You are here

Link.ui.com compared to real results?

7 posts / 0 new
Last post
ng1p
ng1p's picture
Link.ui.com compared to real results?
I now have my first 2 perm mesh nodes setup in Maine (3rd one on the way). One at my house and one that I just installed on Oak Hill in Brunswick Maine at the top of a 250 foot tower. The signal strength is not even close to what link.ui.com shows. here is a shared url for the full details  ​https://link.ui.com/#p=4b9c07a2225f43d5a259ca2ef3fb49ed

Looking at this path link.ui.com shows that I should be at -70dBm for signal strength at both locations. Here is what I am currently seeing:

At Oak HIll with a Rocket M5 and AMO-5G13 13dBi omni at 250 feet I see -84dbm (11 dB SNR) output set to 27dBm (noise floor at -95dBm)
At my house 5 miles away with a Nanostation M5 at 60 feet I see a sig of -88dbm (7 db SNR) out put set at 27 dBm   (noise floor at -95dBm)

The LOS is clear including the fresnel zone. Why am I seeing such a difference from the predicted with link.ui.com and real results? And why are the two nodes reporting different levels (would assume they would be the same?)  Do I have a bad Rocket, Nanostation or AMO-5G13? I wonder if anyone can share and compare their results from like equipment and distances? I still need to do some more testing from other locations but this seems to be strange.

Also I will add that when I got the new node (Rocket) up and running yesterday on CH180 the new node had great -95dbm noise floor and was able to hear my home (Nanostation) node but my home (Nanostatoin) node was not able to the Rocket. I then switched to channel 171 (thinking perhaps the return loss on the devices at that frequency was too high?) On channel 171 both nodes could hear each other but it was on the edge. I then switched to channel 145 and left things on that channel as it was slightly better but not great as the numbers show above.

Any thoughts? I would be very interested it hearing thoughts on comparing link.ui.com vs. real results. I feel that I should be seeing much better signal levels. Am I wrong???

Thanks you Bill NG1P
kj6dzb
kj6dzb's picture
At 5 miles your lucky your
At 5 miles your lucky your picking up the signal at that distance.  I would have installed a sector on the tower and used a lhg hp5 to hit the tower. 

you might also consider that the UI  calculator doesn't account for the radiation pattern  resulting angular variation of the radiation lobs of antenna between 2 3D points in space. Rather the algorithm is using the flat plane peak to calculate path loss. (this is a problem happens in sound reinforcement modelings) the path loss is on a flat plane is not the same at 30deg of the center plane of path loss. The model may miss represent the coverage, in the off  parallel vector.

Think about this.... if the omni antenna is mounted level on the tower, and your qth is below the stie's elevation, then your station may be below the radiation plane of the Omni antenna.  

Do you have a photo for the tower? 


73 Mathison KJ6DZB

 
K6AH
K6AH's picture
Link budget calculations

Here's a spreadsheet I created to do link budget calculations for AREDN equipment.  Check out the Link Budget tab.

https://docs.google.com/spreadsheets/d/1GXzWaceQa5FZnfG9i1i9XwN98ToxnID4ysKHmD4RRoY/copy

Andre, K6AH

WU2S
WU2S's picture
AMO-5G13 Downtilt

Looking at the Ubiquiti datasheet for this antenna reveals that it has a builtin 2 degrees downtilt. Would that affect your predictions?
Ubiquiti Omni antennas datasheet

ng1p
ng1p's picture
Update
Thanks for all of your replies. I did some testing this afternoon by swapping out my nanostation with another one and saw the same results. So I moved the nanostation up higher from 60 feet to 65 feet and the signal level got better by 5 db (that's no bad 1 db per foot :)) My LOS must not have been as good as I though or I'm too low to see the 7 deg beam width.(this is why I asked about link.ui.com as it says I'm good) Yes the 2 deg down tilt helps also. The link.ui.com site accounts for the down tilt and beam width (i think???) of the AMO-5G13. Just to verify this if I test a path just a 1/4 mile away from the high node it shows over - 20 degs. Using the link.ui.com to my house at just over 5 miles away is shows -.84 (so less than 1 deg).. 

So in review by moving my nanostation up 5 feet:
Nanostation is now at -83 dBm and 12db SNR vs. 7 SNR before the move (5 db better)
Rocket is now at -79 and 16 dB SNR vs. 11 SNR before the move (5 db better)

My rule of thumb is to be 15 db SNR to account for fading due to rain and snow. So perhaps just a few move feet will do it. :)

Thanks would love to hear comparisons of link.ui.com to what you see in the field. I just want to know if I can trust what it says,


 
AE6XE
AE6XE's picture
802.11n devices are unable to
802.11n devices are unable to measure absolute or very accurate signal strength.   They don't need to, rather work to maximize SNR.  What they do is take a measure of the noise after a clear channel assessment, meaning no one else is transmitting.   The device then calls this -95dBm.   All received signal strengths are then a relative measure above this reference point.

Thus, if the node is at a site with higher levels of noise,  the SNR values will be proportionally lower.  RF shielding can help.

Joe AE6XE
ng1p
ng1p's picture
Interesting
I did not know thats how that works. Simple and effective for sure. Thanks Joe. Today i went up another 2 feet and got another 2 db for SNR better so i am seeing an average of about 14 db SNR with peaks to 15 db and a rare 16db peak. On the other end its at around 18 with a rare peak to 20 db. I cant go any higher on my end but this is a good signal level at this point.
73 Bill NG1P

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer