Bullet Distance Questions (Power, Fog, B/A, Antennas)

[cajoel]

I have a straight shot line of sight of about 3 miles that I want to cover using a Bullet and some grid antennas.

It's over a residential area that's loaded with b/g wireless
(downtown San Francisco)

It frequently (20% of the time) may have some fog in the way.

I'm currently looking at some grid antennas from Hyperlink:
http://www.hyperlinktech.com/productfamily.aspx?id=109

So here are my questions:
- Should I wait for the Higher Power (400mW vs 100mW) versions?
- Would the 5Ghz A model be less interference prone than 2.4Ghz?
- Is fog going to be a problem?
- Grid vs Yagi? (grid seems most cost effective)
- Actual performance expectations?
(anyone done something similar in an urban environment?)
- Any recommendations for stand alone mounting poles?
(tripod, sandbag type?)

Thanks,

Joel

[MaximumISP]

I dont think power will be a big factor with high gain antennas
Fog can be a problem but I dont have much experiance with that
so I cant comment on it
I would choose a Grid over a yagi
performance unpredicatable till more is known
Tripod and mast is what I typically use

The biggest factor is going to be interferance
what would concern me most is whats out there I can only imagine whats already there in San Fran

I recommend you get a spectrum analyser and do some thorugh scanning
or at least put up a NS2 and NS5 temporarily and do some scanning in both HZ and verticle to get an indication of how bad things are spectrum wise
Keep in mind things like a motorolla canopy network wont be seen from anything but a spectrum analiser or by using the analyser built into a moto canopy SM but you would need a 2.4 a 5.2 a 5.4 and a 5.8 canopy sm to check everything properly

Its perhaps doable but you need to know exactly what your up against in this enviourment

[WHT]

- Should I wait for the Higher Power (400mW vs 100mW) versions?
As I posted in another thread:
For 2.4 Gig, we are limited to 36 dBm or 4 watts EIRP.

That means for a Point-to-multipoint omni antenna application, we can use the Powerstation full power output of 24 dBm and a 12 dB antenna for an EIRP of 36 dBm.

If we use the Bullert HP's full output power of 30 dBm or 1 watt, we are limited to using a 6 dBi omni antenna. That means we loose 6 dB of gain on the received signal side.

I've always maintained its better to achieve your output gain with the antenna as it also increases you received gain.

- Would the 5Ghz A model be less interference prone than 2.4Ghz?
Most certainly, but in a dense urban area, there may be a lot of 5.8 Gig backhaul pollution.

- Is fog going to be a problem?
No, but the atmospheric conditions associated with fog can cause severe refraction. You'll need at least a 20 dB fade margin.

- Grid vs Yagi? (grid seems most cost effective)
Each type have specific advantages and disadvantages, if....Ahem...IF the forum search engine as a wee tad bit better, I could find my post about antennas.

- Any recommendations for stand alone mounting poles?
All depends on what you can use...no way is better than the other, but some are worse than others, e.g. you would not want to mount a penetrating mount on a rolled rubber mastic roof.

[cajoel]

Unfortunately I don't have a NS2 and or NS5, or spectrum analyzer to do testing. I'm just a network hack, not a radio guru.

If the HP bullets aren't 2X the cost, then I hope to keep the radio and antenna budget under $120/site.

Maybe someone else in the bay area wants to lend me some gear?
Hint, hint... :)

WHT: I'm also confused about your spec of 1watt for the higher power model. Other threads have suggested ~400mw.

Regarding mountings, can anyone suggest a vendor for a tripod style mount kit?

[WHT]

WHT: I'm also confused about your spec of 1watt for the higher power model. Other threads have suggested ~400mw.
http://www.ubnt.com/products/bullet.php says "Zero-Variable Wireless Infrastructure Deployment, the Bullet features up to 1000mW TX power", but the 1000mW refers to the HP model. The PDF Data sheets list 15 to 20 dBm for the Bullet 2 and 17 to 22 dBm for the Bullet 5.

Tripods can be found here, http://www.channelmasterintl.com/mountaccess.html and here http://www.antennasystems.com/broadband/masts_tripods_towers.html and here http://www.dennysantennaservice.com/1079219.html

[HozicEmir]

hmm interesting is to hear that someone use more than 100mW on 2.4, in my experience over that power in multipoint one just generate noise and lose signal quality nothing else, and fog should not be problem if u use high gain antennas, we somethimes get even beter signal on wet wether becose beter reflection on all surfaces.

[UBNT-Mike.Ford]

Hello Hozi,

People use thousands upon thousands of our high power cards/cpe devices without causing self-interference.

Thank you,

Mike

[HozicEmir]

don't want to argue but U are wrong Mike.
I had in mtiks some 400mW ubnt mini pcis, on max str in 2.4 band are useless and over 20db, less power less noise, all cards had to be changed to standard r52(not r52H those are crap to).
Also on local forums is argued that is common with those cards in already noisy environment with many networks around.
Also I'm sure that in some large outdoor without other networks grate power is useful but this topic was about city area with probably very crowded air.

[UBNT-Mike.Ford]

Hello Hozic,

I will let the other WISPs chime in here.

Thanks,

Mike

[WHT]

The wost offenders are external amplifiers. Motorola CanO'Pee is pretty messy too.

But with any RF engine, including Atheos, the off-band emissions get messy when pushed. Yer better off using lower transmitter output power (TPO) and a higher gain antenna.

[rconaway]

I think the interference he was discussing was internal to the router with multiple cards. For example, Vivato units with dual radios will have approximately 20dB higher noise floors when each radio was on the same channel and I think 10dB more when on different channels (been too long to rememeber the exact numbers). Their panel which had 13 AP's in it, started having interference issues if the transmitters were higher than 30mw due to cross channel interference.

Back to the issue at hand. I would suggest short of renting a spectrum analyzer, your best bet is to purchase 2 NS5's, set them up, and get some real numbers. An alternative would be to get a multi-band PCMCIA card and one of the free wifi analyzer software packages (can't remember the free one since we use Air Magnet).

Anything else you will be just guessing at.

[wispwest]

you'll avoid ALOT of interference if you use your antenna's to propagate the signal HORIZONTAL. All routers, cordless phones, etc have a vertical polarization. I've dealt with 2.4 interference for many years, horizontal is the way to go when shooting over all that crap.