Here’s the juicy stuff, performance! I tested TEW-691GR’s wireless at 5 different distances:
|Distance||Number of Walls|
|80+ feet||Too many to count|
I used my laptop with two different wireless adapters. For all Wireless N transfers, I used the TRENDnet TEW-687GA 450Mbps Gaming Adapter. I connected the adapter to the laptop’s Gigabit Ethernet port with a Cat 5e cable, so theoretically it could achieve its 450Mbps max speed. I also tested Wireless G with my laptop’s built in adapter, Intel Pro Wireless 3945ABG. When I ran the Wireless N tests, the laptop’s built in wireless adapter was turned off.
I had some trouble with LAN Speed Test while testing the wireless. If I chose large data chunks, I’d commonly be greeted with, “Error writing to server”. To avoid this error, I settled on 5MB data chunks, being written/read 10 times. I repeated the test 5 times in each location with Wireless N & G. Then I congregated all of the data, and generated a bunch of stats including, average, median, standard deviation, max and min speeds. Below is the graph for averages, N vs. G.
As you can see, the comparison isn’t even close. Wireless N destroys G in every way, which is expected. It has better range, better throughput, and even though I don’t have a graph for it, better ping times. Wireless G failed at distances of 40 and 50 feet. It was able to establish a connection and obtain an IP, but once data started flying over the connection, the connection would drop. I was able to start a couple of simple Windows file transfer tests at 40 and 50 feet, each averaged about 30KBps before dying. Therefore, I couldn’t accurately measure Wireless G speeds at those distances.
You might be asking, where’s the 80+ feet category? Well, neither N or G connected at that distance. I think there were just too many walls for either signal to survive. I could avoid the walls by going outside, however, it’s been raining for days, so going outside wasn’t a valid option during my testing time.