WiFi spectrum readings :)

[wpte]

So, I got hold of a wi-spy 2.4i
The cheapest wi-spy available, tho the hardware would probably cost you just 10%

Anyway, since I've never seen measurements (couldn't find anything on the forums) from the wifi of asus routers I thought it would be nice to do a little investigation.

The thing with wi-spys is, that it can't scan all channels at once, so I left every test running for 10minutes since it seemed a quite clear result after 10min.

Also, I ran the test in a microwave. Why a microwave? well... I thought it would act as a nice Faraday cage since it's quite crowded with wifi here
Now I did test if any noise would be generated with a router inside, but I couldn't find any difference, only many other AP's where filtered away a lot.

The firmware is oleg r2249 (from google code)

First of all, A screenshot from inside the microwave without my WL-500W running:


WL-500W default settings:


WL-500W defaults, but with upper channel


WL-500W defaults, but with lower channel


WL-500W defaults, but only 20MHz (none channel)


It's clearly visible the WL-500W generates a lot of noise around channel 6
But more interesting is that it seems like it doesn't matter how you set things up... it will always use a 40MHz bandwidth

For the people who can't see it, an smaller shot with lines I drawed:

Between the red lines is 20MHz, the yellow lines are an extention to 40MHz so that the WL-500W can serve 270Mbits

What do you think, more tests?
if anyone has a better idea for test setup let me know

p.s. not to be negative about oleg firmware & wifi obviously, I haven't seen results from other firmwares. Also it might be a good chance for wifi tweaking

[theMIROn]

nice test! thank you
whats about rt-n models?

[wpte]

nice test! thank you
whats about rt-n models?

I've only did a quick test on my rt-n16.
also I should put some load on the AP next time, it's supposed to create a flat line around the central channel

[theMIROn]

So, awaiting

[wpte]

So, awaiting

Did some tests on the RT-N16, and found out I really have to put load on it to make the true signal visible

Yet, for the ppl who like to see the raw values:
rt-n16 default: http://i52.tinypic.com/2rzuyib.jpg
rt-n16 upper: http://i55.tinypic.com/2i052jc.jpg
rt-n16 lower: http://i55.tinypic.com/2mpztpz.jpg
rt-n16 20mhz: http://i55.tinypic.com/2yukaco.jpg


Now, I measured the following results by downloading a file from the usb stick

defaults with load (channel 11, lower 2nd channel):


lower, surprisingly it shows the higher channel is used.
central channel is 6:


higher, surprisingly it shows the lower channel is used.
central channel is 6:


20MHz


for the 20mhz band it looks like it's just adding a secondary 20mhz band on the same frequency range

[theMIROn]

Great!
We've already discussed that choosen channel is the primary (control) channel, and upper/lower options means that exactly primary channel will be upper/lower, but not secondary or central.
So, I'm thinking about to change SubChannel name to something like Primary, or exchange lower/upper values. But the second will confuse users, am I wrong?

BTW, can you test official RT-N16 FW 1.0.1.9, there're changes related to wifi power saving.
They could change 20Mhz freq usage map, I hope.