Nighthawk Isn't Improving My Home Networking Issues

[HanSabaacSolo]

Hey all,

So, in a short synopsis, I had a 4th Gen Airport Extreme for many years (purchased at initial release). Here lately, the router has been crapping out on me, so I decided to make the plunge and future-proof with the Nighthawk. After going through the initial setup, then going a bit deeper with some more technical tweaks, I am not seeing much of an improvement...

I have our iPhone 5's connected to the 5GHz band, as well as two Mac Pro's. As far as the 2.4 band is concerned, I have the PS3, Vita and PS4 connected.

We have 60 Mbps with Charter Spectrum however I am only getting speeds of 14 Mbps on the consoles, and around 30 Mbps on the Mac Pro's and iPhones.

Any advice would be excellent because I feel like I didn't necessarily need to spring for the Nighthawk.

 

[thiggins]

I think the 4th Gen Airport Extreme is N900 class, i.e. simultaneous dual band 3x3 radios.

Unless your devices are 11ac, you should not expect that a higher class router is going to improve things much. And in many cases, it can be a step back, since 11ac routers sometimes don't play as nicely with older devices and N routers do.

5 Tips For Buying A Future Proof Router

 

[sfx2000]

Not sure if folks want to go back and edit - I believe the Airport being mentioned is the 4th Gen Airport Extreme (not express) - and as Tim H. suggested, it's a dual band 3*3 802.11n router.

There's only been three generations of the Express - the original 802.11g, the 802.11 a/b/g/n single radio (N300) , and the current 2012 edition that is simultaneous dual-band (N600 class)

 

[azazel1024]

Depending on where you are testing those clients, that isn't suprising. If those Mac Pros aren't knew enough to have 11ac clients in them, in all likelihood, any performance differences would be mild.

I've certainly seen some performance and range increases moving from an older router to a newer router utilizing the exact same protocol, band, channels and channel width, but they've generally been mild.

From my old Netgear N300 where I could get about 22MB/sec on 2.4GHz 40MHz, to my newer N600 WDR3600 I get around 23-24MB/sec and with my brand new AC1750 router I get around 26-28MB/sec. That is only about 16% from "mid aged" router to newest and right up on the limits of what should be theoretically possible with a 300Mbps 11n connection. About 27% from a 5 year old (6?) router to a brand spanking new one. Performance at long distance is up a little more than that between the generations and actual maximum range is up a little too, but that can be more about the quality and gain of the antennas, signal amplifiers and broadcast power than it is about real differences between newer and older routers (or 11n vs 11ac, since I am mostly looking at the 2.4GHz band here).

5GHz is up HUGELY in performance, but that is comparing 300Mbps 5GHz to 867Mbps 5GHz as my laptop has an Intel 7260ac 11ac adapter in it.

My tablet with a 150Mbps dual band adapter in it sees pretty much no performance difference between my N600 WDR3600 and my AC1750 with the exception that at longer distances, the AC1750 seems to perform very slightly better. Maybe 10-15% better at medium to extreme ranges on both bands. Maybe. My iPhone 5, my wife's iPhone 4s and her iPad 2 see effectively no difference, but I also haven't done much performance testing with them.

They do work better than with my old N300 2.4GHz only router, especially since my phone can connect at 150Mbps on 5GHz and only 65Mbps on 2.4GHz, but compared to my N600 router/access point, the performance is not noticable at all, though I haven't tested at really long distances from the AP and router for those devices (just tablet and laptop, as those are the ones I really care about).

The only real way to improve performance significantly is to change both the basestation (router/AP) AND the client so that they are on the same protocol. If you started with an 11n wide setup and moved the router/AP to 11ac, the current clients will benefit little, if any. Any improvements are going to be related to possibly slightly better radios, amps, processing power, maybe better/bigger antennas (supposing the older ones weren't good to begin with), etc, which are all likely to be minor improvements in speed.

Now moving the basestation to 11ac AND one more more clients is likely to see the clients that are upgraded having a significant performance improvement.

My laptop saw ~16% on 2.4GHz from N600 to AC1750. Maybe +25% at the absolute best location far from the router. On 5GHz my laptop (whcih already had an 11ac adapter, so could take advantage of the better protocol for 5GHz) sees about a 120% performance improvement close to the router and around 140% at the best location at a medium/far location (smallest difference is around +90% faster).

 

[drawz]

How are you testing? Maybe you should also try a wired connection to eliminate wireless as a possible cause of trouble. You could also try connecting directly to your cable modem (also wired) to eliminate the router as a possible problem.

Could you also define "crapping out on me"? Have things gotten worse over time?

The other thing to check is the router settings - consider resetting to defaults. Is there any QoS enabled that might be throttling? Are you using a 40 MHz wide channel? Are you on a crowded channel that is used by a lot of other routers in the area? Do you live in an apartment?

I have used several N150 to N750 routers running OpenWrt and Tomato on connections up to 50mbps. The 2.4 GHz band is pretty crowded around me, so I usually couldn't get the full 50 mbps, but no issues with the 5 GHz band running a 40 MHz wide channel.

To be honest, if you don't have 802.11ac devices (and it doesn't sound like you do), the Netgear probably doesn't give you any major advantage.

 

[remixedcat]

No 40Mhz in 2.4 and throwing high gain on one router isn't the answer. Instead use access points to increase coverage, have those on different channels and use 20Mhz for 2.4 and 40/80 for 5.0. Turn transmit down to create smaller cells that border eachother (size may depend on clients) and configure roaming for such.

 

[stevech]

No 40Mhz in 2.4 and throwing high gain on one router isn't the answer. Instead use access points to increase coverage, have those on different channels and use 20Mhz for 2.4 and 40/80 for 5.0. Turn transmit down to create smaller cells that border eachother (size may depend on clients) and configure roaming for such.

Agree. Except turning down transmit power shouldn't be done.

 

[remixedcat]

Agree. Except turning down transmit power shouldn't be done.

only when you have multiple and you want seemless roaming and no "sticky clients"

enterprise APs have controls to mitigate that, however consumer grade ones don't. so ...

 

[azazel1024]

only when you have multiple and you want seemless roaming and no "sticky clients"

enterprise APs have controls to mitigate that, however consumer grade ones don't. so ...

Depends on the use case, but sometimes.

I did that with my AP when I relocated it. Now it isn't in the RF shadow of my massive fireplace (only a very tiny part of two rooms are, like 10% of those rooms). The increased signal reduced roaming on clients at the far end of my house where I wanted them to roam to my router the floor below them (directly below them), as it is faster and divies up the wireless bandwidth better.

I turned down the 2.4GHz Tx power to minimum from max. Medium didn't seem to do much, but minimum decreased RSSI by 3dBm supposedly, and my iOS clients do roam much more readily.

Pre-AP relocation, I'd walk over to the far side of the house and they'd roam in 8-12s. Post-AP relocation, more like a minute+ before they'd roam. At medium, maybe 30-45s. At minimum it is down to 10-15s. Windows was happy roaming at nearly the same speed at high broadcast power, but it is a little snappier on low broadcast power.

RSSI pre-relocation was -78dBm, post relocation -61dBm. Minimum power broadcast it is -64dBm.

5GHz was fine and I didn't touch the broadcast power there, as pre-relocation you couldn't see 5GHz (or it would show up as -92dBm or so and windows wouldn't show it). Post relocation at max power it is -81dBm and if a device was connected on 5GHz, it roams just fine.

I still have yet to do much testing, but in the relevant areas cranking down broadcast power has had less than a 5% impact on performance, as most areas I care about are same room, or ajoining room from the access point (these are the "larger" rooms in my house, kitchen/connected open dinning room, living room and my kids play room/family room). In the far locations, if I set seperate SSIDs for testing (because they should be connected to the router), minimum power dropped performance about 20% on 2.4GHz, but of course Tx only. Rx has seen no impact at all in any location.

 

[remixedcat]

I have a tempermental laptop WLAN card as well as two phones that have bad sticky client issues so I have it calibrated to those since they are very heavily used devices.