Wireless AP buying advice

[fizikz]

I'm looking for a wireless AP that will get the best coverage when placed near the center of the house. It needs to serve the main floor, a basement, and parts of an upper floor. It will be connected via Powerline AV to the router (Engenius ESR 9850) which will be placed in a corner of the basement near the modem.

The important factors are:
1) best wireless coverage
2) reliability/stability

I don't need many other features such as gigabit LAN, usb ports, 3 stream, etc.

What would be the best option for a set-it-and-forget-it device that will last a good long time before needing an upgrade? The Asus RT-N53 looks interesting. Any comments or alternatives?

 

[stevech]

any WiFi router can be reconfigured to be an AP.
There's a FAQ on that here.

Also, there are some products that plug into the wall and have IP over power line and a WiFi AP all in one. Don't know how well their WiFi antenna (internal) works. I'd be tempted to use one of these on an AC extension cord to elevate the WiFi antenna. I think I've seen these from Netgear and some of the commodity WiFi vendors.
Of course, there has to be a mating unit near the WiFi router or ethernet switch.

 

[fizikz]

I know I could use a router as an AP, and I would consider getting a router with good wireless performance. I just figured that if it's cheaper to get just an AP, I might as well, since that's all I need. The Asus RT-N53 I mentioned is a router.

I have read a little bit about some Powerline AV / Wireless AP units. I'm just not sure how well they perform as APs. They almost seem like an afterthought or a matter of convenience, but my main concern is about coverage and reliability.

The Ubiquiti UniFi looks interesting too, but it is a bit pricey. Its performance is apparently pretty good, and it can be powered over ethernet.

 

[stevech]

beware Ubiquity: Tech support via user forum only, I'm told by Ubiquity's distributor.

You could buy a WiFi+PowerLine IP all-in-one, elevate it via extension cord (say, on top of tall furniture or some such). See how it works.
Return if poopy - if you buy from a retailer, say, Fry's, Newegg. Maybe not BestBuy.

 

[fizikz]

I may end up buying from a retail store with the option to return since there's no way to know how it works in my case until I try. Still, I'd like to know what has a better chance of working well.

I thought Powerline units must be connected directly into the wall outlet since extension cords can degrade their performance. There doesn't seem to be much selection of Wifi+Powerline units, but I'll have a look around.

 

[stevech]

no, what degrades performance in power line IP are surge supressors and the like that are inside modern TVs, PCs, some plug strips, etc.

So power wiring devices should not share an outlet or plug strip with these.

plain old extension cord is benign. But don't plug it into one of the above.

Less so, noise on the lines from light dimmers.

 

[CaptainSTX]

Extension Cords

In my experience extension cords both the cheap flat cord types for lamps and the heavy duty, typically orange cords, drastically reduced throughput.

 

[fizikz]

Are you sure plug strips / surge protectors can't be plugged into the same outlet as a HomePlug AV unit (HomePlug not plugged into the surge protector)? It's pretty hard to avoid this, and also from what I've read and from my experience, plugging wall wart power adapters into surge protectors helps improve the throughput of the HomePlug units.

Other alternatives I thought of to the Asus RT-N53 and Ubiquiti UniFi is the TP-Link powerline range extender TL-WPA281, and the inexpensive TP-Link router TL-WR841N which has good reviews and apparently good range.

 

[stevech]

extension cords w/o surge suppressors are no different than the Romex in the wall.

 

[fizikz]

extension cords w/o surge suppressors are no different than the Romex in the wall.

That is what I would think, but maybe the gauge and quality of wire makes a difference?

In the end, I went with the TP-Link TL-WR841N as my AP (and convenient switch) since it is half the cost of the Asus RT-N53 and much less than the Ubiquiti, and has good reviews, particularly regarding wireless range. My setup is as follows:

Router (Engenius ESR9850) --> Powerline AV (TP-Link PA511) --> AP (TP-Link WR841N)

The WR841N has 5dBi antennas compared to the ESR9850's 2dBi, so I was curious about the range and signal quality. My initial finding is that the WR841N's wireless is at least as good as the ESR9850 and perhaps a bit better, but not by any huge amount. Nevertheless, now with two wireless units on opposite ends of the house, the coverage is very good.

I was having issues with the WR841N having wild swings in signal strength with my laptop in the same room, going from less than -40dB to -90dB. It was on auto channel selection mode. I found that only on channel 11 did those drops stop. Now it's quite stable at -30dB or better.

Currently I have both the router and AP using the same channel 11 since this keeps the signal most stable. They are both using the same SSID, although I may change that later. Some clients switch fairly well while others stick to the weaker one when moving around. Manually disconnecting and reconnecting usually prompts them to choose the stronger source.

 

[stevech]

Some clients switch fairly well while others stick to the weaker one when moving around. Manually disconnecting and reconnecting usually prompts them to choose the stronger source.

Yep. It's unpredictable.

 

[coxhaus]

You might try using the same SSID with different channels. The only channels to try are 1, 6, and 11. Mine work that way well enough to make my wife happy. Once in a while they are connected to the weak wireless unit but in my house they still function OK. I have not tried using the same channel as it seems like they would work against each other, but I am no expert on wireless.

 

[stevech]

same SSID, different channels, won't improve a WiFi client device that has a lazy strategy for "best-AP".

 

[coxhaus]

same SSID, different channels, won't improve a WiFi client device that has a lazy strategy for "best-AP".

You are correct but don't you think using different channels would be better than using the same channel?

 

[CaptainSTX]

You are correct but don't you think using different channels would be better than using the same channel?

All other things being equal yes.

 

[tipstir]

My experiences using the same SSID and same Channels are mixed.

If you stick with 802.11g or .11G

Example.

first floor
SSID: 11G
Channel 1

second floor
SSID: 11G
Channel: 1

third floor
SSID: 11G
Channel: 1

or

first floor
SSID: 11G
Channel 6

second floor
SSID: 11G
Channel: 6

third floor
SSID: 11G
Channel: 1


or

first floor
SSID: 11N
Channel 11

second floor
SSID: 11N
Channel: 11

third floor
SSID: 11N
Channel: 11


Depending on the surrounds and the amount Channels being used in your area you can check to see the distance or range that 1, 6, 11 channels are being used.

Using the same Channel makes it easier to connect also the same SSID unless you have 11G or 11N mix. Today I just use 1 High Power AP to control both 11G/11N.

 

[stevech]

Most WiFi clients don't use sensible logic on when to search for a better AP.

So using a unique SSID per AP allows a human with a brain to do best-AP selection.

Channels don't matter; SSID does.

 

[coxhaus]

I have to say I am no expert in wireless but in my house with wireless on opposite ends of my long house what seems to work best is the same SSID and different channels. My house is on a large lot in a neighborhood with large lots so I don’t have channel contention with the neighbors. If you are connected on one end of my house and walk to the other end the client will change wireless units and channels automatically if the signal gets weak. If you are in the middle of the house the client tends to hold on to the connected wireless unit. This is the reason I came to the conclusion, same SSID and a different channel.

 

[CaptainSTX]

Do the Laws of Physics Apply

For those of you running multiple APs and considering what channels they should utilize I would suggest you do your own tests.

First shut down all WiFi radios except one. Then run a LAN speed test writing and then reading 100 packets to an USB drive attached to your router or use a NAS if you have one. Do the test using WiFi with your PC in the same room as the AP, but 10 -15' away from AP

Repeat this test for each AP one at a time with it being the only AP activated AND being set to the same channel. Use the Same PC and move it to the same room as the AP you are testing.

Turn all your APs back on.

After you have all the statistics get another PC with WiFi. Connect the first PC to the first AP and the second PC to a different AP.

Start the LAN speed test simultaneously on both PCs which are each connected to a different AP and compare the results to your one at a time tests.

Then change the channels on your APs so they don't overlap and see what the impact is.

 

[stevech]

the behavior of the clients in selecting an AP will vary a lot in time, and by which one you use (laptop, phone, etc).

I would not draw conclusions based on a few tests.