Need a good WiFi extender

[Mike Henderson]

Is your shop a separate building?

No, but the house is old and has plaster walls, which means there's a metal mesh behind the plaster which greatly attenuates the WiFi signal. I do have a sub panel in the shop but I tested the HomePlug devices I have and they provide connectivity to the outlets in the shop. The HomePlug units I have are the older 200Mbs units (the new ones are much faster, and the 200Mbps doesn't really give you 200Mbps) but that's fast enough for what I want to do in the shop.

Mike

[Curt Harms]

<snip>
My point was that there are a lot of unsupported routers and its definitely not a safe assumption that you can load a router with DD-WRT. Even some versions of the WRT54G can't be loaded.

Right. Chipset, amount of NVRAM and SDRAM in the router all enter into it. I know Linksys went away from using Linux to using VxWorks as an O.S. on at least some of their newer routers. I believe VxWorks has lower hardware requirements. It's also not open source so difficult or impossible to add functionality which may be seen as in Cisco's best interests.

[Mike Henderson]

I bought another HomePlug unit and got it installed on my HomePlug network. The signal is strong in the shop (compared to at my TV). Then I bought an inexpensive 802.11n router and hooked it to the HomePlug unit. Works great. It's a different network name compared to the one in the rest of the house, but my phone automatically connects to it when I go into the shop. I ran a speed test on it and the HomePlug is a lot faster than my DSL connection.

So, all in all, I'm satisfied.

Mike

[Matt Meiser]

I ran a speed test on it and the HomePlug is a lot faster than my DSL connection.

Wait, what? Are you saying the network speed from point A within your house to point B within your house is a lot faster than your DSL?

[glenn bradley]

Wait, what? Are you saying the network speed from point A within your house to point B within your house is a lot faster than your DSL?

It certainly should be. Even at 10Mbps, most short-haul ethernet devices would exceed your DSL speeds when simply running between each other ;-) Or did I misunderstand?

[Matt Meiser]

That's what I'm thinking he meant? Because it shouldn't test faster than the DSL line on an Internet speed test.

[Mike Henderson]

Wait, what? Are you saying the network speed from point A within your house to point B within your house is a lot faster than your DSL?

Yep, the HomePlug is about 60-70Mbps, while my DSL connection is maybe 16Mbps.

WiFi is a lot faster than my (or almost anyone's) DSL connection.

One way to know if your internal connection is a lot faster than your DSL connection is to run a DSL test with your computer connected to your DSL modem. This will give you the real speed of your DSL connection. Then run a speed test across your network. If your network is not a LOT faster than your DSL connection, the speed test will give you a slower speed than you got in the first test. This test is more sensitive to contention networks (such as multi-drop Ethernet or WiFi) than if you have a point-to-point Ethernet connection.

Mike

[Matt Meiser]

What are you using to do a LAN speed test?

[Mike Henderson]

What are you using to do a LAN speed test?

I didn't run a specific LAN speed test. See my post above for my reasoning about the LAN speed.

Mike

[Dan Hintz]

I'm confused about your testing methodology... if I'm reading that right, I'm not so sure you're testing what you think your testing...

[Mike Henderson]

I'm confused about your testing methodology... if I'm reading that right, I'm not so sure you're testing what you think your testing...

Yeah, you're probably right. I was thinking that with a contention based link, such as WiFi or HomePlug, you would have times when you would be delayed in sending a packet, and that delay would result in some "missed" bits on the slow portion of the link (the DSL link). However, in doing a speed test, the data is buffered before it is presented to the slow portion of the link. So unless the delay in the contention portion is significant, the slow portion will just continue to send data from the buffers.

So my thinking was wrong. To see a slower speed test, the in house link would have to be pretty slow compared to the DSL link, and that's really unlikely.

Mike