Mounting nanostations in 360* Ring

[laser3]

Does any one know the spacing distance for mounting nanostations in a 360* ring the way Motorola canopy's are done? can they be as close as the canopy's are monted?
and will it have an effect on radio performance? If I mounted on top of a 20 storie tower block in 360*, I think the nano's have 60* H-beam so that 6 nano's does any one know what radius I should be able to cover with LOS and zero Fresnal. assuming the whole setup with either nano2's or nano5's what would the coverage radius be? 8) also is there a piece of hardware to mount them in a Rin config?

[/img] http://www.motorola.com/staticfiles/Business/Products/Wireless%20Broadband%20Networks/Point%20to%20Multi-point%20Networks/Canopy%20Products/Canopy%20Advantage/_Images/_Static%20files/Canopy_Advantage_MD_US-EN.jpg [/img]

[pjm]

This may help...
http://forum.ubnt.com/forum/viewtopic.php?t=5636&highlight=spacing
Found it reading other places in the forum.

[kijoma]

hi,

We certainly would not do that, the signal fed to the neighbours receiver will be very high in transmit, you have intermodulation issues, desensitisation etc..

It may be better if you could stagger at least alternate radios vertically but to be honest I do not hold much faith for performance or integrity of that layout without some serious high order receiver design (not an integrated chip!) and/or hi quality channel filtering.

if you can seperate the nano's around the top of a building or mast so there is some reasonable isolation (metal/brick/concrete in the way) then it will work as expected.

remember 60 degrees is the 3dB point (half power), there is still a lot of power coming out at 180 degrees if the neighbour receiver is so close

that's my zero pence worth :)

[WHT]

http://forum.ubnt.com/forum/viewtopic.php?t=5636&highlight=spacing
Spacing of back to back (not referring to the side by side co-phased array) at 1.9 Gig cellular is on the order of ten foot plus. And keep in mind they have cavity filters AND the radios are well shielded from each other in RF partitions at the base of the tower.

At 2.4 and 5.8, it would be closer, but you have to deal with near-field emissions (not so much from the radio engine, but saturation of the receiver front-end). That means we're getting into a lot of unknowns.

I'd settle for, withstanding further testing, a minimum of three foot spacing of back to back antennas.

As for vertical separation, you get quite a bit of isolation at ten wavelengths. Which would be about 47 inches at 2.4 Gig and 19 inches at 5.6 Gig.
That's my estimate based on cellular systems and the Moto Can-O-Pee costs several hundreds of dollars more.

[laser3]

Thanks guys this was really helpful it would have been great to have them clustered like the canopy. it would have a much smaller footprint on top of buildings an towers. that ring in the photo that the Nano's are attached to where could I go about finding one of those. I wounder if the Ubiquiti guys have done any testing of there own? and lets say I was using a mesh firmware like Robin on the NS's and they were repeating the same signal would I be able to cluster them any close with out seriously affecting performance


I sometimes have my mesh repeters very close when tesing in the same room....or in the next room and the still seem to work fine, sometines they are right next to each other, but I guess the radios could be colliding, I may have to test this scenario http://www.telecomanddata.com/nsarrayfa8.jpg[/img]

[guidonet]

I experimented with 3 NS2 separated by an aluminum baffle plate and mounted in a pipe of 2.5 inches and another group similar to 2 feet of distance but under the pipe. I obtained very good result, but the truth is that it is but economic a MTk with one omni of 15dB… and so far is but administrable.

[WHT]

laser3...

While you may not notice any interference between units close to each other in the lab, there may be sufficient interference when trying to recover a signal 40 or 50 dB lower from a remote client.

Corrected link - BTW, that wasn't my picture, someone else posted it and can't find it to give them credit.

http://www.telecomanddata.com/nsarrayfa8.jpg

Here is a good picture,

http://forum.ubnt.com/forum/viewtopic.php?t=2783&highlight=cluster

http://web.aanet.com.au/drewaustralia/NB5%20Array%20Tower.jpg

[laser3]

Hey WHT

What If the radios are clustered and repeating the same signal , like a mesh network would the signal be affected then. The reason I asked is because most people use their Nanos as individual radios talking to other individuals over the AIR but the Mesh is like a Choir where all member sing together or the same song. so would the same not apply if I was using my nanos with a mesh frimware like Robin? and all the nanos clustered to make a 360* signal pattern like an omni antenna would this eliminate the Multi-signal path issues if i have just a single signal?

[kijoma]

hi,

if they all transmitted at exactly the same time and received likewise then all would be well. but this can easily be accomplished with.... an omni :)

[guidonet]

That´s what I´m trying to say
but this can easily be accomplished with.... an omni
...and cheapest too

[kijoma]

hi,

i think the issue is people think by having say 6 14dB aerials in a ring they somehow have 6 times 14dB! . the only possible advantages are each radio can run at the regulatory limit so you have effectively 6 times the output power and of course you share the CPU load.

in reality though the laws of physics get you as far as crosstalk/intermod/blocking and it all goes horribly wrong.

[brasileottanta]

hi,

if they all transmitted at exactly the same time and received likewise then all would be well. but this can easily be accomplished with.... an omni :)

or like similar to Syncmesh of Skypilot ..... but is a syncronus modulation ..