Someone help me, I'm having a router-buying addiction.

[VirtualLarry]

I just bought like, 4-5 different router models to test out. (Just signed up for Gigabit FIOS again.)

Curious how some of them perform. Most were in the $30-40 range on ebay. Gigabit LAN/WAN, AC1200/1750/1900.

Got some Asus, some Linksys/Cisco, some Netgear, some TP-Link, and I'm not sure what else.

 

[Gryz]

If you are really testing them out, please write down your findings. And please share them with us. I'd be curious to hear if there are any differences.

(I got a 7.5 Mbps ADSL2+ line at home myself. So my interest is purely academic. )

 

[mxnerd]

LOL. VL you probably not only have router-buying addiction, you also have motherboard-buying addiction.

 

[mnewsham]

I feel like you probably could have bought the hardware for a nice custom Pfsense build with that money... But hey, to each their own.

 

[FreshBross]

I feel like you probably could have bought the hardware for a nice custom Pfsense build with that money... But hey, to each their own.

I think both outcomes of spending time testing/hobbying around with hardware models and Pfsense can provide good info and experience. Now you made me want to learn more about setting up a Pfsense computer router

 

[shortylickens]

This got me curious. I went into my router settings for the first time in years, noticed a lot of features I never saw before. Didnt change anything, but it was good to refresh my memory.

 

[PliotronX]

First off, those are "routers" so you've got a toy addiction second, there's nothing wrong here. You will find the one you like and hopefully it takes full advantage of the fiber but I am guessing that most of them will have trouble delivering.

I think both outcomes of spending time testing/hobbying around with hardware models and Pfsense can provide good info and experience. Now you made me want to learn more about setting up a Pfsense computer router

Better to do it soon, my first time with it was throwing an extra NIC into one of those grey optiplexes with a core 2 duo and it worked great but the upcoming v2.5 will require AES-NI on the CPU. This doubly sucks because I currently have it in a VM on a Phenom X6 with Sophos UTM in bridge mode. Opnsense doesn't have this on their roadmap however so I will have to mess with that.

 

[aigomorla]

mmm.... once i went over to the enterprise level solution... ie.. smoothwall at first, then pfsense now, i can never go back to the commercial router.

Even my grandmother's house is running a tiny Pfsense box with a dedicated switch, and AP, just because i prefer the no nonsense / robust reliability that route offers over the consumer stuff, which is important when you have IP cameras monitoring her in case she has a fall.

 

[dave_the_nerd]

Christ, man. You don't have to buy one of everything to know how they work - they're 95% the same hardware. A review site and a spec sheet will tell you which one is better. Read the manuals to check feature support. You can figure this stuff out without hands-on.

You have a problem.

 

[JackMDS]

Christ, man. You don't have to buy one of everything to know how they work - they're 95% the same hardware. .

While it is true that there is only 6 manufacturers that produce WIFI Chipsets that also write write the core Firmware and the base Managing software, Wireless about transmission.

I.e., variables (that are strange to us) and thus ignored by your average computer "enthusiasts" are what make the difference even between similar core hardware devices.

Antennae, their quality and position make a lot of difference (it is not part of what the chipset manufacturers provide). Since the Quality of transmitted signal is dominated by Signal to Noise Ratio (Not by the "silly Bars"),the quality and position of the element in the plastic box plus the quality of the PSU and any other hardware added will make difference in the level of the electronic noise and would affect the final signal performance as well.

Most review site are skewed by Business arrangement decisions, the very few that are slightly more objective than the others tend to deal more with the Upper price Hardware.

As for refurb. it really any body call, you can buy and try few units of the same make and model, one or two will work superb, others might be Duds..

Then comes the enviroment while is Not affecting regular Computer hardware functions it is dominating the WIFI functional capacity.

Did One ever wonder why Apple Computer with WIFI while using the same chipset provide better Functional WIFI connection?

 

[VirtualLarry]

Well, thanks for coming to my defense, of my "seat-of-the-pants / experiential" methodology for testing things.

Also, thank you Dave for telling me I have a problem, I tend to "overdo" things, sometimes.

Anyways, yesterday, I received,
1) Asus AC1200 router, refurb
2) Asus AC1200G router, refurb
3) TP-Link Archer C900 AC900 router, refurb

Today, I received:
2) Edimax BR-6428nS V4 (kinda rolls off the tongue, NOT), new

And I'm expecting several more routers throughout this week.

I made the mistake of ordering the Asus AC1200 non-G model first, as it is limited to 10/100, as well as some Linksys E6100 AC1200 routers, which are likewise, 10/100 LAN/WAN only.

Why they make modern AC routers, with only 10/100 wired ports on them, I don'\t quite understand. Seems quite a bit limiting.

I've discovered that my main rig's "Aukey" 3x3 AC USB3.0 wifi, only seems to do around 200-250Mbit/sec, even with the router a few feet away, so wifi testing will need a better adapter, I have some AC1200 TrendNet media bridges, one of which I was using for a while to hit around 500-600Mbit/sec with my AC1900 Asus AC68R router. I also intend to test WAN-to-LAN wired speeds, using speedtest.net and my Gigabit FIOS connection. Testing that may require staying up late-night, so that the "internet tubes" are less clogged.

Edit: Also got in an Asus "Dark Knight" N66U (used, from private party - hey, it was $20 + reasonable ship), and a Netgear R6300v2 (which... seems to be a "Charter" version, not really specified in listing, I don't know if it has Charter or Netgear OEM firmware, hard to tell, seems like OEM firmware maybe).

Interestingly enough, the N66U isn't AC, but I get the same sorts of wifi speeds from that, the TP-Link C900, and the Asus AC1200G, around 220-250Mbit/sec with the Aukey AC 3x3 USB3.0 adapter with Win10 out-of-box drivers, but with the R6300v2, I was getting 400Mbit/sec with it. Also, it listed FOUR frequencies being used by the 5Ghz signal, which I didn't understand. It's not a tri-band (dual 5Ghz wifi), nor MU-MIMO that I know of, and only AC 3x3, so why four freqs.? I'll have to research that to understand it, I think. I thought only 4x4 routers used four freqs.

 

[JackMDS]

I made the mistake of ordering the Asus AC1200 non-G model first, as it is limited to 10/100, as well as some Linksys E6100 AC1200 routers, which are likewise, 10/100 LAN/WAN only.

Why they make modern AC routers, with only 10/100 wired ports on them, I don'\t quite understand. Seems quite a bit limiting.

10/100 is cheaper to make, most normal (None Enthusiasts) Consumers do not use more than 100 Mb/sec bandwidth.

While in the past technology was dominated by the Tech Pros (Engineers and the like) at the begging of the 21st century the power was transferred to the None tech Directors and Marketers.

To add to my previous post.

Most computer chipsets have Temperature control protection internally. I.e, they reduce Speed/function above certain heat production.

We know about it only when devices/component get to hot to touch but the Internal Mechanism reduces function way before we can sense it.

As a result a lot of small packaged WIFI work as their data sheet claims for very short time before the heat control kicks in.



P.S., I know that my info is hard to deal with in a society trumped by wishful thinking, by that the way it is. ---

 

[dave_the_nerd]

To add to above: average internet connection in the US is still only ~50mbps, and most routers never actually have anything plugged into the LAN ports anyway (everything's over wifi - laptops and tablets.)

vis a vis used routers - crack them open, replace the TIM and blow the dust out before you bother doing anything with them.

 

[VirtualLarry]

As a result a lot of small packaged WIFI work as their data sheet claims for very short time before the heat control kicks in.

That's very insightful, Jack, I hadn't even considered that. Very few of my routers actually "run hot".

What would be a good test of that, Jack? A speedtest over AC 3x3 wifi (or media bridge), when "cold", and then listen to internet radio for an hour (a constant load), and then another speedtest or so, and compare the speeds? I now have Gigabit FIOS, so that would make a good upper-bounds for the wifi to test against.

To add to above: average internet connection in the US is still only ~50mbps, and most routers never actually have anything plugged into the LAN ports anyway (everything's over wifi - laptops and tablets.)

vis a vis used routers - crack them open, replace the TIM and blow the dust out before you bother doing anything with them.

That seems slightly pessimistic, but you're right, there's probably more rural folk than city folk with high-bandwidth metro connections. I could, in theory, get a GigE internet connection from both FIOS and Comcast, and round-robin them. Would make for some amazing speedtests (since they often use multiple streams, so I've gotten speedtests before that were a combination of my FIOS and Comcast connections, at the time).

As far as routers and dust inside, probably a good idea, overall, but I'm hardly that meticulous. Although, that might be the reason that my refurb Asus AC68R running Tomato, occasionally drops the wifi for a few seconds at a time.

 

[dave_the_nerd]

That seems slightly pessimistic, but you're right, there's probably more rural folk than city folk with high-bandwidth metro connections. I could, in theory, get a GigE internet connection from both FIOS and Comcast, and round-robin them. Would make for some amazing speedtests (since they often use multiple streams, so I've gotten speedtests before that were a combination of my FIOS and Comcast connections, at the time).

https://www.statista.com/statistics/204952/average-internet-connection-speed-by-country/

Even us city-folks don't often have a good incentive to get more than the basic 20-50mbps packages.

As far as routers and dust inside, probably a good idea, overall, but I'm hardly that meticulous. Although, that might be the reason that my refurb Asus AC68R running Tomato, occasionally drops the wifi for a few seconds at a time.

If you end up passing these along to other people, you probably will want to. Heat-related failure is what kills routers (and prompts people to get new ones.) Avoid that.

 

[VirtualLarry]

If you end up passing these along to other people, you probably will want to. Heat-related failure is what kills routers (and prompts people to get new ones.) Avoid that.

OTOH... most of these were "factory refurbs". If overheating was the problem, shouldn't they (the "factory") have dusted and re-pasted the chipset(s), as part of the Refurbishment process? Or do they just swap plastic shells, to give it that "new" look.

 

[mnewsham]

https://www.statista.com/statistics/204952/average-internet-connection-speed-by-country/

Does that maybe include mobile connection speeds? Because speedtest.net's global index shows fixed broadband (as in a home wired connection, not a mobile connection) to be much much faster on average than that.

http://www.speedtest.net/global-index

 

[Ranulf]

Also, thank you Dave for telling me I have a problem, I tend to "overdo" things, sometimes.

I totally do not believe this. By the way, have you seen linksys' ebay page? Decent deals on refurbs sometimes.

https://www.ebay.com/usr/linksys-official-store?_trksid=p2047675.l2559

 

[killster1]

you didnt even say what your tests will be! just stay up late and run speed test over and over .. Sounds like a good use of fiber@home i personally only have 18mbit and seem to use more than 1000gb a month (so att keeps telling me) so i wouldnt mind no cap compared to higher speed. (18mbit max offered here) .. So funny yes you have problem with overspending on cheap things when you could have just bought one quality router and not bothered with middle of night speed tests and auky wifi adapters

 

[VirtualLarry]

I apologize, I intended last month, to do a round-up review sort of thread about the 10 or so AC / Gigabit routers I picked up for under $40 on ebay.

This project is on hold indefinitely, until I can clean up some space in my apt., and get my infrastructure put away properly. Right now, I'm on a semi-skeleton network setup.