Poor WiFi -- Lintels/steel beams

nyck

Active Member
We added a 2 storey side extension to our 3 bed house using Heavy duty lintels for the Arch/door ways from old house to extension about 10-15 years ago

The Wifi was never great so about 3 years ago i added 2x TP link eap 225 v2 , one in the downstairs hall way one in the upstairs landing ( both in the original house ) all good.
We've had a few issues with Virgin over the years ,damaged cables , upgrades , lock down over load etc . So when the kids have been moaning about slow internet ive always put it down to Virgin .
We've recently been redecorating so been using the rooms in the extension side more and noticed issues browsing/streaming etc via wifi .
ie our kitchen is split , half in the extension half in the original house .Using Speed test app via my phone In the original part i get the full 200Mbps , move 3-4ft right into the extension part it drops on average to 20-30Mbps , sometimes its upto 150Mbps but not often .
Anyone else have this issue ???
I can add extra AP's but it seems a bit excessive having 2 AP's in therory on the same upstairs landing 6 ft apart
 

mickevh

Distinguished Member
Wi-Fi hates anything wet or metal. It works best with direct, unobstructed line of sight between communicating peers - more so in the 5GHz waveband which penetrates "stuff" (walls, doors, air) less well. Steel lattice structures ("frames") can also effect a Faraday cage which is also not so good.

Internet "speed tests" are not a very good way to test local Wi-Fi performance - using such you don't know whether the performance observed is attributable to your local Wi-Fi or the ISP link. To test local Wi-Fi performance, it's better to use apps like NetIO or iPerf (both cross platform I believe) which takes the ISP link out of the test path. It's pretty much like hosting the "speedtest" server locally, ideally on something wired to your router. Both are free.

Also, when testing it's best to ensure you eliminate as much air time competition as possible - turn everything Wi-Fi off except the AP---client pairing you are using to test.

Rather than guessing or buying new kit, you could experiment with what you have in hand: Unplug your router from the Internet and move it around and see how the Wi-Fi link rates vary. It doesn't matter that there's no Internet connection, but you won't be able to run a "speedtest" (local or Internet) but you will be able to see how the link rates change on client devices as you experiment with positioning. Don't worry about the absolute numbers, you are looking for trend. Of course, with no Internet connection you will need to "warn the household" whilst testing. You could also do the same with one of your AP's, but I am presuming they are less convenient to temporarily demount.

Thence, hopefully, you'll have some evidence of whether more/moved AP's will make a difference which will inform a decision on whether you need to buy more AP's or get better result moving what you have.

With multiple hotspots, you may also be suffering with co-channel interference issues if you haven't paid attention to the channel plan. (Basically, each AP/router need to be tuned to a different radio channel.) If not, that can rob throughput (which is what "speedtest" measures.) We can get into RF channel planning if you like.
 
Last edited:

neilball

Well-known Member
You extension is on the other side of what was once an external wall, so it’s that wall structure that will be affecting the wifi performance - dense brick/block/insulation blocks the high frequency radio signals from your APs compared to the internal walls in the original house. You will find that wifi performance is best where your position lines up a straight line from an AP through a doorway into the extension, and be worst where the radio path passes at an angle through the original external wall (as opposed to at right angles to the wall).

So if you want better performance in the extension you may well need additional APs, with the channels set to avoid interference with each other.
 

captain morgan

Active Member
I’m presuming you have the eap225’ so hard wired not in a mesh config?

I too use the eap225 I’m in a three bed with a gf extension and open plan gf so I have three sizeable steels there.

My ap is mounted in the loft in a central position and I get good coverage throughout the house, extension and garden.

The wifi signal generally has a easier time penetrating floors than walls so by mounting in the loft it allows the beam to form before hitting significant obstructions.

Admittedly I only get ~60mb/s from my fttc connection but that is reliable in any location.

It might be worth relocating one to your loft and see if it improves the situation

It could also be that devices are not roaming correctly and hanging onto ap’s at grater distance
 

nyck

Active Member
I only know a little about WiFi and thought i was in a mesh network
All AP's are hard-wired so I guess its not :confused: .

I might make up a long Cat5 lead and move the AP around to find the best position ( I thought I did originally ) . Upstairs I could possibly mount it under the metal beam so the AP is firing into the original and extension at the same time ?? or bite the bullet and buy some extra AP's .
The upstairs is going to be easy as the modem/router is upstairs and everything is fed up into the loft and then back down to each room , the downstairs might be a bight more tricky without lifting multiple floorboards and carpets

I forgot I stuck a extra AP in the conservatory to cover that and the garden , can you have to many AP's ??

Thanks for all the Info/advice
 

captain morgan

Active Member
Yeah you are confusing wifi roaming with mesh not the same thing at all.

I’m managing my system with the tp link Omada controller so I’m not sure if all of these options are available to you as I assume as you have the v2 ap’s you are not using Omada management but.

Not sure if this a option with non Omada setups but turn on fast roaming, this can help force clients to move from ap to ap more seamlessly.

I have the option of looking at what clients are attached to my ap so you should be able to see the same when a device had poor performance what ap are they connected to is it the closest

I have the option of doing a wifi scan from the ap’s which will guide you what bands are most congested in your area.

Ideally you’ll place each ap onto its own pair of channels 2.4 & 5 ghz

The scans take about 5 mins per band per ap.

Once you have them split if you still have issues then perhaps look at lowering the tx power, it seems counter intuitive but if all the ap’s are screaming over each other it can be hard for the client to work out which ap to attach too.

If that doesn’t change it then revert to max and it’s time to look at rssi thresholds this forces weak client connections this is a bit of a art but here’s a very basic overview.

Ideally if a change hasn’t improved the situation then you should back it out and try another apart from the manual channel allocation, though you may have to do this more than once as the neighbouring routers might auto adjust to use space you created.

If none of them solve it then try combining lower power with fast roaming etc.

Good luck!
 

nyck

Active Member
RE - Omada management
Do you have to have it running live on a PC 24/7
My PC is left on most of the time but I do turn it off now and then

I could possibly stick and old PC in the loft or my data cabinet
 

captain morgan

Active Member
Yes but only because I have a low powered device serving other sw.

I believe that fast roaming is the only one of those settings that need the controller to be permanently on.

I believe that you need v3 ap’s to use the sw controller but might be wrong.

They do a hw controller for ~£60 that might be better than having a additional pc on 24/7 but I’d test with the sw controller before spending on one.
 

nyck

Active Member
Update
Upstairs I moved the AP direct under the lintel so i guess its now firing into the extenstion and orginal house. My sons bedroom is in the extension has said its dramatically improved and the signal strengh seems ok in the rest of the upstairs .
just need to sort out down stairs , ill try a few of the above suggestions

thanks
 

jimscreechy

Active Member
Yeah you are confusing wifi roaming with mesh not the same thing at all.
Actually Mesh isn't a technically defined term at all, it is more of a marketing one that manufacturers seemed to have employed. At present there industry agreement in what it's meaning is supposed to specifically constitute, and Its loosely defined interpretation and can be (and often is) morphed to fit all sorts of descriptions to suit promotion.
 

captain morgan

Active Member
Actually Mesh isn't a technically defined term at all, it is more of a marketing one that manufacturers seemed to have employed. At present there industry agreement in what it's meaning is supposed to specifically constitute, and Its loosely defined interpretation and can be (and often is) morphed to fit all sorts of descriptions to suit promotion.



These seem to be good and understood definitions of mesh networking.

Obviously that does not preclude companies marketing teams (mis)using the term.
 

jimscreechy

Active Member
Just to be clear, I didn't say there are no definitions of Mesh, I'm speaking of technically defined terms which companies and bodies have a agreed understanding.
 

mickevh

Distinguished Member
Also, the term "mesh" is not particularly a "Wi-Fi" thing - mesh as a network topology has been around forever - long before Wi-Fi (as we know it at time of writing) got invented. The take home message when making purchasing and/or deployment decisions about Wi-Fi equipment is to not "assume" that any given product has any particular capability just because it claims to be a "mesh" system - I suggest it is best to dig into the datasheets, specification and/or (ideally) manuals to see what we are getting for our money.
 
Last edited:

The latest video from AVForums

Paramount + UK launch: Halo, Star Trek and Beavis, and all the latest 4K + Movie/TV News
Subscribe to our YouTube channel

Full fat HDMI teeshirts

Support AVForums with Patreon

Top Bottom