anoclon
join:2004-10-12
Guatemala | reply to John Galt
Re: Remember "3 town wisp setup" ?
Thanks john, your opinion is very important to me. I'd like you to reply to this new topic of mine:
»Amplifiers of 1 and 2 watts. |
|
John Galt
Forward, March
Premium
join:2004-09-30
Happy Camp
 ·CenturyLink
| reply to anoclon
said by anoclon  :Just for the record... Thanks for keeping us posted on how your project is coming along.
--
A is A |
|
anoclon
join:2004-10-12
Guatemala
| reply to anoclon
Just for the record: I added an amplifier on my tower number 2, thinking that maybe the throughput was too low because signal strength and quality were very low too (74% or 75%). With this amplifier my signal strength and quality got to 86%. Even so, throughput was basically the same; about 140Kbps. |
|
anoclon
join:2004-10-12
Guatemala
| reply to anoclon
I did the tests this past weekend, 8 miles away from my pop I got better throughput than I got 4.5 miles away where my other tower is, even when signal strength and quality are basically the same (74% or 75%), throughput turned to be about double (about 120 Kbps). |
|
anoclon
join:2004-10-12
Guatemala
| reply to anoclon
I'm about to do some other tests using sane D-Link equipment (silly me). A couple of months before I had this new tower, when I was looking for the right place to put it, I remember downloading a big file from the internet at a 60Kbps transfer rate in another location 8 miles away from my base station and using same D-Link DWL-G520 pci wireless adapter. I'll go to the same spot this coming weekend and do the downloading test again. I'll keep you all posted. |
|
nboxmedia
join:2004-07-10
Mount Olive, NC
| reply to anoclon
Proper engineering is the key. Just because you can do it doesn't mean it's the proper way. The radios I have right now are only 80 mW (19 dB) and I can achieve up to a 12 mile link with them no problem, while maintaining a good SNR and fade margin. |
|
totalaccess
Premium
join:2002-10-04
Elgin, TX
| reply to anoclon
What did you replace it with, another Dlink ? I would suggest a better AP like a Universal Access Point by »wirelessworlds.com , or AirMatrix at »defactowireless.com and quit using the cheapest stuff one can find on the planet for your base stations.
Here in an insert from Defacto on their product.
Powered by our unique software, AirMatrix Home includes such abilities as; software selectable 2.4, 5.3 or 5.8GHz, Bridging, Routing, DHCP server, NAT, Access Control, and tons of other features which will keep you set apart from the competition. Providing real-life speeds from 3 to 35Mbps and greater, AirMatrix Home is a perfect solution to provide reliable broadband connections to your customers.
Again I highly recommend getting away from the $29-79 dollar base stations. The old saying you get what you paid for is really true about WISP base stations in my experience.
--
Deploying Global Solutions: »www.wirelessworlds.com |
|
anoclon
join:2004-10-12
Guatemala | reply to anoclon
I switched radios the past weekend and throughput was basically the same. At least I proved to myself that D-link radios can "almost" make it at a 4.5 mile link. I guess I'm gonna have to use a better radio to increase throughput. |
|
public
join:2002-01-19
Santa Clara, CA
·DSL EXTREME
| reply to VariableARK
said by VariableARK :i had a similar issue, i was using cb3+ deluxe(.g) radios, and i would only get 300k throughput on a one mile link with no interference, and 24dbi to a 16dbi directional, couldnt figure it out either, finally decided the radios suck, maybe its your case too. The RF design of a data link is only a half of the task. The other part is the radio software. What is often overlooked is that the retail grade products are designed for short range, and do not have a long enough window to listen for packet acknowledge. For a longer link the delay is too long for the driver, and packets are assumed to be lost and retransmitted. That is one of the reasons why the throughput is far lower that expected.
Some of the dlink products have extended range setting. There was a thread on the dlink forum. |
|
anoclon
join:2004-10-12
Guatemala
|  reply to superdog
I've just found out this topic: »Wireless users, READ THIS! you posted on another forum:
said by superdog :What is the latency (ping time) from your computer to the Wisp's gateway? Ping times to the Wisp gateway should not exceed 20 ms. We have clients who traverse over 5 tower hops exceeding 80 miles in distance who have ping times less than 15 ms. Let me tell you that the ping times from my server 4.5 miles away from the tower are, most of the lines, less than 10ms and some few lines =16ms. So if times are really low, why throughput it's still so low? Remember I'll be doing some tests this weekend and keep you posted. |
|
anoclon
join:2004-10-12
Guatemala
| reply to pointrow
The tower is only 33 feet high; grids are of course on top and radios are in the middle of the tower with 19 feet Belden 9913 cable up to the grids. With the same 19 feet belden 9913 I've tested the wireless pci adapter in the computer with better results. I've never tested de dlink radio dwl-2000ap+ past the 1/2 mile until now. What I've tested before is only the wireless pci DWL-G520+ card inside the computer even at a distance of 8 miles and downloading an mp3 file from my server at that distance took only seconds. |
|
pointrow
Premium
join:2003-09-23
Poplar Bluff, MO
| reply to anoclon
just keep changing channels till u find the right one. i dont think a switch will change things at all. might just be radio problems.
question, is this radio on the ground with coax running up the tower?
did it ever go faster past the 1/2 mile mark before? |
|
anoclon
join:2004-10-12
Guatemala
| reply to superdog
Is there any interference that may cause problems on channel 11?
not that I know, but I'm gonna try using another channel (6) and see if it works better on it. I'll do it this coming weekend and I'll keep you all posted. I'll also try switching both radios and using a switch instead of a cross over cable and connect the computer directly to the switch.
what u failed to tell us is the speed u get at the main pop. and how are u relaying signal to the tower in question from the main pop. do u have good speed when pluged in on the ground at the link in question?
Yes, I have a really good speed and throughput in clients within a 1/2 a mile of my pop using the same dlink radio. A videoclip or mp3 file goes from my server to any client just in seconds! so, I know I really have a good network. |
|
VariableARK
join:2003-03-17
USA | reply to anoclon
try swaping out radios, thats how i knew it was my radios, because when i changed radios to a different brand i had full bandwidth. |
|
pointrow
Premium
join:2003-09-23
Poplar Bluff, MO
| reply to anoclon
i think the design of the network has more to do with speed than the radios themselves. if u know what i mean. not that certain radios wont go faster than others. poor design will lead to poor performance.
what u failed to tell us is the speed u get at the main pop. and how are u relaying signal to the tower in question from the main pop. do u have good speed when pluged in on the ground at the link in question? |
|
VariableARK
join:2003-03-17
USA
| reply to anoclon
i had a similar issue, i was using cb3+ deluxe(.g) radios, and i would only get 300k throughput on a one mile link with no interference, and 24dbi to a 16dbi directional, couldnt figure it out either, finally decided the radios suck, maybe its your case too. |
|
superdog
I Need A Drink
Premium,MVM
join:2001-07-13
Lebanon, PA
| reply to anoclon
said by anoclon : I'm using now the wireless pci card from my computer connected directly to 1 of the grid antennas where the radio to receiving the signal was. Guess what, connected this way, the computer downloaded the same file in a couple of minutes!! This means the throughput increased a lot. Why is this happening? Well, there are a lot of reasons as to why this is happening, but my guess is the hops it has to make. Each hop will inject a little bit of latency, but not enough to make a 9Mb file down load in half the time as You mentioned. Is there any interference that may cause problems on channel 11?
--
»www.wavecrazy.net |
|
anoclon
join:2004-10-12
Guatemala
| reply to anoclon
Talking about throughput, maybe I´m setting up radios the wrong way. I´m not using radios in "bridge mode". What I´m doing is this: in my base station I´m using 1 radio set up as "access point". 4.5 miles away, I´m receiving the signal with another radio set up as "wireless client". These 2 radios, of course are in the same channel (#1) and using vertical polarization. Now, the 3rd radio is setup as "access point" in channel #11 and is connected to the 2nd radio by a crossover cable and rebroadcasting but in an horizontal polarization. Now, I come with my computer just a few meters away from where both radios are, and start downloading a file from my server. it took about 12-15 minutes to download the 9mb file. Now, here comes the 2nd scenario: forget about these 2 radios, I'm using now the wireless pci card from my computer connected directly to 1 of the grid antennas where the radio to receiving the signal was. Guess what, connected this way, the computer downloaded the same file in a couple of minutes!! This means the throughput increased a lot. Why is this happening? |
|
superdog
I Need A Drink
Premium,MVM
join:2001-07-13
Lebanon, PA
| reply to anoclon
said by anoclon :That´s what I don´t understand!! Maybe you can explain me why if I have a good signal strength and a good signal quality, and with this I mean a good Link, why is it matter the kind of radio I use? What´s the important? What if a 100dBi grid antenna exist? wouldn´t be enough to point and face 2 of this 100dbi antennas against each other in a 4 mile link with a D-Link radio? Well, To start with, power is not always the answer, and I personally would rather have two 30Mw radios talking at 5 miles with 2 24dBi grids and a really high receive sensitivity(Which I have done without a problem-2Mb thruput in a high noise environment), than I would 2 radios at 200Mw and low gain antennas, as the amount of noise they will pick up is crazy at the higher power level. It really is all about engineering?. The only reason that a lot of OEM gear makers have come out with 500mW Ap's is because we will always have a mindset that bigger is better, and in the radio world, it is just not true!. It has been said here over and over again that antenna gain is preferred over radio gain any day of the week. When I read posts that say they can't get a link with a 15dBi omni and a 1 watt amp with a 200Mw client radio 3 miles away, I almost lose my mind!. I am not yelling at anyone, I am just trying to make a point, and the point is this:
Use the best radio that You can, with the lowest power possible for a stable link. If You can get away with a 30Mw radio and a 26dBi dish, then do it!. It will work 10 times better than a 200Mw radio with a poor receive sensitivity and a 19dBi grid any day of the week.:)
--
»www.wavecrazy.net |
|
VariableARK
join:2003-03-17
USA | reply to anoclon
and different radios use different chipsets, one chipset may be faster than another, just like AMD is faster than intel (i had to rip on intel there). radios are just designed differently. |
|
