This Wi-Spy device is a lot of fun to monkey with ... I think it really appeals to the not often used physics part of my brain (recall my degrees are in physics not I.T.)
Saturday I spent a couple minutes in efforts to quantify why my wifi would drop at the house anytime our old 2.4GHz phone was in use. First I tested our new since Christmas 5.8GHz Uniden phones ... Uniden claims they have absolutely zero wifi interference. I can qualify that since I've had no wifi issues since Christmas. Joshua Wells left a comment on the my last Wi-Spy post about some 5.8 phone systems still using 2.4GHz for handset-to-handset communication while base-to-handset is 5.8GHz.
So I fired up the Wi-Spy about 4 feet from the phone base and turned on a handset ... nothing nada zippo. So then I used my handset to connect directly to the handset in the kids room (this "room monitor" function is great to eavesdrop on the kids at night) ... again no signals showing up.
Now here's one drawback to the Wi-Spy - it only does 2.4GHz and not 802.11a (5.8GHz). Of course .11a isn't as prone to interference problems like .11b/g as .11a has many channels available and doesn't have the terrible channel bleed of .11b/g
Anyways, back to the experiment. So I dug out the old 2.4Ghz phone system, fired up Wi-Spy and then hit TALK. Wow, check out this graph of a 30s capture ...
Compare that to the phone off ...
Good grief that is one noisy phone! No wonder it knocked me off wireless all the time! I did try different channels on the AP long ago, but they never seemed to make a difference. So why wouldn't putting my AP on ch11 not have helped before? According to the graph, there's little noise up around 11 ... I was stumped for the moment. Man, I could have swore I tried using ch11 before and it didn't help. Hmm
So I moved my AP over to ch11, started streaming some tunes from Yahoo Music and then hit 'talk' on the old phone ... audio kept on streaming. Grrr. Hit 'talk' again ... still no problems. Arrrrgh ... you gotta be kiddin' me! For kicks I hit 'talk' one final time ... bam, wifi dropped. What changed?
I looked at the graph and there was a new curve present ... almost the same shape as the prior, but the peak was shifted up towards channel 3/4. Ah hah! I'd forgotten that this old 2.4GHz phone had several channels it could use. Not only is there is channel button on the handset to force channel hopping, but apparently the handset hops every so often on it's own.
So here's a graph created by hitting the channel button on the phone every 15s
The upper graph shows the signal peak starting below ch1, then hops over to ch7, then down to 3/4, then ch10/11, then far right of ch11. Further testing showed those same signal peaks, but the order of the hops was different. I'll also note that the only "phone channel" that didn't kill my wifi connection (when set to ch11) was the peak below ch1. So in the end it didn't matter where I would have set my AP ... eventually the phone would hop and wipe out any wifi connection.
Fun stuff ... boy, this would make a kewl science fair project :-) You can bet I would have used this for some lessons in my physics classes.
I did end up moving my AP from channel 6 to 1 after looking at the prior graphs from our microwave (aka "the sterilizer"). At channel 6 it wasn't getting disconnected but the signal dropped significantly ... at ch1 it's affected the least.
Comments