Large historical home - wifi quote

[HCHTech]

We had our first onsite meeting today with a new customer - or rather with that customer's vendors. Architect, project manager & electrician. This will be the largest Unifi setup I've done.

It is an 1865 brick home, 3 stories plus basement, plus separate 2-story garage with basement, plus outdoor spaces on each side of the home. They are 6 months into a 2-year renovation of the entire space. I can't even imagine what the pricetag must be.

They dug down 2 feet and remade the entire foundation to support 8-foot ceilings in the basement. Stone foundation, and stone interior walls on that level separating a total of 5 rooms that need wifi plus mechanicals space. There will be a full network rack in the mechanicals room.

The first and second floors have interior walls made from bricks, all 9" thick, being faced with studs & drywall. 5 rooms that need wifi on the first floor, 4 rooms that need wifi on the second floor. 15' ceilings on the first floor, 12' ceilings on the second floor.

The third floor is normal wooden studs & drywall, 5 bedrooms that each need wifi. 9' ceilings on the third floor.

The garage needs wifi in the basement, where there will be a shop of some kind, in the main car space for Tesla's OTA stuff, and the 2nd floor will be a greenhouse. That will probably get signal from the exterior AP on that side of the house.

Some fun restrictions from the architect: No ceiling-mount APs, which would "break the aesthetic" and be difficult to service once in place (he's not wrong about that with those high ceilings).

I took an access point I could setup and then measure the signal as I walked around, and it is quite clear that wifi doesn't go through stone or brick (no surprise). I'm going to quote Unifi In-wall units - probably 19 in the house & 2 in the garage, then 4 outdoor APs for the exterior spaces.

That totals up to a 318W PoE budget, so I'm planning on using 2 of the Unifi 48-port Layer2 switches - they each have 32 PoE ports and a 195W budget. This will give us 75 ports leftover for the multitude of ethernet runs over and above the APs.

Suggestions welcome - I do have some concerns about the length of time likely involved in this project, and how that will impact availability & pricing, etc. Should be fun, though!

 

[YeOldeStonecat]

So if the saucer ceiling mounts aren't allowed...yeah the In Wall, or the FlexHD/Mesh6...they can be wall mounted...or nicely stood up on furniture like wall units, or shelving. And they will give better performance than the InWalls.

For better penetration dial the 5 radios from 40 to 20 wide channels...

And as you did, bring a pre programmed AP (or a couple of APs)....with POE injectors, and plug in in various locations. Doesn't need to be connected to internet, just...plug in...walk around with a laptop or handheld and measure signal strength. See where things fall below say, -65 or -60. Talk to owners about where HD TVs will be...you want much better signals there...better yet..ethernet those if you can.

Re: In Walls...remember they are generally "low" in the wall...where you prefer higher. And often they end up having things "in front of them". Couches, wall units, whatever..which of course really puts a hit on their signal.

 

[HCHTech]

For the many rooms with TVs, I like putting the InWall behind the TV. That way the TV can be wired and the InWall will still give wifi to the space. I did talk about the Mesh6 (I wish they would pick a name and stick with it), but the architect clearly didn't prefer that look. I don't have any idea if there will be any built-in bookcase space where the Mesh6 could be more hidden along with it's connection cable. The other folks in the meeting were rolling their eyes and told me later that the architect was one of the most inflexible folks they had worked with - haha.

Also, we do have some control over box placement since the whole place is just studs now. He was fine putting one at switch height in the kitchen, for example. I'll underline that fact in our cost estimate. One limitation we have is that we only have vague ideas where eventual furniture will be, so I'm sure there will be some adjusting to do once the owners actually move in.

 

[britechguy]

Amen @Computer Bloke!

I'm sitting at this very moment in my den, in a brick home, where my WiFi access point is upstairs in the bedroom nowhere near to close to me and is almost certainly traveling through brick to reach me. I have often been mystified by assertions that WiFi cannot travel through solid walls of brick, stone, as my own life experience has clearly demonstrated otherwise. Although the signal is not as strong, I can even catch the same WiFi signal in my basement, two floors below the WiFi access point, and where the basement is entirely underground.

@HCHTech, do you ever meet with the actual clients (as in the homeowners)? There really are times where wired is a much better choice than wireless and where it can be achieved with minimal to zero visible artifacts.

The architect sounds like a Frank Lloyd Wright personality type. He was known to come into the homes he'd designed years after the fact and throw out things he believed "did not belong" and somehow befouled his aesthetics. It's someone's HOME, for heaven sake, not a temple to your genius, regardless of how great your genius might be.

 

[YeOldeStonecat]

re: utilizing multiple floors, I find it better to try to avoid stacking APs in the same vertical column...I try to stagger them. The floor above...would have APs sorta...in between the APs on the floor below or above.

 

[HCHTech]

Well, I'm not a lab, so can only go by my own experience. The problem isn't that the signal can't go through brick (I think stone is pretty much a blocker, though), it's that the attenuation is great enough as to drastically affect performance. The worst material I've ever run into was concrete. I have a piping engineer client (think gas plants), whose home is poured concrete & steel. Beautiful and must have cost a fortune, but getting wifi useably throughout that house required putting an access point in every room. We could turn down the signal strength since we only needed to cover that one room with any particular access point, but that was the solution. Oh, log homes suck as well -haha.

Anyway, I very rarely DON'T deal directly with the homeowner. In this case, I was hired by the project manager, I don't even know the homeowner's name. They are out of state, I heard. So we want to give the best coverage we can the first time around. They are spending a ton of money and it just needs to work. In that set of circumstances, I'm going to overbuild...especially after today's results with an actual access point onsite.

 

[britechguy]

it's that the attenuation is great enough as to drastically affect performance.

Not arguing with you, at all, but it's what gets considered "drastic" and by whom that really matters.

I have the feeling I'm a lot more tolerant of "meh" performance than many people might be. Quick check at this moment and download speed is 44 Mbps. Some would consider that glacially slow.

 

[HCHTech]

but it's what gets considered "drastic" and by whom that really matters.

Oh, you're not wrong there. We have some folks that know all of the speedtest sites and call whenever the measure something they think is "too low". Ugh. Most folks, though don't know or care about speed. They just want Netflix in one room playing without stuttering while Prime video or Hulu is streaming on another TV without stuttering. Or two different kids playing fortnite (or whatever MMORPG) at the same time without lag. We get more way more calls about internet speed on the residential side than we do on the commercial side for sure.

 

[YeOldeStonecat]

Throw on smart queues/bandwidth management...and watch their complaints fly. "How come I only get 20 megs?!~ When I have 500 megs from Comcast?!"

...and then I give the speed about bandwidth management and how you don't want to allow one or two users to grab all the bandwidth they can (hog it)...and everyone else suffers. "Like a good traffic cop at a busy 4x way intersection!" I tell them!

 

[YeOldeStonecat]

Forget to type in above...I did twitch a bit when I saw the zip ties on the cable bundle...ugh, I want to zip tie the installers throat when I see that. Velcro! no zip ties!

 

[Blues]

I prefer zipties but you have to do them right, not too tight or too loose, and I accept they are throw away as when I need to remove them they must be cut and I have perfect tool for that.

 

[britechguy]

I have perfect tool for that. [cutting zip ties]

A pair of small wirecutters is ideal. Nip and you're done! I'm like you in that I consider zip ties disposable and whether I've used them versus velcro has depended on my own estimation of the probability of "need for revision" (addition of cables in particular) at a later point in time.

 

[Blues]

True there is a measure of estimating need for change and yes a small pair of flush cut wire cutters is perfect for it.

 

[Markverhyden]

I want to zip tie the installers throat when I see that. Velcro! no zip ties!

Especially when they're every 6 inches or even less in some cases. I had one today like that but at least they had velcro so it wasn't so time consuming. BICSI even says to use velcro - https://www.bicsi.org/standards/bicsi-standards/resources/standards-frequently-asked-questions

 

[YeOldeStonecat]

My Simply45 snips are great for zip ties, and I remove them on any/all new clients we get. However, over time, they're too narrow and hard, they're prone to "pinching"...which changes the shape of the twists inside. Not to mention, just a pure pain in the butt when it's time to change things/grow the network (which is...really quite common, esp with successful clients). While sure I know how to "snip" carefully, you have to assume someone less careful than you may work on the network at some point in time, why make it more prone to error.

Take inspiration from some of the "best of the best, high end cable installers" like Chris Tiffanys crew from SprinterIT. His crew does some incredible work, some huge scale work too. And always...velcro, never zip.

I'm not a tree hugger but I am mindful about cutting down on one time use plastic.

 

[Markverhyden]

The very real problem with zips is simple statistics. All things being equal the more often "something" happens the great the probability the outcome will be bad. As in not only snipping the zip but some cable as well.

 

[frase]

Wish you well, in my edperience people with $$ are a pain in the arse.
I have no real input rather than to sit back and enjoy the show, I personally use felt ties.

My annoyance onsite is someone who used dead zip plastiic zip ties grr..