Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[nrms]

The three aerials on the DIR-855 are rather short, and the WiFi coverage area doesn't seem very good compared to my previous WiFi router (a Vigor 2600G).   

A few years ago I bought a pair of high gain aerials for the Vigor 2600G, they are almost twice as long. Is it possible to extend the range of the dLink DIR-855 by replacing the two outer aerials with my pair of high gain aerials?  Or would all three aerials need to be replaced?  Or is the whole idea a waste of time?

Thanks
NigelS

[funchords]

The three aerials on the DIR-855 are rather short, and the WiFi coverage area doesn't seem very good compared to my previous WiFi router (a Vigor 2600G).   

If that's based on signal strength alone, then ignore it.  The MIMO antenna arrays give misleading strength because it's measured passively (it's not measuring transmissions from the 2 antennas that would be used to communicate with your particular station).  The proof is in how well and how far you can communicate when associated.   

A few years ago I bought a pair of high gain aerials for the Vigor 2600G, they are almost twice as long. Is it possible to extend the range of the dLink DIR-855 by replacing the two outer aerials with my pair of high gain aerials?  Or would all three aerials need to be replaced?  Or is the whole idea a waste of time?

Experimenting is good, but any antenna you try will have to be dual-band -- 2.4 and 5 GHz. You do not have to replace all of them, but not doing so will make the footprint less symmetric (which is sometimes good).  Vertical antennas radiate sideways.  To get a nice round sphere (if the router is in the center of a multi-story house), all of these antennas should be at an angle, including the center one.  See http://www.youtube.com/watch?v=5zIrr91E4Kc for an example.  To better prefer a more horizontal plane (like a single-story house), point them mostly vertical or (perhaps better) just slightly off vertical. 

[nrms]

Many thanks for your insightful reply. I am connecting mostly 802.11n clients on the 2.4GHz channel and the signal strength (based on the Network & Sharing Centre in Visa is 5 full green bands and described as Excellent), when I am some 20 feet away from the dLink transmitter but in another room on another floor. When I connect to the 5.0GHz channel and go and sit in the same place, I only get a Fair signal & 3 green bars. 

I am not sure, but I doubt whether the old high gain aerials I had were dualband, 2.4/5.0GHz. I will have to check with the supplier.

NigelS

[nrms]

Well, I tried experimenting, and it would appear that my older high gain aerials do not make any significant difference to the connections I can make with my laptop.

I am finding that when I connect to the 5.0GHz band radio, I have excellent reception and a 300 Mbps connection when I am a few meters away from the router in the same room. However, moving downstairs to another room, where I am no more than 8 meters away by direct line, but working through two walls, I only get a "Fair" signal and only a 120 MBps connection as measured by Vista network connection status.  Is this normal? 

Do dLink sell replacement high gain aerials that are dual band compatible that would boost the 5.0GHz signal considerably?

Thanks
NigelS

[Fatman]

5Ghz suffers more from degradation by passing through unfavorable mediums that 2.4 ghz.  It is advised for line of sight operations, and where 2.4 Ghz is just too busy.

Any time you are moving through a floor we are in dangerous territory, there are enough unfriendly surfaces in your average multi-story divider to make connecting iffy.

How do you do at 2.4 Ghz?
How are your aerials positioned?

[nrms]

In the same location downstairs I get "Excellent" (5 bars) and a 130 Mbps connection on the 2.4GHz band.

The aerials are arranged thus: \ | /

I did experiment with moving the middle aerial of of plane backwards (to the notch) as recommended by funchords' U-Tube video; but that didn't make any noticeable difference that I could see.

[Fatman]

Is your router on the stand or lying flat on the desk?

If there were a Cartesian coordinate plane with it's origin in the centre of your DIR, it's X axis parallel to the LEDs on the front of your DIR, it's Y axis parallel to the shorter side of your DIR, and it's Z axis coming straight up and down from the centre of you DIR.  Where approxomitely would you be trying to connect from in meters?

For example (1 (meter to the right),3 (meters in front of the DIR),-3 (meters below the DIR))

[funchords]

I am finding that when I connect to the 5.0GHz band radio, I have excellent reception and a 300 Mbps connection when I am a few meters away from the router in the same room. However, moving downstairs to another room, where I am no more than 8 meters away by direct line, but working through two walls, I only get a "Fair" signal and only a 120 MBps connection as measured by Vista network connection status.  Is this normal? 

Completely normal -- it's a feature designed to keep your thoughput as high as conditions will permit. Don't worry about bars or "Fair" or "Excellent."  Worry about results -- throughput.  If that's falling, then these other indicators are helpful as to figuring out why.

The best datarate performance (small b is bits, large B is bytes) is only possible in the best conditions.   If D-Link were to fix the datarate to 300 Mbps, then the error rate would shoot through the sky as you moved about and reflected/refracted signals caused the digital peaks and valleys (1s and 0s) in the signal to fuzz or blur.  (this is called multi-path phasing and its about signals bouncing off of stuff - walls, humans, dogs, wires).  So what they do when the error rate passes a threshold is to lower the datarate until the error rate drops.  If you go one more room, it might even drop lower.

You'll also get temporarily knocked off of 300 Mbps when a neighbor signal is heard, when there's a random bit of static on the power line, or if your station (your computer) gets busy with the CPU.

THE GOOD NEWS IS THAT your throughput isn't necessary cut when the datarate gets cut, unless you were transferring data with another network device capable of those speeds.  So if you're transferring at 5 MB/s (40 Mbps) at 300 Mbps datarate, you'll still do 5 MB/s (40 Mbps) at 120 Mbps.   

Do dLink sell replacement high gain aerials that are dual band compatible that would boost the 5.0GHz signal considerably?

Someone does, if not D-Link.  You're looking for RP-SMA or Reverse Polarity-SMA connectors for the antennas, and they must be 2.4/5 GHz dual.  Any antenna that works for A/B or A/G with that connector will work, even if it doesn't mention N.  The only problem with that idea is that higher-gain antennas work by "flattening" the spheroid pattern of a dipole (they send and receive even farther side-to-side BECAUSE they send and receive less above and below).  The footprints of typical dipole verticals are more sphere-shaped and the higher-gain verticals are more pancake-shaped.   As a result, switching to high-gain antennas could make use in your home seem "spotty."  Try it -- just know that experiments are good -- but they're not always successful.

[nrms]

Is your router on the stand or lying flat on the desk?

If there were a Cartesian coordinate plane with it's origin in the centre of your DIR, it's X axis parallel to the LEDs on the front of your DIR, it's Y axis parallel to the shorter side of your DIR, and it's Z axis coming straight up and down from the centre of you DIR.  Where approxomitely would you be trying to connect from in meters?

For example (1 (meter to the right),3 (meters in front of the DIR),-3 (meters below the DIR))

Lying flat.

x: 0 , y: -6 z: -8 approximately. The router is sited against an outside brick wall. I am in a conservatory against the same outer wall.

[nrms]

So if you're transferring at 5 MB/s (40 Mbps) at 300 Mbps datarate, you'll still do 5 MB/s (40 Mbps) at 120 Mbps.   

I tried downloading a large file (7GB) from my NAS yesterday. Sitting about 2 meters away from the router, with the 300 Mbps rate, I got about 7MB/s transfer speed (as measured by Vista dialog box); moving down to the other location with 90-120 Mbps rates; the file transferred at about 2MB/s. 

[Fatman]

OK, starting with the aerials all at 0 degrees (pointing straight right as seen from the front of the device) you might want to try moving the aerials (from left to right) as so:

1 rotate 90 degrees, articulate 0 degrees
2 rotate 90 degrees, articulate 45 degrees
3 rotate 0 degrees, articulate 0 degrees

And to elaborate on funchords cautionary tale about gain, in your case if you have higher gain aerials it will almost guarantee worse performance.  You need the lower gain spread to have any chance of getting reception in your unusual connection locations.  If all your clients were to intersect an imaginary plane (for an impossibly high gain aerial, more commonly it would look like a torus with an diameter proportional to the gain and a radius inversely proportional to the gain, that is being squashed along the Z axis at a rate proportional to the gain) perpendicular to the middle of multiple of the aerials, then we could talk gain.

what I believe funchords is trying to explain is that if you had the same EMR conditions and you dropped phy speed it would not necessarily be detrimental to performance.  Not that you should get the same performance no matter where you are.

What I am saying is that your environment is poor and the source of your problems.  Period.

Moving the aerials to match the above will give you better performance for that position at the expense of the general position mentioned before (which is expecting you to be on the same Z more or less as the router).  Even the layout I just gave you pretty widely spread, it is just less so that the default.  This does not mean that you will see better performance however, that is up to your environment, which no amount of forum posts can properly elaborate.

[funchords]

I tried downloading a large file (7GB) from my NAS yesterday. Sitting about 2 meters away from the router, with the 300 Mbps rate, I got about 7MB/s transfer speed (as measured by Vista dialog box); moving down to the other location with 90-120 Mbps rates; the file transferred at about 2MB/s. 

That's awesome.  7 MB/s is 56 Mbps -- that's rockin!  2 MB/s is 16 Mbps -- which is very good as well.

Was the NAS attached to the router via Ethernet or is it wireless?   (Please don't make me jealous and say "wireless") 

[nrms]

Sorry, but wireless... This is using the DIR-855's dlinkmedia 5.0GHz 802.11g channel connected to the Intel 4965AGN wireless card in my dell laptops. I set Auto 20/40 and it connects at 40MHz channel width (wideband) and reports a 300 mbps connection for my laptop network connection.

Sadly, though I am still getting occassional stuttering when trying to playback blu-ray films wirelessly from the buffalo terastation live NAS; though I suspect this is either due to the rather poor performance of the NAS  device, or else in the decoding of the ISO on my laptop (the film is ripped to my NAS as an ISO file and mounted on Daemon Tools Lite Virtual Drive and played back using Cyberlink PowerDVD Ultra.

NigelS

[Fatman]

Blu-Ray over IEEE802.11 standard or draft wireless is a bad idea.

That statement requires no qualifiers, however I will explain quite lengthily below.  You could quit reading now and get what I am trying to say though.

Blu-Ray 1X is 36Mbps
Minimum bandwidth for BD-ROM movies is 54 Mbps
The Blu-Ray association themselves admit that 2X is the lowest allowable speed for real world applications.  That is 72Mbps.

Notice how even if the movie itself is limited to 54Mbps, you still have to read faster than that due to meta data and other extraneous variables.  In this case the total throughput from NAS to converter needs to match that requirement, as that is effectively emulating the optical system in a Blu-Ray player in this environment.

Even if you had a strong 300Mbps connection the nature of wireless is inherently susceptible to interference:

(2.4) If your cell phone or blue tooth devices hop to that frequency, you will stutter.
(Both) If your cordless phone is on that frequency, you will stutter.  Possibly only on a heart beat, or only when someone is talking (and it could be when either one talks, when one side in particular talks, or other combinations), or only while ringing.
(2.4) Your microwave will cause stuttering.
(Both) Any other Draft N devices will cause stuttering.
This list goes on.

This list is biased to 2.4 Ghz (even though you mention 5Ghz) because I have been more concerned about that frequency for a while (and it is more common to my knowledge), there WILL be examples specific to 5Ghz.

And before you tell me you live in an empty box with none of these devices, remember that neighbors may not live such spartan lives.

Most of the time you won't notice this because you are either not using real time applications, or they aren't taking any significant amount of bandwidth.  With Blu-Ray going you are going to be using way more real time bandwidth than I would ever count on.

While we are at it, using a file transfer in windows to measure throughput is fraught with peril, though that point is moot by now.

Also, you have me slightly confused with this:

DIR-855's dlinkmedia 5.0GHz 802.11g channel connected to the Intel 4965AGN wireless card in my dell laptops.

I assume you mean 5.0Ghz Draft N, however I feel compelled to add that IEEE802.11G is 2.4Ghz and 54Mbps only.

Throughput on the NAS makes sense as a bottleneck.
As does wireless Draft N specification.
As does specific wireless environment.
As does decoding and playback on your PC.

*** Edited to make my statement about Blu-Ray over wireless more true, I realized that there were some more exotic technologies that would be used that are "wireless" when I got home. I changed it to specifically reflect IEEE 802.11 standards and drafts. ***

[funchords]

First-class job, Fatman!