Hi all,
For P2P links, can people recommend antennas that would work well at up to 70km ? Anyone who has tried it, that is? (theory is easy: reality differs)
Customer prefers 5.8GHz, and the towers proposed are not very high, so 5.8 might be better than 2.4 considering fresnel zone, antenna beamwidth - hough 2.4 is still an option.
What realistic bandwidth can you get through such links?
Would really appreciate comments from people who have tried such things …
Regards
CableFree Solutions
Hi Stephen,
I seem to remember someone posting about a 90 km link he had done here in the forum. I think it was somewhere in South America, from one mountain top to the other.
Apart from that, no personal experience from me with such distances (although I’d love to try one 
).
Best regards,
Christian Meis
Well, 3m dishes from both sides, SR5 and if you get 6 mbit/s you should be very happy…
We haven’t done one quite that long, but the best “antenna of last resort” we have found is the Gabriel high-performance (drum style) quick fire series. Runs about $3000-$4000 per 6ft dish, gets you about 37dBi of very clean gain, weighs about 300lbs, and they are almost invurnerable to noise.
At 70km, with SR5s right on the back of the antenna, and VERY careful aiming; I would expect these to link at at least 12Mbps, perhaps 18 or 24. Higher rates may be possable, but I wouldn’t count on more than 12 (~8Mbps real throughput).
–Eric
I would not think you should have a problem with a 70km link.
We have tested 60km with some equipment in the workshop.
WRAP with SR5 Cards each end
Solid 1.2M Dish, 26dBi antenna with 1/2M run LMR-400 coax
19dB 2.4GHz Parabolic reflector, 26dbI antenna with 1/2M run LMR400 coax
Signal strengh, around -78 to -85 dB rx each end.
Managed 8MB/s half duplex.
Not to bad for assembling a few bits a peices laying around.
Thank you very much everybody, very helpful and much appreciated.
Will pitch these suggestions in and see what happens …
Regards
Stephen
This is our experience in long MT links
90 Km, Atheros, 2.4 Ghz 24 dBi Grid:
rx-current: 4.4Mbps
rx-10-second-average: 3.7Mbps
rx-total-average: 3.7Mbps
lost-packets: 0
random-data: no
direction: receive
rx-size: 1500
status: running
duration: 4s
rx-current: 3.8Mbps
rx-10-second-average: 3.8Mbps
rx-total-average: 3.8Mbps
lost-packets: 0
random-data: no
direction: receive
rx-size: 1500
53 Km, Atheros, 5.8 Ghz 29 dBi Parabolic:
[admin@REICO-POAS] > /too bandwidth-test 172.23.0.134
status: running
duration: 11s
rx-current: 15.2Mbps
rx-10-second-average: 9.4Mbps
rx-total-average: 9.4Mbps
lost-packets: 0
random-data: no
direction: receive
rx-size: 1500
62 Km, Atheros, 2.4 Ghz 24 dBi grid with nstreme dual:
tx-current: 2.3Mbps
tx-10-second-average: 2.4Mbps
tx-total-average: 2.4Mbps
rx-current: 2.6Mbps
rx-10-second-average: 2.2Mbps
rx-total-average: 2.2Mbps
lost-packets: 0
random-data: no
direction: both
tx-size: 1500
rx-size: 1500
wow, nice
90km single faster than 63km dual? hmmm… sounds like interference or something else, but this is not “normal” 
were these 2.4GHz links with g, g-turbo or just normal plain b mode?
Both in 2.4-b
Yes, 90 Km is more efficient that 63 Km. The 90 Km link have better fresnel zone.
I saw that the test were only running for 4s and 11s. It may take longer for the bandwidth test to get to the full speed of the link. Also, you should add any ‘normal’ traffic that is already going over the link.
Nstreme might speed this up if you are not already using it. If the signal is good, you should get 20Mb/s plus.
You might also try the new 2.9.12/13 with the 5 or 10MHz channels. It concentrates the power on less channels and should help get a better signal over the distance.
John
well , if its thats easy to make a link of 90Km , so why i cant make it with a 7KM !!! i’m totaly lost with this distance , it’s ok for an hour or two then it’s gone , after that i need to restart my d-link 2100 AP to get back the connection , and what about the fresnal zone ? is it so important ? what if the height of one of the 2 points are not equal to the 2nd point height ? or its lower than the fresnal zone ? what about amplifires ? i’m using a 1W with the d-link2100AP , is it wrong ??? please i need HELP !!
Hello samsoft08
First of all, most in the wireless industry will say that WiFi amplifiers will do more harm than not. The worst thing about them is they amplify not only the output signal, but all so the noise being received by your WIFI radio device. I have not needed to use them to date so can not comment to much.
What type of antenna are you using?
The type of coax and the lenght?
Did you make them yourself?
Have you done a scan in both areas for other WiFi repeaters etc? You may have some interferance.
What are you current signal strenghts?
Matt
Please can you tell us the TX output power in db that you are using?
Thank you!
Some of our links at 65 km have throughput of 50 Mbit/s tcp.. The main thing is “The bigger the better”, but we don’t use antenas greater than 2 meters. You have to have very low loss cables, a good feed. This link has the same throughput with CM9, put we’ve upgrated put SR5 just to be sure that the link will be stable in the winter
Yup but the bigger (antenna) the higher db. I am asking you this because in some countries there are output power limitations. So I was wondering how much dbs you are using (antennas + wireless card). For example 28db antennas (at 5ghz) are common but if the output tx is over 3 or 4 dbs (considering some cables and connectors attenuations) we are over the 30dbm which is the legal maximum limit. So that’s why I am asking
this is the answer :
a- point 1 :
1- Outdoor AP 1W transmision power ,2.4Ghz .( wireless interactive ).
2- 15 dbi omni antenna.
3- LMR400 less than 1 m cable.
4- 45m height .
b- point 2 :
1- D-link 2100 AP. 2.4Ghz
2- 1 W outdoor amplifire.
3- 24dbi grid antenna
4- cables are LMR400 less than 1 m .
the distance is 7Km near line of site .. its two points inside a city crowded with 2.4Ghz signals , which i’m using .. i’m using MT as a gateway , DHCP , and a web proxy
Hi there
With that setup you shouldn’t have much problem, while i haven’t done that many long distance setups though, the last I did was a 4km this weekend and that worked real sweet.. could even reduce output power to fit inside regulatory demands anyway, back to your issue…
Since you are using the 2.4GHz band… and since you are in a city (which I presume is full of things like microwave owens, wireless phones, hearing aid circuits and such, all on the same band) have you looked at the radio spectrum to see if there is any interference?
On my MT boxes I would just start the windows interface and hit a button to see the “load” on each channel… not sure if any of your boxes has that kind of capability.
/Jörgen
well , be sure that in my city ( Baghdad ) there is 100’s of wirless clinets and wisp , some of the wisp’s using power up to 4W !!! i’m sure there is an interference , the question is how to overcome this problem ?? i cannot say its a bad link , coz its good for hours before hang up !!! restarting the d-link solves the problem ..
Hmmm… any special time of day you loose connection?
When you loose the link, is there any kind of debug/log/message/status information on the DLink unit stating why it lost the link? Normally it would just reestablish the link in case of disturbances so it sounds strange that you have to restart it…
What kind of signal do you get at each end? With that kind of output and antennas you should be somewhere around… -50 dBm, is that the number you get? If not you have a problem somewhere in your setup, here’s a very good linkbudget calculator http://www.antennspecialisten.se/se/support/WDT_linkbudget.html (requires Java)
Since a reboot cures the problem it sort of sound like there might be a problem with that unit… do you have another unit available that you can test with? Have you tried running without the external amp?
Is there any dispaly in your units that can show the load on various channels? Sometimes ou can solve a problem by moving to another channel.
Best regards
/Jörgen