2.4 vs 5.8 NLOS link question

[TimS]

Can anyone out here advise me o the following, we have a 3-mile 2.4 ptp link setup with a pair of 24 db pacwireless grid antenas, 300 mw radios (qnode mesh on miniITX). Its punching through several stands o tall trees and a lot of clutter from buiuldings and so on. I'm able to pull a little under 3 Mbs consistenly from this setup. But the client wants more, so we are considdering 5.8 starOS with atheros 400mw. I dontthink the 5.8 will work through allthe trees so I'm thinking bout bouncing the 5.8 off a building tha can be clearly seen from either side of the link, its actually the local airport control tower and the angles on the tower look like they are in the right position to reflect the bouince. Has anyone out here ever don this kind of thing sucesfully at 5.8 or i it just pipe dream?

[lonnie]

Get some 3 foot dishes from PacWireless and use horizontal polarization. I think you'll get 20 mbps at that distance with 5.x GHz in a NLOS situation if you get nearly full throughput from standard B mode and likely in vertical polarization.

[mp3turbo]

nobody on this planet can answer this question - you have to try it.

[go.fast]

If you do get this link to work bouncing off the airport tower, PLEEEEASE report back here and gives us the details.
Thanks
:)

George

[robertm]

I am doing a similar link using 23db 5.8 panels and CM9. I am losing the signal output over only 40 feet of LMR 400, I can only get an association if I hook the panel directly to the pigtail, Senao 2511 is working at -59db over same cable, what could be up?
Does the CM9 have a limit on cable length, or should I be looking somewhere else for the problem ?
I will be mocking up the cable lengths on the floor in shop tomorrow and will re-post results, but if anyone can comment in the meantime it could save me a lot of heartache, I really want these units to work, I just bought a bunch of them
r :-<

[lonnie]

Try the other antenna connector. It sounds like you are using the wrong one.