Wired home - Mesh vs. AP?

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.

adambean

Distinguished
Mar 8, 2010
306
0
18,780
(tl;dr, mesh or traditional in a wire house?)

Hey all,

When I moved into my house, I made sure we ran ethernet to all rooms. It's a 2,400 sq foot home, minus the basement (which is ... 1,200 I believe).

I've had a few different setups through the years between dual routers, router + ap, router + repeater and just a really expensive router. Sadly, to date, I seem to one way or another, have issues.

Currently, I am running an Asus AC5300 and it covers most of the house and parts of the outside relatively well. As of late, I seem to continually have issues adding new devices (4 in the past week), in that they simply won't connect. This issue may unrelated all together, but, I do want/need better WiFi on 1st and 2nd floors. Recently started playing with baby monitors due to newborn, and solid WiFi is a must right now. I have issues getting a solid connection to 5ghz when on 2nd floor and outside. Streaming videos over WiFi is a hit or miss ... sometimes great, other times awful. This is why I largely rely on everything being wired.

Reality is, just ... frustrated with several failed routers through the years. They've all been Asus, perhaps that's my core issue. Or, I'm just stupid and missing something all together.

Anyway, I just started looking into these new mesh networks and they appear to be very appealing on paper. Quite a few competitors and right now appears Orbi is the best option. Yet, being that I do have ethernet run to pretty much every room, is this a dumb investment? Should I be looking at a different option?

I just pulled the trigger on the Netgear RBK50 + RBS50. It won't be here until next week as they're Warehouse Deals, so I have some time to cancel if this isn't the best choice. Being that I have less than 4k square foot, do I really need that third satellite? I figured, better safe than sorry ... or is that just dumb? Will try without the third on setup to determine if needed.

One last note, we are planning to move in a couple years. No idea if it'll be a new home or not and if the later, what options will exist to run lines. Hard to take that into consideration given the unknown, but part of me does say it leans a little towards the Orbi.

Any feedback is most welcome.

Thanks
 
@gregjar - Is there degradation per run? In that, what If I ran to 5 different rooms? The runs are pretty long too. Go out the basement, up a pipe, through the attic and back into the rooms. Not sure how to determine length, but I'd probably say ~50 feet on some of 'em.
 
If you're hooking the AC5300's antenna ports up to cable runs on a 1:1 basis without a splitter, 50 feet shouldn't be problematic. The degradation (attenuation) on 2.4 GHz over 50 feet of RG-6 cable is around 7.9 dB. To compare that to wireless alone (looking at SmallNetBuilder's review of the AC5300), your router can lose about 60 dB at 2.4 GHz before the RSSI is too low to support data traffic. At 5.8 GHz, there's about 13.6 dB loss over coax, while the router can lose about 39 dB before conking out.
 
gregjar - should I do this: http://www.ebay.com/itm/Coaxifi-Whole-Home-Wi-Fi-over-Coax-Router-Range-Extender-6-Antenna-Kit-/182655840438? or just do the straight shot? In reviewing, I'm thinking I could/should do 4 runs (half the router antennas) and that gets me one to first floor and three to second floor. I'd just like to try this before I commit to the AP.

Also, this review doesn't seem overly promising: https://www.digitaltrends.com/networking-and-wi-fi-reviews/coaxifi-wi-fi-over-coax-review/.

So I guess I have to ask (dollars aside here), is this Coax option 'better' than an AP? Seems like everyone votes for an AP over a whole new Mesh system, but now we're adding this new player to the mix. Seems interesting, but also odd that it's not widely available or talked about.
 

If you don't need 6 outlets and/or don't have a shortage of antenna ports on your router, I'd recommend using the 4-way kit since the insertion loss should be a bit less. With the AC5300, it may be that 4 of the antennas do 5 GHz and 4 do both 2.4 and 5 GHz (I believe it's the front of the router that has the interleaved 2.4/5 GHz antennas). You could always try out the StraightShot kit first.

Thanks for the link to the Digital Trends review. A few things look very weird - sending just 1 MB at a time is obviously not a legit throughput test, there seems to be a lightning arrestor or attenuator between the router and their cable that isn't in the Coaxifi kits, and never in my life have I seen a cableco leave coax runs dangling from the side of a house without at least mounting an enclosure. His throughput results for the WRT1900ACS routers are also suspiciously lower (about 5 times lower) than either Tom's Guide or SmallNetBuilder in every review, and always in the reviews where he compares it to the Ignition Labs Portal mesh. Some review sites have a pay-for-play model (BGR has an affiliate program, etc.), so you'd have to ask Simon Cohen if Ignition Labs has their fingers on the scales, or if he just doesn't know how to throughput test with two APs in bridge mode.

Subway and train engineers use 'leaky' coax for Wi-Fi and cellular in tunnels. Cell phone companies use coax for DAS in urban small-cell deployments. It's more of an engineering solution than a consumer electronics solution so far, since Silicon Valley isn't big on tech that doesn't have planned obsolescence built in. Actiontec and Teleste are two 'topside' brands that sells Wi-Fi over coax products, but AFAIK both of their options clone the SSID so you wouldn't have Fast Roaming. Dual-Comm has a more comparable Wi-Fi over coax solution, but it costs $199, is limited to 2.4 GHz, and is geared for people needing to diplex cable modem or TV signals with Wi-Fi on the same cable.
 


I'd use the StraightShots for 1:1 with no split required. The splitter kit would be more useful for routers with only 1 or 2 antenna ports, or if you wanted to combine a bunch of outdoor antennas on one router/network adapter.
 
So according to this article as well (https://electronics.stackexchange.com/questions/261558/), seems like you have to have very specific cables in play?

I guess I'm having a hard time signing off on the idea when it generally isn't well reviewed, options are very limited (an eBay seller with low reviews), and doesn't seem to be widely accepted.

In your opinion (or others here), is this 'better' than an access point or merely a cheaper alternative that may or may not yield positive results?

Thanks
 


We get what you pay for with free online advice...
If you use the calculator at http://www.timesmicrowave.com/calculator/?productId=118&frequency=2400&runLength=50&mode=calculate#form, you can see that 2.4 GHz over 50 feet of RG-6 is estimated to lose only 7.9 dB. With a router's output power and antenna gain, that loss would be even less. I think people are confusing how much a cable is sweep-tested for with the notion of some sort of frequency cut-off, as if the cable had a stop-band filter. The Yagi antenna in my yard wouldn't work if the online peanut gallery were right, nor would cellular DAS or MoCA or DOCSIS 3.1 cable-modem Internet (all established technologies, natch) as those readily exceed 1 GHz.

In my opinion, in a purpose-built home or office, you'd ideally use wired POE access points with 802.11/k/r/v Fast Roaming. But if the UniFi wasn't working well for you, then the challenge is in finding a set of extenders with the holy grail of high backhaul throughput and Fast Roaming. Maybe in a few months or years, these will exist and cost less than $100-200 per AP. In the meantime, the most efficient path to the second floor is through the walls. Given that inside every dipole Wi-Fi antenna is a thin cable (usually RG178 or weaker), putting Wi-Fi on RG-6 cable is actually sort of an upgrade.
 
Gotcha. Sadly I don't understand everything on this arena quite yet, but ... you made mention that the PoE AP's are a bit more optimal. Fast Roaming and Zero Handoff probably not overly necessary in a home, but would require not using my AC5300 broadcast. Currently using both, it seems like a first come, first serve. If I reboot one, they all go to the other. Establishing new connections is a bit of a gamble.

In regards to your statement about the high backhual throughput and fast roaming, if the Orbi enables ethernet backhaul, wouldn't that be one of the better/easier solutions to that problem?

Reality is, the AP is ... serving the purpose. It got me what I needed (a much stronger WiFi signal for the nursery to use a 1080p baby monitor). But now that I've gone through this, I'm in debate about what's the "best" option. Best in regards to strength and reliability, but also for resale. We'll only be in this house a couple more years and the UniFi AP is quite cool, but it's also a lot more overhead and administration, especially for someone that doesn't know that world well. As it is, I know enough to be dangerous.

The AP is cool for me, but simply having it already makes me want more. I need the SG to be able to do anything via mobile and gain deeper insight on the data. I hate that I can't just manage it anywhere on my network (desktop or mobile) vs. having to rely on the controller software (unless I'm missing something). It seems like it makes odd choices on what channels to put devices on (I'm hoping this thing is smarter than me). Then it drives the question of do I mount another AP and disable the router?

Your solution, using the coax antennas, is all things considered ... cheap and easy. Just doesn't seem to be a popular option and a handful of places seem to poop all over it. The concept seems legit, just wish I could buy them on Amazon so I could easily return 🙂. Still, $40 ... not the worst gamble if it doesn't pan out.
 
Well, I ordered 4 of the coax antennas + the line tester (gotta figure out which run is which), for $40. Not too bad. Definitely cheaper than the $97 I spent on the UniFi AP. Guy said I could return if it doesn't pan out, so either way, I'll stick with one of these solutions I think.

Toying with the idea of trying out one of the other Mesh systems ... both Google and Linksys offer Ethernet backhaul. Still being stupid and not fully understanding pro's and con's of each, I just feel like having all those nodes would yield better results in the big picture. Still, probably just stupid and overpriced being that I have the ability to do whatever with all the cables I ran.

Thanks
 
So did some testing with the antennas and could use some help here validating the data and understanding of this router. I ordered a secondary router to convert to an AP + new UniFi AP. I'll test each appropriately, but the Coaxfi is all things considered the cheapest and somewhat easiest. Before I commit to any of the three options, would love some feedback from the below rant.

Looking at the router right side up (ASUS is shown the proper way), the antenna numbering starts in the upper left at #1, ending at #8 on the side upper left.

Antenna #2 is a must, only way to have a broadcast happen. In each of the following tests, #2 was used in conjunction to each of the numbered antennas. Router is in the basement, and I walked roughly 20 feet away to conduct each test at the same location every time. Summary of what appears to be each antenna purpose + Test runs below.

5.1 vs 5.2 is each of the two 5ghz channels that this router broadcasts.

Three questions come from this data.
1) Being that the data seems quite odd for a couple antennas (primarily #3), is it safe to assume that the antennas are made to be paired with one another in some fashion?

2) Given the nature of this crazy router, 8 antennas, all serving different purposes with three different broadcasts ... is Coaxfi a smart solution?

3) If #2 is a YES, how do I know which antennas to convert and why? In theory, using these new antennas, won't I hurt my range somehow in regards to #1 potentially?

Again, thanks for all your help and sorry for all the pestering. You've helped me to understand more than I did before. I hope that ultimately I pick your solution, but as of now, I'm simply uncertain.

Thanks!

My interpretation of the data below:
#1 = 5.2
#3 = No clue, this is odd
#4 = 2.4 + 5.2
#5 = All
#6 = All
#7 = 5.1
#8 = 5.2

Tests:
#2 (stand-alone)
2.4 = 63-70
5.1 = 82-84
5.2 = 64-67

#1
2.4 = 61-67
5.1 = 83-85
5.2 = 54-57

#3
2.4 = 63-68
5.1 = 83-85
5.2 = 63-66

#4
2.4 = 49-53
5.1 = 83-85
5.2 = 56-57

#5
2.4 = 58-62
5.1 = 67-69
5.2 = 64-65

#6
2.4 = 50-54
5.1 = 69-75
5.2 = 60-62

#7
2.4 = 70-71
5.1 = 62-64
5.2 = 62-64

#8
2.4 = 67-70
5.1 = 81-82
5.2 = 57-58