cancel
Showing results for 
Search instead for 
Did you mean: 

582n Wi-Fi Test Results

Graham21
Grafter
Posts: 72
Registered: 23-08-2012

582n Wi-Fi Test Results

Hi Everyone,
We received a couple of Technicolor 582n routers from customers to do some testing on.
I tested the two customer routers (Tijara33 and Hairy Biker) along with four random routers (PN2,PN3,PN5,PN7) that were hanging around the office. I also tested the ADSL router that I normally use at home (PN4).
I did this particular set of tests in my cellar. I used Inssider on my Macbook Air to determine that no other Wi-Fi networks were visible. It does not mean that there is no noise, but I think that this is the quietest place that I can find.
I used iPerf. iPerf is free software that you can download for your Mac or PC. You run iPerf as a server at one end of your connection and iPerf as a client at the other. In my case, I used my wife’s netbook (she wasn’t too happy about that) as the server and my Macbook  Air as the client.
I used this command on the server, a Windows XP machine...
iPerf.exe –s
The –s switch makes it run as a server with default parameters, listening on port 5001 and using a TCP window size of 8.00 KBytes
... and this on the client machine:
iperf -c 192.168.1.64 -t 24000  -i 200  -yc
The –c switch makes it run as a client, the IP is the IP of the machine running iPerf as a server, -t is the duration of the test (24000 seconds) and –i is the reporting interval (200 seconds). –yc makes iPerf output comma delimited reports that I could then import into Excel. I probably didn’t need to run the test for so long but as I did most of this overnight, I thought it better to run the test for as long as possible.
Test 1
To get a base line for the testing, I decided to first try with both the server and client connected to the 582n via Ethernet cables. The average result was about 95Mb/s.  The server I used has a 100Mb/s Ethernet port as does the 582n.

Test 2
I started with a Tijara33’s fibre router. I ran the test for 6.6 hours and then immediately ran it again for another 6.6 hours. I wanted to know that I could get the same result given that there are two baby monitors in the house, Christmas lights, mobile phones etc. The other Wi-Fi networks were turned off and as I said, there were no other neighbouring Wi-Fi networks visible to the client machine at any time during the testing.
I made a line graph of the 120 data points and was able to see that the range of results overlapped quite well I decided to continue testing with the other routers.
Test 3
I tested the other 5 routers. Three fibre and two DSL. The line chart started looking quite messy. To make it easier to see what was going on, I  created box plots. If you look at the box plots on the attached chart, the rectangular box shows the limits of the middle 50% of the results.  The horizontal line through the box is the median (middle) value. The ‘whiskers’ show the minimum and maximum value and the red dot is the mean (average).
If you look at Hairy Biker’s router, the minimum result I saw was 76.7 Mbits/second, the max 83.1 Mb/s, the mean was 82.0. 50% of the results were between 81.7 and 82.7 Mb/s. I thought that this was a pretty consistent result. 
I used a tile with lines on it to be able to line up each router in exactly the same way. I also had the Macbook sitting on a tile. I didn’t move it during the testing. I didn’t even move the lid as that’s where the wi-fi radios are.
PN6 and PN7 performed about as well as Hairy Biker and Tijara33’s routers respectively. I did find that three routers performed better but with the best fibre router (PN5) having an average result of 88.4Mb/s and Hair Biker’s coming in at 82.0 Mb/s you can see that there is not much of a difference. It would be interesting to repeat the test in an anechoic chamber with proper electrical isolation but as you might already be thinking, we are getting further and further away from the real world. I don’t think that either of the routers sent in have faults as they are performing well within the range of the test routers. There might be something going on in the home network environment and I would be interested in getting to the bottom of that. I am happy to test other people’s routers to see if we can find one that might have a fault so that we can send it to Technicolor for analysis. I am also interested in replacing people’s routers with known good ones to see if it makes a difference in their particular environment.
I would also be interested in knowing what router Tijara33 and Hairy Biker have been using since sending these ones back and whether or not they are getting better performance.
. I did test some competitors’ routers. I can tell you that none of them performed as well as the Technicolor 582n (the range of the means was 32.1 to 78.0 Mb/s). I would be interested in testing other 2 x 2 n routers so if you have suggestions of good ones, please let me know.
Graham
61 REPLIES
Community Veteran
Posts: 38,460
Thanks: 1,030
Fixes: 62
Registered: 15-06-2007

Re: 582n Wi-Fi Test Results

A question - did you try them with a few walls in between or were all the tests all in the cellar with all the parts in close proximity
Community Veteran
Posts: 5,172
Thanks: 480
Fixes: 20
Registered: 10-06-2010

Re: 582n Wi-Fi Test Results

Is it possible that a Macbook Air itself has better than average wifi?
Community Veteran
Posts: 6,773
Thanks: 257
Fixes: 20
Registered: 16-02-2009

Re: 582n Wi-Fi Test Results

Given that the router that I returned WOULDN'T HOLD a 128KB stream overnight, how long did you test for?
(Previously using a second hand speedtpouch which would keep running for months at a time on wifi)
If you are going to NAME routers you should say WHY they were returned.
Community Veteran
Posts: 5,172
Thanks: 480
Fixes: 20
Registered: 10-06-2010

Re: 582n Wi-Fi Test Results

All these tests have shown is that the TG582n is capable of good wifi throughput (presumably at short range) to a single Macbook Air connected, under artificially clear conditions. I don't think it'll give much insight about when it doesn't.
Community Veteran
Posts: 1,294
Thanks: 30
Fixes: 4
Registered: 22-06-2012

Re: 582n Wi-Fi Test Results

Hi Graham, I wondered what had happened to my router!! Now I know.
I read your results with interest. I returned the 582n at Chris's request because I was reporting poor wireless performance on the ground floor of my house, with the 582n upstairs in the study connected to my master socket via the BTOR modem. On a profile of 45 Mgs I was only achieving 20-25 wirelessly. The distance seemed to be causing the problem.
I repleaced it with an Asus  RT-N56U, purchased from Amazon. This is giving me 43-45 Mgs on the 5Ghz channel & 40-42 Mgs on the 2.4Ghz channel. It has been totally reliable & although not quite as easy to install as the 582n, has a host of really neat features, most of which I haven't a clue what to do with!! As an OAP I'm no techie but I did want as fast a system as possible.
Just for interest I live about 650 meters from my green cabinet on the Fordingbridge exchange. Prior to fibre becoming available last October I was achieving about 2.6 Mgs. Unlimited fibre is the best thing since fried bread!!! Cheesy
Community Gaffer
Community Gaffer
Posts: 5,269
Thanks: 577
Fixes: 5
Registered: 04-04-2007

Re: 582n Wi-Fi Test Results

Quote from: ejs
Is it possible that a Macbook Air itself has better than average wifi?

Good question, and worth a test.
Kelly Dorset
Broadband Service Manager
Community Gaffer
Community Gaffer
Posts: 5,269
Thanks: 577
Fixes: 5
Registered: 04-04-2007

Re: 582n Wi-Fi Test Results

Quote from: Hairy
Given that the router that I returned WOULDN'T HOLD a 128KB stream overnight, how long did you test for?
(Previously using a second hand speedtpouch which would keep running for months at a time on wifi)

6 hours or so each I believe.
Quote from: Hairy
If you are going to NAME routers you should say WHY they were returned.

Not sure what you are meaning?  We have quite a few new routers in the office, as well as returned ones + the ones you and Tijara33 sent to us to test.

Kelly Dorset
Broadband Service Manager
Community Gaffer
Community Gaffer
Posts: 5,269
Thanks: 577
Fixes: 5
Registered: 04-04-2007

Re: 582n Wi-Fi Test Results

Quote from: ejs
All these tests have shown is that the TG582n is capable of good wifi throughput (presumably at short range) to a single Macbook Air connected, under artificially clear conditions. I don't think it'll give much insight about when it doesn't.

Next step I reckon.  Try the test with the connecting device much further away Wink
Kelly Dorset
Broadband Service Manager
Graham21
Grafter
Posts: 72
Registered: 23-08-2012

Re: 582n Wi-Fi Test Results

Quote from: Oldjim
A question - did you try them with a few walls in between or were all the tests all in the cellar with all the parts in close proximity

Hi I was concerned that the post would be too long so I omitted some of the rationale for the test. I had previously tested all the same routers in my house with the routers left in the same position and with the laptop in different positions on different floors. I noticed pretty early on that the results were not consistent. The performance of each router seemed to vary as much or more than the differences between them. I could see quite of few of my neighbours networks. I decided that I would try to eliminate interference and the only place I could do this was in the cellar. Although the results of the test are Mb/s for each router tested, what I was really looking for was 1) were the returned routers any different from randomly selected ones and 2) were the Plusnet 582n's better or worse than other routers that I had access to. I was hoping to find that under the same conditions the router received from Hairy Biker or Tijara33 had significantly less throughput. Then I would have sent them to the manufacturer for further analysis. As it is now, I'm sure that they would not find a fault with these routers.


Graham21
Grafter
Posts: 72
Registered: 23-08-2012

Re: 582n Wi-Fi Test Results

Quote from: Hairy
Given that the router that I returned WOULDN'T HOLD a 128KB stream overnight, how long did you test for?
(Previously using a second hand speedtpouch which would keep running for months at a time on wifi)
If you are going to NAME routers you should say WHY they were returned.

Hi. I can confirm that I tested for 6.6 hours and recorded the average throughput every 200 seconds. When you say stream, can you be sure that the stream was persistent? If you send me the URL, and more details about the environment I could try to recreate. Have you been able to solve this problem using another router from us? Very keen to get to the bottom of this so I appreciate your help. I'l bring routers to your house if that's what it takes but right now I have no reason to believe that there is anything wrong with the router you sent in.

Graham
Graham21
Grafter
Posts: 72
Registered: 23-08-2012

Re: 582n Wi-Fi Test Results

Quote from: ejs
All these tests have shown is that the TG582n is capable of good wifi throughput (presumably at short range) to a single Macbook Air connected, under artificially clear conditions. I don't think it'll give much insight about when it doesn't.

Hi, you are partly correct. All I was trying to do was compare routers to each other under the same conditions. The problem with doing this under *real* conditions is that the conditions change and nullify the test. We would have to build a house (or at least some rooms) inside and isolated chamber. Then we would be debating whether the walls were the right thickness ; )
If we find a router in a given environment which is not performing as we think it should, we can replace the router with one that I can show works as expected and see if it makes a difference. If it does then we can have the router analysed by the vendor to find out why it does not work as expected.
Graham
Graham21
Grafter
Posts: 72
Registered: 23-08-2012

Re: 582n Wi-Fi Test Results

Quote from: Kelly
Quote from: ejs
Is it possible that a Macbook Air itself has better than average wifi?

Good question, and worth a test.

This is an interesting point. One of the routers that I tested was set to 40MHz (so it used two channels simultaneously). I realised too late and decided to run the test anyway to see what would happen. It made no difference because my Mac doesn't support 40MHz in 2.4GHz N mode. So I could have gotten better performance if I had been using a different kind of laptop.
Again, I wasn't trying to say that the 582n can deliver 90Mb/s in every situation, I was just trying to compare various test units in a reproducible way.
If I was to do further testing, I would actually use a USB Wi-Fi adapter so that I could fix it into position and not have to worry about keeping the test machine in exactly the same position for long periods of time.
Graham
Graham21
Grafter
Posts: 72
Registered: 23-08-2012

Re: 582n Wi-Fi Test Results

Quote from: tijara33
Hi Graham, I wondered what had happened to my router!! Now I know.
I read your results with interest. I returned the 582n at Chris's request because I was reporting poor wireless performance on the ground floor of my house, with the 582n upstairs in the study connected to my master socket via the BTOR modem. On a profile of 45 Mgs I was only achieving 20-25 wirelessly. The distance seemed to be causing the problem.
I repleaced it with an Asus  RT-N56U, purchased from Amazon. This is giving me 43-45 Mgs on the 5Ghz channel & 40-42 Mgs on the 2.4Ghz channel. It has been totally reliable & although not quite as easy to install as the 582n, has a host of really neat features, most of which I haven't a clue what to do with!! As an OAP I'm no techie but I did want as fast a system as possible.
Just for interest I live about 650 meters from my green cabinet on the Fordingbridge exchange. Prior to fibre becoming available last October I was achieving about 2.6 Mgs. Unlimited fibre is the best thing since fried bread!!! Cheesy

Looks like an interesting router. It has 5 antennas where the 582n has two. It might also be set for bonding, can you check? Your speed increase is essentially double what you were getting before which makes me think it might be set for bonding whereas the 582n is not by default.
G.
Dodgexander
Grafter
Posts: 37
Registered: 14-01-2013

Re: 582n Wi-Fi Test Results

Quote from: Graham21

Looks like an interesting router. It has 5 antennas where the 582n has two. It might also be set for bonding, can you check? Your speed increase is essentially double what you were getting before which makes me think it might be set for bonding whereas the 582n is not by default.
G.

By default? How can we turn on channel bonding?