|
Written by djhurt1
|
|
Monday, 02 October 2006 |
"'ve asked this question in the past and got good answers but thought I'd throw it out again just for knowledge purposes. I've encountered several situations where a 12db flat panel will by far out perform a 15dbi grid antenna. Now this could be due to manufactuerer quality but there really is'nt much inside a passive antenna from what I've seen. One instance tonight was clear LOS except a single power line about 50' off across the road and the panel did much better. I've seen examples of the radiation pattern but wanted to hear more details why in general, a flat panel out performs? I should also mentione that this is off an sector AP antenna. I figured you people would be the best to ask!"
milbrath:
My guess is wider beamwidth. A panel would be more forgiving than a grid.
djhurt1:
I considered this, but we spent some extra time at our most distant client today just toying with it. I informed her prior that we'd like to try some things and she was just as curious as we were if it would work etc. We had the 15dbi grid and the 12dbi panel and tweaked each ones settings ever so meticulously and the panel just blew the grid away on both RX and TX. I forgot to mention that with the panel, we set the power to be right at 36db and did not change this once we put the higher gain grid up and the panel still produced a stronger RX at the AP.
SipSizurp:
To add some spice to your findings, it would be interesting to know the specific manufacturer/models of the antennas you tested, and whether or not the feed lines and pigtails were the same for each antenna.
Lligetfa:
You didn't mention polarity. My guess is the panel is more polarity agnostic while the grid needs to be more precise. I heard that a few degrees off on rotation can really drop off on a grid.
When a sector is down-tilted, the further out from centre you go, the greater the polarity rotational skew.
Also, the panel can make more of multipath signals than can a grid. Keep also in mind that the Rx levels reported are just raw signal strength not reflective of signal quality, so multipath signal that does not help the link will still be seen in the Rx strength.
I think a better head-to-head shootout would be to compare grid to grid and panel to panel between two towers where they are straight shots with little multi-path.
Original thread location |
Sample Networks 
