Realistic max range for PS5 PTP link?

[mohave_steve]

I have a backhaul link I need to complete and am considering a pair of the PS5-22s. The link is clear LOS between two mountaintop sites 22.5 miles apart. My path calcs tell me I should be able to establish link at 24mbps with a 14db fade margin. I have never attempted a link of this distance on 5.8 before and I have not yet used the PS5-22.

Are my path calcs realistic? Have any of you run a link at this distance without the use of external antennas?

Thanks

Steve

[WHT]

14 dBm fade margin is about right for what you are describing.

However in practice, fade margins are not a linear progression, rather a proportional progression. The longer your distance, the MUCH MORE greater the fade, not just a proportional (to the distance) greater.

You might predict a 14 dB fade for a half mile path and never experience more than a few dB of fade, but when it comes to a long shot like what you are doing you could very well see 20 dB of more fade instead of the predicted 14 dB fade.

You may want to consider a dual radio solution like what Mikrotik uses. At the headend transmitter radio, use the maximum certified antenna gain of 22 dBm and for the receiver radio, use a 32 or 36 dBi solid dish antenna. Likewise at the tailend.

[mohave_steve]

When looking at that option I may be as well served with going to 3.5ghz.
The added cost & complexity of the dual radio would offset the added expense of a pair of routerboards & XR3 radios. The reduced path loss at 3.5 vs 5.8 wolud improve my fade margin.

Question: Would it be enough? What would by your goal for fade margin on a 23 mile link?

[WHT]

Not really, it will only get you about four dB more gain.

The free space loss at 3.8 Gig is identical as 5.8 Gig (the inverse squre law is not frequency or wavelength dependent), but the total path or link loss is about four times less...because the 3.8 Gig antenna is about twice the physical size and therefore four time the square area available to intercept the incoming radio wave.

[WHT]

For a 23 mile link, I personally wouldn't go below a 26 dB fade margin...but that's just me. Others might be comfortable with 20 dB.

Where 20 dB is about ten minutes a day annual outage, then add another 6 dB to the margin due to the distance. For 10 mile link, I'd be comfortable with a 23 dB fade margin.

[mohave_steve]

Hmmm....

I had allways assumed that path loss had a direct corelation to frequency.

So: the internal 22db antenna on the TX side and a 32.5db on the receive side will give me 25.5db fade margin.

I am not clear on your service level statement "Where 20 dB is about ten minutes a day annual outage" ? Do you mean 10 minutes (average) every day? 3650 minutes a year? 60+ hours?

Thanks

Steve

[WHT]

Path loss does have a correlation to frequency, its that free-space loss doesn't . Where path loss includes free-space loss AND receiving antenna aperture.

Yup...the 32 dBi dish will add 10 dB more to your signal.

William Barnet with AT&T Labs came up with a formula almost forty years ago for reliability metrics

This is a snapshot of a spreadsheet I use for System Operating Margin (SOM) with a -83 dBm receiver sensitivity. You can see where I get the 10 minutes per day number.

But in reality, fade is more a seasonal thing and the chart can be misleading in that a 20 dB fade margin does not mean you will always have ten minutes per day every day of the year...rather closer to half an hour in the morning after sun up and half an hour after sun down for several days in the spring and fall.

http://telecomanddata.com/WiFi_SOM.png

[mohave_steve]

Where path loss includes free-space loss AND receiving antenna aperture

I understand. Thank you for explaining. Makes sense now. I had been under the impression that free space loss changed with frequency.

The results from your spreadsheet are very enlightening as well. It looks to me that the numbers would suggest that a 28db SOM would be a worthwile target for my backhaul. Perhaps a difficult one to hit but obviously the "10-15db fade margin" goal is not nearly high enough.

Thanks again.

[mohave_steve]

WHT,

Seeing where you are from you may get a kick out of this:

In mid 2000 I approached a gentleman who ran a successful ISP in Lubbock. I spoke with him about my idea for using the new 802.11b gear that was becoming available to provide Internet service to outlying areas that had no broadband access. I had put a great deal of time and effort into making my case. I really believed it was a great idea. He nearly fell on the floor laughing at me. He made it quite clear that it was a rediculous idea that would never work.

So I guess we are all wasting our time as this wireless stuff will never work!! ':D'

[Erick]

I would strongly suggest that you use mikrotik for your link. RB433 with XR5 plus 26 dbi single feed parabolic dish. We run 63km p2p links with this configuration. Its stable with 26 mbps throughput. Don,t get me wrong as we use ps5 for cpe's for our clients as its easy to setup and online support is easy if something go wrong and we get 30 km p2mp links that is quite stable. We got bullet 5's and 32 dbi grid antennas today and we are going to try to get a 100km link up and running this weekend for a backup link to a remote site. Good luck with your link.

[WHT]

We got bullet 5's and 32 dbi grid antennas today and we are going to try to get a 100km link up and running

We would love to run a PS5 with external 32 dBi antenna, but since we have to deal with the FCC Rules & Regulations...we can't. The same situation with the Bullet.

[rconaway]

WHT,
if you drop some dues into my political campaign, I'll make sure nobody in Chicago touches you.

signed,

Blagojevich
(still the Governor)

[Erick]

[quote="WHT"]We got bullet 5's and 32 dbi grid antennas today and we are going to try to get a 100km link up and running

Where we go no one has even attempted any wireless setup before. So no interferance of any kind. Do you know how it feels to scan in 5 or 2.4 ranges and you do not even find one AP of any kind. I know some other dude tried it but couldnt get the link working and his comment was that if he cant make it happen nobody can. So i must be "nobody". XR5 cards rock and with duel feed 32 dbi parabolic dish you get stable and good throughput. The bullet id going to be a backup link if something would fail.

[rconaway]

I have tested the NS2's for 10 miles across vegas using 32dBi antennas and got 24Mbps connections. This was to hold us over while we had a pair of Motorola radios repaired. Worked great.

[mohave_steve]

10 mile link, 32db antennas, -86 RX sens.....

Let's see if Steve has figured the math out?

Max legal TX w/ 32db antennas on PTP=21dbm

Assume feedline & connector loss 2db each end.

Fade Margin @ 2.4ghz=42db

It would appear to me that you could be connecting @54 w/-74 RX sens and still maintain a 30db fade margin??

[rconaway]

It's directly across the center of Vegas which is not exactly the lowest noise point in the U.S. with several WISP vendors with 5.8GHz solutions and quite a few fairly tall buildings. I'm also running a 20MHz channel.

[WHT]

Lets' see if Justin has figured the math out?

NS2 at both ends for a 10 mile link at 2.4 Gig at 24 Meg throughput, where 14 dBi external antenna is the highest gain antenna allowed by FCC rules (as it is equal or less than the radio's certification).

NS2 TPO +26 dBm -4 dB conn. loss +14 dBi antenna -124 dB free space loss +14 dBi antenna -4 dB conn. loss +86dBm receiver = +8 dBm

8 dB fade margin over the -86 dBm receiver sensitivity = 90% reliability (about 1½ hours per day down time)

[rconaway]

I think he is saying that I should have done better at 26dBm. Actually, we ran them at 22dBm and figuring cable loss, I'm guessing my output was about 48dBm.