Industry Forums > Wireless Service Providers > UBNT 900mhz

UBNT 900mhz

Nope, Only one per tower site here...

In the UBNT forums i've seen a few topics of people having major issues with 900 mhz co-location with 3 antennas... The answer was wait for GPS which is a long ways off...

reply to wirelessdog
Even with shields 3 antennas per tower is awful. We are running 8Mhz channels and using channel shifting trying to get everything far apart, but it still has issues.

reply to wirelessdog
Not trying to be a dick here... Honestly not...

But I screwed the pooch when I said that Ubi would not have this 900 MHz stuff working by June of last year, didn't I?

I'm surprised it has taken this long and I'm expecting that Justin will square me away...

And I won't mind if he does.
--
Once we IPv6 enable every device on the Internet we will have toasters, baby monitors, and security cameras joining the bot nets which today are populated only by idiots that can not refrain from clicking, "Yes I would like to see those titties..."

reply to wirelessdog
The 900mhz works fine, but yea, one AP per tower.

reply to raytaylor

Justin... absolutely no offense intended...

No vendor gives a shit about us.. not in this forum.

They just want us to talk up their crap and sell it for them.

That ain't your fault mate. Naw, it is their fault.

And you are not some dumb shit kicker... the vendor is...
--
Once we IPv6 enable every device on the Internet we will have toasters, baby monitors, and security cameras joining the bot nets which today are populated only by idiots that can not refrain from clicking, "Yes I would like to see those titties..."

But I do think Canopy's way of doing it with a single GPS module and handing off the timing to the APs is less CPU intensive than having each AP process the GPS sync.

reply to wirelessdog
Ubiquiti says they have GPS working in 5GHz but it still isn't working in 2.4GHz very well. Since 900MHz is simply down-converted 2.4GHz radios, when that gets fixed, 900MHz gets fixed. When it does, the right way to deploy it is four 5MHz channels unless someone else comes out with a true 120 degree dual polarity antenna. I'm kind of waiting for someone to come out with a 900MHz dual-polarity omni first.

reply to wirelessdog
900 is downconverted 2.4?? Last I was told by Ubnt it was 5.8???

Yea its downcoverted 2.4 afaik.

I can sometimes see a 900mhz AP on a 2.4ghz scan if its really really close or a 900mhz card is in the same box as a 2.4ghz card
--
OptionsDSL Wireless Internet
»www.optionsdsl.ca

reply to voxframe
I know the XR9 and SR9 were down-converted from 2.4. I believe the M900 stuff is as well. I think the 3.65 stuff is down-converted from 5.8.

reply to wirelessdog
AHH YES that was it sorry!

Has anyone had trouble co-locating 5.8 and 3.65 stuff?

reply to wirelessdog
We've struggled with the 3 Sector problem for nearly a year. Shields help a little. We took 3 Loco's and ran through Cavity Filters and that worked however its single chain stuff. We actually found that turning the Transmit power DOWN on the Rockets helps too as it keeps the AP's from being overdriven by the adjacent AP and therefore helps it hear the CPE in the field better. Of course the RSSI on the customer side is lowered and so customers in a noisy environment are affected. We erected a 190' tower last week with Shields on the sectors and the antennas vertically separated by 5 - 7 feet as well. We're hoping the combination of shields, front to back and the umbrella affect of vertical separation will allow us to operate at full power. We're waiting on electricity to the sight however and that may take a couple of months.

I just put up a tower with 3 Rocket M900's, 120 degree 13dBi sectors, and the rfarmor shields. The sectors are mounted on H-mounts that put the sectors about 30 inches away from the tower. They have about 4 or more feet horizontal spacing between the edges of the shields. So far it is working.

So far I have 4 CPE's online. One for each sector + one poor signal (trees, etc) that I am playing with to see what it takes to improve it. One of the other CPE's is at 3.25 miles (5.23 km) with a 13 dBi Yagi (-70 dB) and the internal antenna (-81 dB) which gets about 800k upload and more than 1 Mbps download (have to remove traffic shaping and retest tomorrow).

It may get worse as more CPE's are added. I will find out soon enough.

Book, were your problem sectors mounted directly on the tower or offset like mine?

As for the poor signal install, it is currently using a 13 dBi Yagi. I can only see 3 options (antenna choices) for improvement. UBNT's dual-polarity 16 dBi Yagi, Lanbowan's 12 dBi dual-polarity panel, or the Nanobridge NBM9 dual-polarity dish.

The Yagi may suffer from wet snow.
The panel may be better but harder to source.
The NBM9 may be the best option.

Ours were mounted directly to the tower as the tremendous wind load and twisting leverage the shields added concerned us.

reply to bburley
As far as CPE's go... The dual polarity yagi is supposedly getting amazing results from a few posts I gathered in the UBNT forums. Keep in mind that two come in a box so adjust the 200$ price accordingly.

I would highly recommend NOT using the NBM900 I have had very disappointing results in two seperate cases. I saw 1-2 db gain over the locoM900 at best. When UBNT released these they were listed as 13dbi gain antennas, and thats what most vendors have them listed as. BUT sometime in the last few months they quietly reduced the gain on their spec sheet to 10-11dbi. Not really worth it when the loco is 8dbi.... Go with the Yagis and you will not be disappointed.