Testing a 900Mhz Canopy system. I have a 10db , “Polarization Vertical” Omni on the AP and 700 feet away I have a SM ( factory Maxrad 10db panel antenna ).
I just noticed as I walked by today that the MaxRad was oriented Horizontal so I rotate it 90d ( Vertical is at the top now ).
To rotate all I did was remove the screw at the bottom and rotate it, so the panel is still pointed / aligned exactly as it was pre-rotate. And at 700ft away LOS I see it is pointing right at the Omni.
Now, my Jitter is all over the place.
Omni Vert - Panel Horizontal
RSSI - 2200 (-51db)
Jitter - 3
Pings to AP - 5ms %99
Omni Vert - Panel Vert
RSSI - 2762 (-38db)
Jitter - 3 - 12
pings to AP - from 3ms to 17ms
Shouldn’t jitter be better / the same now that both are Vertical ? Am I blinding the SM now ? Possibly the AP ?
At 700ft should I lower the power on the SM ?
Yes, -38 is WAY too hot. You were doing yourself a favor with the alternate polarization. In fact, I’d say -51 was still way too hot. The SM can see as much power as you want, but it’s the AP that will be overwhelmed by an overpowered SM. Definitely lower the power in the SM, and if this is more than a test site, (in other words will continue to be operational with other clients on the system, then you may want to switch 90 degrees AND turn it down.
Jerry has written the conclusive tutorial on balancing your SMs. Read it here.
Hey, Moto… Can we sticky that post or something? Took me 10 minutes to find it…
cvs wrote: Yes, -38 is WAY too hot. You were doing yourself a favor with the alternate polarization. In fact, I'd say -51 was still way too hot. The SM can see as much power as you want, but it's the AP that will be overwhelmed by an overpowered SM. Definitely lower the power in the SM, and if this is more than a test site, (in other words will continue to be operational with other clients on the system, then you may want to switch 90 degrees AND turn it down.
Jerry has written the conclusive tutorial on balancing your SMs. Read it here.
Hey, Moto... Can we sticky that post or something? Took me 10 minutes to find it....
Hmm, I read the post you linked and I understand the concept of why you don't want an SM blasting the AP but there isn't really anything on how to control it.
Turning down the power starts causing other problems before the DBs get to the 70s.
With the SM Panel left Vertical (same as AP) I can drop power down to it's lowest setting (3 , if I set it any lower it just defaults back to 3) and the SM is hitting the AP at about 60db and my jitter is all over the place ( 2 - 12 ).
So I set the SM Panel back to Horizontal ( AP is Vertical) and with the power still at 3 the SM is hitting the AP at about 70db with Jitter jumping back and forth between 4 and 11 , 4 then 11, 4 then 11.
The best I can do is have the SM panel Horizontal to the APs Vertical and set the SM power to about 10. At this point it is hitting the AP at about 60db and jitter is around 2 with jumps to 11 ever 10 refreshes or so.
If I turn the power back up to 26 then the SM (still horizontal to the APs vert) then I hit the AP at about 50db but my jitter is rock solid at 2 or 3.
I try pointing the SM away from the AP until the DB drops to 70 then my Jitter goes back up around 11. This is regardless of Horizontal or Vertical.
Is the SM in 2X mode? If so, a jitter of 9 or 10 is perfectly acceptable.
Depending on the speed you are providing, 2X may not be needed. Switching to 1X will make make the jitter far more stable.
Turn the power back up and have the backside of the antenna facing the AP to lower the power.
I still have not been able to see any difference in performance by balancing the power levels.
mattmann72 wrote: Turn the power back up and have the backside of the antenna facing the AP to lower the power.
I still have not been able to see any difference in performance by balancing the power levels.
I don't know enough about RF to argue one way or another. Initially my understanding was that the AP controls when the SMs can broadcast and uses time slots to prevent them from all talking at once. Combined with my impression that it also knows how far away the SMs are and also uses that information to assign said time slots. So my logical conclusion was that one SM hitting the AP like a Super Nova wouldn't blot out a SM with it's little Bic Lighter since they would not be broadcasting at the same time.
Now, the information in Jerry's post mentions an effect similar to a bright light hitting a humans eyes and your eyes needing a few secs to re-adjust before it can see lesser light (of course this is taken down to milliseconds for an AP/SM). If that effect is a reality then it makes sense that one radio could blind an AP for a millisecond or whatever and cause problems for the SM with the Bic lighter.
Since dropping the power on the SM backhaul I haven't really had time (and it's COLD outside !) to do much testing and see if it made any difference in how well the other SM can communicate with the AP. I really need to find a spot were my test SM is at the extreme end of almost not working while the backhaul SM is blasting the AP and see if dropping the power or even turning off the backhaul SM makes any difference in the test SMs connection.
Then again I'm testing with just 2 SMs and the effect may not show it's self unless there are a lot more SMs vying for those time slots and/or firing much more rapidly from a lot more points.
I’m no expert, but I suspect that it’s possible for interference to cause high jitter even while packet loss is low. Does anyone have a spooky story about the mysterious ways of interference? In your case, perhaps changing the polarization has exposed your link to interference that wasn’t there before, even though the polarization was wrong. Of course, the obvious question is what could interfere with such a short link. Whatever the case, I’d be interested to see how the spectrum looks like for both polarizations.
Tommy wrote: I'm no expert, but I suspect that it's possible for interference to cause high jitter even while packet loss is low. Does anyone have a spooky story about the mysterious ways of interference? In your case, perhaps changing the polarization has exposed your link to interference that wasn't there before, even though the polarization was wrong. Of course, the obvious question is what could interfere with such a short link. Whatever the case, I'd be interested to see how the spectrum looks like for both polarizations.
Spectrum looks pretty much the same the same Vert or Horizontal with almost everything in the 90's. Numbers are also pretty much the same at both the AP and SM location.
900.0 -95 -95
901.0 -93 -93
902.0 -94 -91
903.0 -95 -94
904.0 -95 -94
905.0 -94 -94
906.0 -93 -92
907.0 -96 -94
908.0 -95 -92
909.0 -95 -95
910.0 -93 -93
911.0 -94 -92
912.0 -94 -93
913.0 -94 -94
914.0 -93 -91
915.0 -95 -94
916.0 -93 -91
917.0 -93 -93
918.0 -90 -90
919.0 -90 -83
920.0 -89 -89
922.0 -91 -89
923.0 -90 -82
924.0 -93 -92
925.0 -96 -91
928.0 -95 -94
929.0 -97 -92
930.0 -97 -94
I had never noticed until now that 921 , 926 and 927 are missing. I don't remember seeing anything in the manual that explains the reason for that.
<jealous>
Wow, that’s clean!
</jealous>
8)