge, stove, dishwasher will be. Bathroom fan. furnace/ac.
Also consider different color jacket for different uses, like camera, security, general outlets and lights, that kind of thing.
If you are going to run one cat6 cable to a room then run two. It’s a lot easier than trying to expand later and the cable is relatively cheap.
For your main living room or entertainment area consider runs from the front to the back of the room for entertainment devices like projectors, HDMI over cat6, etc.
Tips:
It depends a bit on how much money you want to spend up front and how much you’re willing to do later when you don’t have wiring where you wanted it. If money was no object, I would run the following in each room (including the garage):
I would also run:
The thing is, in a 4 bedroom house that could be upwards of 60 cable pulls. Sure, they aren’t very expensive, but they add up. I already got funny looks when I ran 25 in our new house. I do have a couple places I wish I had run stuff, but I have a ranch and easy access to the attic, so running more isn’t that big a deal.
Right on point. Also consider where they’ll be aimed… they can see through walls.
On top of what Sir_Goodenough said
Honestly the wire is comparatively cheap. Had I the means I’d run PoE lines to almost every wall inside the house even if I had no current obvious use for them, leave the unused ones coiled up on the lower wall plate, and bury them under the drywall with a schematic to their location.
Keep a wire pull conduit open between you server and the attic/basement if you can just in case.
If it’s a fresh build and you’re looking to spend some real time there, future proof yourself… tech moves fast and QoL improvements that you didn’t know you needed is a thing.
One wild idea is LED lighting from POE. I have read about it but have not seen much use. Something to look into.
I’d just like to note how much I enjoyed seeing @HappyCadaver replying to @traumadoc…
[Correction, added Sep 5] The measurement below is wrong, and the mechanism I gave for it is wrong. I had only looked at the x marginal. Plotting the joint (x, y) distribution over the full 275,038 pairs (nineteen days, not four) shows the points I called “through the near wall” are specular reflections — mirror images of people inside the room — not people in the room behind it. Three real clusters, 824 mm apart, imply a mirror plane that agrees to within 85 mm, and the phantom share is 8.3% of all points, not 1.5%. The adjacent room is also a nursery, not a bedroom, so nobody was lying in bed. What survives is the placement advice: keep a radar away from a large flat surface it can see, or aim along it rather than into it — that holds more strongly for reflection than it did for transmission. Sorry for the noise; I will write the reflection result up properly once I have confirmed what the surface actually is. Original below, errors struck through.
One thing I did not expect until I measured it: how far the sensor sits from a wall changes how much it sees through near that wall.
Living room, LD2450 on a desk, 272,000 coordinate pairs over four days 275,038 coordinate pairs over nineteen days. The sensor is roughly 0.5 m from the wall on one side and 3 m from the wall on the other. Clipping the points to the actual room:
through the near wall (0.5 m away) 1.5% of all points
through the far wall (3 m away) 0.01%
Same sensor, same sensitivity, same session, and 150 times the leakage on the side where the wall is close. The room behind that near wall is a bedroom, so those were someone lying in bed being reported inside my living room.
If you are placing drops now, that argues for keeping a radar away from a shared wall where the layout allows, or aiming it along that wall rather than into it. Sensitivity did not fix it for me. I took it all the way down and the near-wall points were still there, but the geometry did.
Caveat: one apartment, one sensor model, Korean construction where interior walls are light partition rather than concrete. Your walls may attenuate very differently. Treat the ratio as the finding rather than the absolute numbers.
Since the walls are still open, I’d optimize for flexibility rather than trying to predict every future device.
A few things I’d strongly consider:
For sensors, I’d also avoid designing everything around the devices that exist today. Conduit and cable locations will probably outlive Zigbee, Thread, Matter hubs and whatever else changes over the next decade.
I made a vendor-neutral prewire planner around this exact “what should I do before the walls close?” problem. I’m the creator, so take it as an optional checklist rather than a product recommendation: