[moep]

I recently bought a new router to replace my old WRT54GL.
It’s a WRT610N with dual radios which allows me to run a b/g mixed as well as a seperate 5Ghz n-only network.

It all works pretty well except for the n-only network. My Unibody Macbook and MBP show a maximum connection speed of 130 Mbps rather than the 300 Mbps I expected. Actual throughput with bulk file transfers is around 50-60 Mbps which seems about right for a 130 Mbps network.

But does anyone have an idea why I’m not getting the 300 Mbps Link speed of a regular 5Ghz Draft-N network, even with as little as 1m between the router and client?



The n-only network is set to 5Ghz, 40 Mhz Channel Width only.
Interferences shouldn’t be an issue because there is no second WLAN within 200 meters range, no cordless cellphones, no microwave, no wireless surveillance cameras, nothing.

[Simon]

I don't know where the industry pulled that 300 Mbps figure from. I have never seen an n network get more than 130 Mbps. And 50-60 Mbps real world sounds just about right. The best I have ever measured (and that's with client and router right next to each other) was 65.6 Mbps.

[turtle777]

It is a well known fact that the manufacturers hopelessly exaggerate the speed of the networks and equipment.

-t

[hwojtek]

What your Network Utility displays is the highest speed that a connection has been attempted with. For real connection speed, check on the router side (you have a www-based router setup or something like that for sure).
My computers show 300 mpbs steady as a rock, however the Airport Utility displaying my Airport base stations stats shows that the speed constantly fluctuates between 216 and 300.

Oh, and you will get the 300 mbps if:
a) your router is 100% compatible with your Airport card and
b) you run 5 GHz n-network ONLY.

[Simon]

Originally Posted by hwojtek 

a) your router is 100% compatible with your Airport card and

check, both are Apple's latest products

b) you run 5 GHz n-network ONLY.

check.

130 Mbps is still the limit. 300 Mbps is wishful thinking at best.

[hwojtek]

I have two N-enabled computers and two AEBS'es - they all run at least 216 and up to 300 peak, according to logs from both AEBS'es, so you definitely must do something wrong.

[Simon]

And what real world throughput does hat correspond to?

[turtle777]

Originally Posted by hwojtek 

I have two N-enabled computers and two AEBS'es - they all run at least 216 and up to 300 peak, according to logs from both AEBS'es, so you definitely must do something wrong.

Originally Posted by Simon 

And what real world throughput does hat correspond to?

Exactly. Copy some large files (e.g. movies) and time it, then calculate the throughput.

I bet you never even et close to 300.

-t

[Simon]

Yeah, I think the misunderstanding here is that the link gets reported as having a certain "speed" which can be up to 300 Mbps. But if you then measure actual throughput the very best n networks won't even get 100 Mbps.

[steve626]

I think the original query might not have been fully understood by some of the responders here. One can run Network Utility which may indicate 130 Mbps (N router, 2.4 GHz) or 300 Mbps (N router, 5 GHz). These numbers indicate only the maximum speed with each kind of connection -- the actual speeds realized are always lower than these values, but they do indicate what type of N connection is being made. The actual speeds realized are slower, as many have pointed out, typically by a factor of 2 or so. I think the original query was why he always sees the 130 Mbps number (max value for 2.4 GHz) and never the 300 Mbps number (max value for 5 GHz), since his router was set for 5 GHz, or so he thought. I think the question was: why are his Macs always selecting the 2.4 GHz N signal instead of the 5 GHz N signal?

I am guessing that for some reason his 5 GHz N connection is being dropped in favor of a 2.4 GHz N connection. The router might be deciding that the SNR is better at the lower frequency, but I suspect there is something in the configuration of the router that needs to be investigated. For instance, I would disable entirely the 2.4 GHz functionality of the router as an experiment just to see what his Macs do when presented with ONLY a 5 GHz signal from his router.

There are a number of reasons a Mac may refuse to use the offered 5 GHz signal and instead go for the 2.4 GHz signal from an N router. There could be a very strong 5 GHz signal nearby. If the neighbors' network is not broadcasting its SSID, you might not be able to "see" it using common tools for this sort of thing. Someone nearby could be using a strong repeater that is jamming your 5 GHz signal.

By the way, just walking around my house (all the dwellings in my neighborhood are separate houses, fairly well separated in fact) with my laptop I can see more than a dozen different networks from my neighbors. I go to one side of the house, 10-15 different ones pop up, and on the other side of the house there are a different set of a dozen or so. I have no idea which ones are 5 GHz or not, but they are becoming more common. In an apartment or condo complex, the situation could get much more congested even.

I recently purchased a Linksys WRT160N router, and it provides 2.4 GHz N signals only. I was initially wondering if I should have gotten a 5 GHz + 2.4 GHz device but after reading reviews and anecdotes on the web, I found that the 5 GHz signals are often not usable because they seem more "fragile" for some reason. For one thing, the shorter wavelength at 5 GHz means that it won't penetrate walls, floors, buildings as well, which may be part of the issue. The poster said he wasn't getting on 5 GHz even when 1 meter away, however, so I think some experimentation with router settings might be in order here.

[Cold Warrior]

It's unlikely all the wifi points you're seeing are 5 GHz. That freq does not propagate as well as 2.4 GHz (lower frequencies propagate better than higher ones). Also the channel numbering is different -- use a utility like airmoose to examine the wifi points. Channels 1 through 11 identify 2.4 GHz points.

[mduell]

802.11n (draft whatever-we're-on) is ~150Mbps (can be 130Mbps for reasons I don't understand) per stream, with a maximum of 4 streams (600Mbps). Typical throughput is only 150/130Mbps due to the use of cheaper (2 or 3)x(2 or 3):2 (where at least one of the (2 or 3) is usually a 2) radios instead of the more expensive 3x3:3 (like Intel PRO/Wireless 5300) or even 4x4:4 (specification maximum) radios.

[gaming4fun]

Agreed. Theoretical max is between 200Mbps to 500Mbps. Actual speed is much lower depending on the environment between the AP and your computer.

[midasnick]

Actually, what you're all perhaps not aware of is that some Intel adapters do not support the broader 40 Channel Width option. This (by design) limits their connection speed to 130. Check the specifications of your Intel adapter to ensure that it is capable of bonding on the 40Mhz range to achieve the full 300Mbps, probabable limitations on the macs.

[Simon]

I'll only add to that that Apple does not use Intel wifi circuitry. Not even in MBPs that otherwise used Intel chipsets. Also, no Intel circuitry in AP or TC devices. Atheros has been very popular with Apple though. There's more, but I forgot which ones and I'm too lazy to check right now.

However, it's clear that if somebody is using a Mac with a third-party access point that makes use of Intel's wifi chipsets, what you mention could be a problem.

[midasnick]

Hi Simon, nice to meet you. I'm on a MacBook Pro and every now and then i use pc, and a PlayonHD Media Player
The pc/PlayonHD transmit between 270-300MB/s but the mac always peaks at 130MB/s. I'm beginning to get curious as to why. Everything is running on N mode.

[turtle777]

Originally Posted by midasnick 

The pc/PlayonHD transmit between 270-300MB/s but the mac always peaks at 130MB/s.

MB as in Mega Bytes ? Are you sure ?

-t

[Simon]

No, quite obviously he's talking about Mbps = Mb/s = megabits per second.

802.11 ratings are always given in Mbps.

[midasnick]

correction, Mb/s, thanks for pointing that out.