Router Advice

[tomliv]

I'm looking at our first home modem and router (have been renting a Comcast combo modem/router for years). I've been doing a lot of research over the last couple weeks, read many reviews, and have spent a fair amount of time reading reviews/tests on SNB. I thought I would now ask some advice in this forum.

Our house is two levels, 100+ years old with plaster and lathe walls, about 1200 sq. ft. Neighbors 30-40 ft away on both sides. Comcast is our provider, and we have 75 Mbps download/5 Mbps upload service. I think our electronics using the Internet are pretty typical:

• Three personal computers (one Macbook, one HP laptop, and a desktop). I believe that all have 800.11n wireless cards
• One work computer, 800.11n wireless card. I use a work VPN to connect to work from home several times a week.
• HP wireless printer
• 20-year old has a couple game systems - Playstation 4, XBox
• Roku 2, Kindle Fire, and several cell phones

I did buy an Arris SB6183 modem thinking that Comcast was going to end of life the retail SB6141. After talking with Comcast, I don't think that is going to be the case, so I plan on returning the 6183 and picking up the 6141 to save some money. I also don't believe that we are benefiting from the 6183 over the 6141.

I also picked up the TP-Link Archer C7 router and have been using it for the last several days. From what I read, it was the best bang for the buck. The router is in the corner of the living room (first floor) because that is where the cable comes in. At this time, there are no plans to relocate the router.

With the Archer C7, I have measured wireless signal strength with Acrylic Wi-Fi Home and wireless speed with speedof.me on my personal laptop (800.11n wireless card). Signal strength is about -40 dB in the living room and -55 dB in one of the upstairs rooms and -65 dB in the 20 year old's room. Wireless speed tests (wireless computer connected to network) show downloads varying from 4.5 to 6.7 Mbps download/2.1 to 2.8 Mbps upload (living room), 5.5 to 8.4 Mbps download/1.6 to 2.75 Mbps upload (upstairs room), and 2.75 to 6.1 Mbps download/2.0 to 3.15 Mbps upload (20 year old's room). The Comcast router was active at the same time but typically on a different 2.4 GHz channel.

The Comcast gear had slightly better signal strength in all the rooms but the downloads were roughly half of the Archer C7's.

I did pick up the Archer C8 from BJ's to try out. It's wireless signal strength and download/upload speeds weren't noticeably different than the C7, and my wife was not a fan of the white packaging. So, it's going back.

I can afford a more expensive router. The C7 was $91; I would pay up to 2x that if the performance upgrade, reliability, etc were noticeable and worth the price difference. I have been focusing on the following routers:

• Netgear Nighthawk R7000 [Amazon, $185]
• Asus RT-AC68U (Amazon, $175), -AC68P (Best Buy, $188), -AC68R (Target, $199)
• Asus RT-AC66U ($130)
• Asus RT-N66U (Amazon, $103)

While I am smart with a technical background in engineering, I have no background in either networks or its technical aspects.

Any thoughts, words of wisdom, etc? Given the potential prices, I would like to get at least two years out of the equipment.

 

[Nullity]

I assume Mbps = megabyte? (If it is "megabit", your network is severely degraded from the norm.)

Is your Comcast 75Mbit symmetrical? (Same upload/download)

Your network seems fine to me. 75Mbit is going to max out at ~9Mbyte/sec, and you almost get that in every room. The upload is low, but I have no idea what tier you are paying for (asymmetric?).


What you need to test, is internal speed from a wired client to a wireless client using something like iperf/iperf3 (or simply just transfer a file and note the speeds). I get ~185Mbit with my RT-N66U and practically anywhere in my 2 story house can get >100Mbit.

 

[tomliv]

We have the Comcast "Performance Pro" package which is 75 Mbps download/5 Mbps upload. Mbps = megabits per second. I'll update the initial post.

I also just measured the wireless performance in the upstairs room from a different website (speedtest.net) and got 11.3 Mbps download and 4.05 Mbps upload.

 

[Nullity]

Oh, if you are getting <10Megabit, that is alarmingly bad.

Can we get information about distance, along with floor & wall materials?

 

[tomliv]

Distance from the router on the first floor to the 2nd floor room is about 30 feet horizontal, 10 ft vertical. Floor is 100 year old wood; wall materials are plaster and wood lathe. It's a 100+ year old house, and the walls are original construction.

However, when I was in the same room (i.e., the living room) about 10 feet away from the router, the download Mbps was still in the 4.5 to 6.7 range. I didn't see a huge difference between the various routers (C7, C8, Comcast provided). When I did the speed tests, all three routers were on and providing a wireless signal (only one was connected to the modem). Currently, I just have the C7 and Comcast provided router still on; they're on different channels though.

 

[tomliv]

Since the Comcast modem/router is within a foot or so of the C7, I turned off the Comcast modem/router and ran a speed test again on speedof.me in the upstairs room on the C7. Rates were 16.25 Mbps download (max 21.3) and 6.27 Mbps upload (max 6.51). A repeat test within 1-2 minutes gave me 7.63 Mbps download and 2.93 Mbps upload. A third test game me similar rates as the repeat test.

 

[Nullity]

What speeds do you get when you run speedtests from a wired computer?

 

[tomliv]

Here you go ...

Personal laptop (HP ProBook 4440s) wired directly to the modem (SB6183): 58.66 Mbps download (93.7 Mbps max) / 6.46 Mbps upload (6.52 Mbps max)

Personal laptop wired to Archer C7 Ethernet (i.e., modem wired to C7; C7 wired to laptop): 8.38 Mbps download / 5.24 Mbps upload

I also ran speedtests on my wife's MacBook and my work laptop (Dell Latitude E5430) using speedof.me and got the following results:

Living room (same room as router):

• MacBook: 30.1 to 35.9 Mbps download / 6.4 to 6.5 Mbps upload (max 9.29 Mbps)
• Dell Latitude (using Internet Explorer): 34.6 Mbps download / 6.5 Mbps upload
• Personal laptop (HP ProBook; using Google Chrome): 7.22 to 7.85 Mbps download / 1.5 to 2.2 Mbps upload

Upstairs room:

• MacBook: 30.3 to 38.6 Mbps download (max 73.1 Mbps) / 6.2 to 6.5 Mbps upload (max 9.9 Mbps)
• Dell Latitude: 18.5 to 27.2 Mbps download (max 31.3 Mbps) / 5.9 to 6.4 Mbps upload (max 7.6 Mbps)
• Personal laptop (HP Probook): 5.8 Mbps download (max 7.6 Mbps) / 1.9 Mbps upload (max 3.5 Mbps)

Given the results of the different laptops, I've got to wonder if the issue is in my personal laptop and not in the network. I also do not understand why the speedtest results of the laptop wired to the router were so different than the laptop wired directly to the modem.

 

[tomliv]

Thinking there may be something wrong with my personal laptop, I checked the wifi adapter. It's a Qualcom Atheros AR9485 802.11b/g/n adapter. The driver date was Aug 2012 (version 10.0.0.75). I discovered how to update the driver via the computer, did so, and now the driver date is Nov 2014 (version 10.0.0.303). One small success.

I redid the speedtest several times in the upstairs room and got typically 26 to 35 Mbps download / 6.3 to 6.5 Mbps upload. Quite a difference from the previous results. The driver update apparently had froze the Acrylic Wi-Fi Home program that I had running throughout all the testing on my laptop. When I restarted the Acrylic Wi-Fi Home program and repeated the speedtest, the download and upload speeds returned to low numbers (4 to 5 Mbps download / 2 Mbps upload). Guess I should only use the Acrylic program sparingly to make sure that I'm on a relatively open channel and then turn it off.

Does this all make sense? Any reason for the wide difference in download speeds (lowest was 13.3; highest was 47.6 Mbps)? Anything else that I can do to maximize the laptop's ability to take advantage of our Internet package speeds?

Also, given the new results, is it going to be worth going to one of the more expensive routers I listed? If so, any recommendation? Why?

 

[Nullity]

I am hesitant to begin trouble-shooting because I cannot tell whether you are having troubles with your internet connection or wireless.

Please test wireless only to trouble-shoot wireless.

 

[tomliv]

Please test wireless only to trouble-shoot wireless.

How do I do that?

 

[avtella]

Don't use a WiFi scanner (Acyrlic in your case) while doing speed tests, as its causing an increase in scanning rate which leads to a drop in WiFi performance and resulting speed tests, this is from my experience using inSSIDer.

As for testing wireless use regular file transfers between a WiFi device and an Ethernet device (WiFi to WiFi will be slower) and or use a software like Tamosoft Throughput Test.

 

[tomliv]

For file transfers, I assume that I have to set up a home network? If so, I'll figure out how to do that. Will I be able to transfer the files between Mac and PC? Will I have to use the Tamosoft to measure the transfer rates or will the transfer rate display during the file transfer?

 

[avtella]

For file transfers, I assume that I have to set up a home network? If so, I'll figure out how to do that. Will I be able to transfer the files between Mac and PC? Will I have to use the Tamosoft to measure the transfer rates or will the transfer rate display during the file transfer?

Tomliv, you can do transfers between your PC and MAC that should be fine, transfer rate should display as your sending/recieveing the file. You don't have to use Tamosoft, I just gave that as an example.

 

[tomliv]

Ok; thanks. Will try to test this out tonight.

 

[stevech]

If file transfers are infrequent and large, a USB3 drive is easier.
Sneaker-net.

 

[tomliv]

Ok; thanks. Will try to test this out tonight.

Well, after much trial and error, I was able to get file transfers completed between MacBook and personal laptop and also between my personal and work laptops. The transfer rates were between 4.5 and 5.5 MB (megabyte)/sec.

I was going to try the Tamosoft, but OS X 10.7 was not listed.

I also ran a speedtest after plugging the personal laptop into the router by cable. It was 80 Mbps/sec download and 6 Mbps upload, so the earlier, slower results I got were a result of having the Acrylic WiFi Scanner on.

 

[fiberboy]

Netgear would be my choice.

 

[tomliv]

Thanks for everyone's input. I picked up an Asus RT-AC66R to compare with the Archer C7 and C8. After working to optimize the conenctions, in the upstairs room, the C7 was consistently getting downloads in the 35-45 Mbps range and uploads around 6 Mbps. The C8 was getting downloads in the 50 Mbps range with uploads the same as the C7. The AC66R was getting downloads in the 85-90 Mbps range with uploads the same as the other two.

The C7 and C8 are being returned due to the performance of the Asus. Also, the Asus was smaller, heavier, seemed more solid, had a removable stand (unlike the C8), and had rubber feet to keep it from sliding on the surface (unlike the C7 which when touched would move).

 

[Trip]

Interesting anecdotal evidence there regarding the variable download speeds. I'd presume Asus or Netgear to have a slight signal quality and coverage advantage over TP-Link for models in the same class, but double the throughput, that's pretty significant, depending on what you're doing with those endpoints.