dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
22207
share rss forum feed


IntraLink
Premium,MVM
join:2002-08-14
Uath Valley
reply to viperm

Re: Installed first UBNT Air fiber unit today

We have two installed, but 2.2 miles and 3.4 miles.

So I've heard that you need to use the low power setting for real close in shots like that.

And the alignment is really focused. And more so close up if that makes sense.

Also make sure they are not tilted at all since the thing has two antennas per side.

When I was aligning I couldn't get one side to get the same power reading, but then I did something to the mount on the Rohn tower and I think it 'straightened' up a bit and then it really changed to good power/alignment.

These things are also a bit finiky in these initial firmware versions, sometimes we've had to reboot them a few times to get them to really lock in.

But they seem to run well for us for a couple months now that they are online and have some time on them.

Waiting for newer firmware for management and features etc. It will come (right?)


Semaphore
Premium
join:2003-11-18
101010
kudos:1
said by IntraLink:

Also make sure they are not tilted at all since the thing has two antennas per side.

When I was aligning I couldn't get one side to get the same power reading, but then I did something to the mount on the Rohn tower and I think it 'straightened' up a bit and then it really changed to good power/alignment.

Agreed. With dual polarity a difference of only a couple of degrees from 'plumb' on each end can make a 20db difference in cross pole isolation on a link. The dual stacked antenna setup would make any side/side or front/back tilt a huge issue I would think.


viperm
Carpe Diem
Premium
join:2002-07-09
Winchester, CA
Hmmmmm okay as we.have some uptilt and downtilt on opposite sides. Your saying to make both sides completly level and keep them level even though there is a small height difference? I thinks that's where we started and had nothing?

I have no problem starting the process again to better learn the nuances of these units..

LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1
I think he meant that if the mount is not plumb, that you can end up with some rotation. Up and down tilt is to be expected but side to side, not.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey


IntraLink
Premium,MVM
join:2002-08-14
Uath Valley
reply to viperm
Yeah, rotation, not tilt, sorry!

Watch for rotation. Tilt is fine of course!


viperm
Carpe Diem
Premium
join:2002-07-09
Winchester, CA
reply to viperm
The master might and I do mean might have a slight skew to it but maybe a 1/2 to 1 degree..

We will know next week when we go back out..


Semaphore
Premium
join:2003-11-18
101010
kudos:1
reply to viperm
Sorry. I don't have an AirFiber unit... but when installing Dual pol Solid dishes the side to side should be perfectly plumb or you'll loose cross polarity isolation really fast - way faster than I ever realized - and the front/back should be exactly parallel on each side. E.G. if the East side is plus 3 degrees tilt the West side is minus 3 degrees.
I use an "inclinometer", (which is just a 36" electric level with degree measurements), vertically on the face of the parabolic for the front/back tilt, and that silly little bubble level on the back of the dish for side/side. Radio Mobile gets the inclination pretty well perfect on every shot so far.

Thinking that the same ideas will apply to a dual antenna device and probably more so (?)


viperm
Carpe Diem
Premium
join:2002-07-09
Winchester, CA
reply to viperm
Okay all is well. We re-aligned the master side the skew was off about 2* once we got it plumb and did our up down and left right on both sides we hit - 41.

We turned down rx gain on the radios and it knocked it down to about -55 on each side and thruput is now 743 megs both rx and tx. Not sure how much that rx gain option is supposed to attenuate does anyone know?

I figured maybe 10db?

So being very plumb is crucial thanks everyone for your input!!!! Customer now thinks he should be getting more since they are gigabit radios. I had to remind him that's aggragate not each way, and referenced the specs on the sheet..
--
»www.accelwireless.com
ComTrain Certified Tower Climber.
Wireless and IT consultant.
Proficient in Mikrotik

Newbie

join:2011-04-18
Thanks for the update!

Oh if he thinks he should be getting 1gig symmetrical maybe this thread will peak some interest?

»forum.ubnt.com/showthread.php?t=58841

But i'm sure if you give him that he'll need more......


viperm
Carpe Diem
Premium
join:2002-07-09
Winchester, CA

1 edit

1 recommendation

Click for full size
Click for full size
Click for full size
Picture

BBJunkie

join:2012-10-02
Montgomery, IL
reply to viperm
Any thoughts on how far it will can really go? they claim like 13 KM but I wonder if that is true.


battleop

join:2005-09-28
00000
reply to viperm
Some people have an uncontrollable desire to respond to a post no matter how useless their response is. Even if it asking questions that were clearly answered in the original post.
--
I do not, have not, and will not work for AT&T/Comcast/Verizon/Charter or similar sized company.


treichhart

join:2006-12-12
reply to BBJunkie
No 13Km is only in desert type climate area if your area is like rain area look only get maybe 1 to 2 miles with airfiber.

some people on ubnt forum are getting 3-6 miles are getting around 200-700meg but not the full 1+gig as ubnt claim that airfiber would do on 13km setup.


viperm
Carpe Diem
Premium
join:2002-07-09
Winchester, CA
reply to viperm
Well back to crap performance again. RX side dropped and now we have latecny and missing packets!! ARGHHHH signal is still rock steady at -49 to -52 at less then 700 feet with RX gain set to low on the slave side.

UBNT swears its alignment but we have been there 6 times now and cant get any better. Then they told us oh your on older firmware there is a bug in that one that reports incorrect RSSI etc. So we spent two hours testing that and sure enough did not work still doing the same issue which appears to happen more when the evening starts?

Any way to test for interference? I mean come on at less than 700 feet even if it was some noise there is nothing between master and slave? Slave is pointing up to the master from the movie theatre, no cell towers nearby and its pointed to houses?

I am at a complete loss so now back up there a 7th time tomorrow morning. I am about ready to pull this POS! and put back something else like a rocket. That one was rock steady but not enough thruput for this guys 300 meg fiber connection.

Maybe I can set it to NO full duplex and try that so my throughput would drop to what 750 megs aggragate? 375 up and down? but then I could at least determine if its that particular channel? What also sucks is I have to travel to each side since inband managment does not pass management traffic to the slave side allowing me to log into both radios at the same time.

Any thoughts anyone?

Thanks

--
»www.accelwireless.com
ComTrain Certified Tower Climber.
Wireless and IT consultant.
Proficient in Mikrotik

prairiesky

join:2008-12-08
canada
kudos:2
reply to viperm
at least setup a temp 2.4 link so you dont have to run back and forth


viperm
Carpe Diem
Premium
join:2002-07-09
Winchester, CA
reply to viperm
I would but there is only one set of cables and they are in us by the AF unit.... The master side is on a rental home and the owner. Wants no exposed cables temp or permanent... The tenant gets kind of touchy when we are there even though its not their house they just rent. But then again I kind of understand...

jcremin

join:2009-12-22
Siren, WI
kudos:3
At that distance and clear LOS can you just throw a wireless router inside at the far end and connect to it from the roof on the other side?

OHSrob

join:2011-06-08
reply to viperm
Is your issue just intermittent by chance ?

VCWireless

join:2010-11-17
Valley Center, CA
reply to viperm
I wouldn't waste your time by worrying about interference, with the narrow beam width you can co-locate several of theses on the same tower.

I would say its one of the following...

1. possible bad unit
2. reflection from the commercial roof ledge
3. Alignment - I know you don't want to hear this but it took myself and my most seasoned employee about 1 hour to align a 2 mile link. I couldn't imagine a 700ft link.

How long is the cable run? How low did you turn the power down?

prairiesky

join:2008-12-08
canada
kudos:2
reply to viperm
can you tape a laser onto the horn and align it at night?


viperm
Carpe Diem
Premium
join:2002-07-09
Winchester, CA
reply to viperm
Click for full size
Okay so we got it working again down to -45 RX set to low and power set to 20 per UBNT support.

We spent a good two hours again back and forth up and down finally nailed it so we think its going to stick this time.. Xing fingers..


IntraLink
Premium,MVM
join:2002-08-14
Uath Valley

1 recommendation

reply to viperm
Sooooo, it WAS an alignment issue after all?


viperm
Carpe Diem
Premium
join:2002-07-09
Winchester, CA
I would have to say it appears soo? as much as I hate to say it yes.. touchy little buggers!

We had it to where it was way to hot last time so we attenuated and everything was good. Went back this time again and had to realign again so not sure what happend. I find it hard to believe they moved any litle bit. We will eave it as a learnign experience..