Archive for November 2009

When is faster actually slower?

Sunday, November 1st, 2009

Speedtest.net is a nice way to make you feel less guilty about spending too much money on Verizon FiOS test the speed of your internet connection. I decided to test things out on my Mac Pro. I tested out two different setups:

Screen shot 2009-10-28 at 7.14.52 AM.png

Screen shot 2009-10-28 at 7.12.34 AM.png

As you can see, the second test was significantly faster. What’s the difference between the two?

The first test was run via my Apple Time Capsule, an 802.11n-based wireless base station. Not only was it running 802.11n, the fastest current WiFi standard, but it was running in dual-band, 5 Ghz mode.

The second test was run using the stock Actiontec 802.11g router that Verizon supplied.

The two routers are connected via ethernet and are only inches away from each other. While the Actiontec is one hop closer to the external network, it really shouldn’t make that much of a difference. The 802.11n router should be faster or, at worst, the same speed as the g router (if the external connection is the limiting factor). The best I can do to explain the fact that I’m seeing an 802.11n router perform worse than an 802.11g router is that the antenna in the Actiontec (which is external) is better than that on the Time Capsule.

I adjusted the Multicast Rate on the Time Capsule to the maximum, which in theory shouldn’t really affect the speed, but it did seem to make a difference.

However, the speed was still a bit worse than the g router.

More detective work to follow…