I’ve noticed that many people are using sectorized omnidirectional antenna arrays instead of a single onmi antenna. Here’s an example of one of the more popular one’s that I’ve seen used: http://www.hyperlinktech.com/web/sector … _array.php
Now, the question I have is how is this better than a single omni AP? Also, what kind of loss do you get from the splitter/combiner? Does this give you better range than the 60 degree integrated antennas? Better range than a single omni?
Thanks for your input!
Ok, here is my 2 cents on this, i tried a 5.8 4 way splitter to hook up 3 90 degree sectors that were rated at 15 dbi, this failed badly, the 10 dbi omni sector had more range by a few miles than this setup, i assume if you used 4 it would be worse, now i dont know what kind of loss happens but it is alot, i have a ap with the 90 sector now and its great, getting shots at over 10 miles with reflectors, so all in all i dont think that this solution can work unless you use a amp, and in my book that is just another thing to go wrong, the problem i have is my omni only can cover up to 6 miles reliably, and i cant afford 4 to 6 aps to cover a 360 cover, so it is kinda a game that has to be played, hope this helps, post if you have more questions
Funny, I just brought this up last week.
First thing that comes to mind is that the highest omni you can find is about 11dB - there might be some slightly higher versions but not much.
The combined sectors give you 17dB which helps you get closer to the max of 4W EIRP (Radio Tx power will have to be reduced a little). Even after splitting the signal 3 ways, you will still have higher EIRP per sector than with just the 11dB omni.
The 6dB of Rx sensitivity can really help pick up weaker SM’s.
The ability to down-tilt helps solve the issue of the signal overshooting customers that are close in.
The ability to scale the number of AP’s as the user base grows is attractive.
Not having to get approval to change the antenna system and incur costs to change from omni to the sectors
The only downside I see is that they are vertically polarized and only are available for 2.4GHz, but if that’s what you are running I see no downside.
The cost to do this with 3x120 H-pol is going to be 3x the cost of a single Omni but that would be the only downside. Just make sure you use a quality splitter.
Thanks for the input. Anyone else have similar problems or a success story?
My advice, omni antennas are only for very low noise enviroments, starting POP…
Jerry Richardson wrote: Funny, I just brought this up last week.
First thing that comes to mind is that the highest omni you can find is about 11dB - there might be some slightly higher versions but not much.
The combined sectors give you 17dB which helps you get closer to the max of 4W EIRP (Radio Tx power will have to be reduced a little). Even after splitting the signal 3 ways, you will still have higher EIRP per sector than with just the 11dB omni.
The 6dB of Rx sensitivity can really help pick up weaker SM's.
The ability to down-tilt helps solve the issue of the signal overshooting customers that are close in.
The ability to scale the number of AP's as the user base grows is attractive.
Not having to get approval to change the antenna system and incur costs to change from omni to the sectors
The only downside I see is that they are vertically polarized and only are available for 2.4GHz, but if that's what you are running I see no downside.
The cost to do this with 3x120 H-pol is going to be 3x the cost of a single Omni but that would be the only downside. Just make sure you use a quality splitter.
If you went with HPOL wouldn't you need to get HPOL converters for all the SMs?
2.4 is V-Pol
Right, but you mentioned “The cost to do this with 3x120 H-pol is going to be 3x the cost of a single Omni but that would be the only downside. Just make sure you use a quality splitter.” I wasn’t sure what you were getting at.
I see the confusion. I was writing that in a hurry.
The antennas that the link refers to are 2.4GHz V-Pol.
They don’t make a 900MHz version of the “sector omni”. It could be done with 900Mhz 120 sector antennas either V-Pol or H-Pol.
Thanks for the clarification.
Have you used any of these in 2.4 or 5 GHz?
Not with Canopy. 10 years ago we would combine multiple radios onto single antennas and run multiple antennas using combiner/splitters. It works fine.
If I needed to set up omni coverage on a starter pop I would absolutely run multiple 15-17dBi panels off one AP with the intent of adding a second, then a third AP as needed for capacity.
I think the reason I did not mention this earlier is that I have no need for it so it was not on my radar so to speak.
Good information, thank you.
Do you know exactly how much loss comes from the splitter? Does it take the power and divide by three?
Power is divided evenly to each leg (Tx -3dB) plus ~.5dB in connectors. The Rx gain stays the same less ~.5dB in connectors.
A Canopy 900 AP puts out 26dB. 26dB going into a 1x3 splitter will yield 26dB - 3.5 - 3.5 - 3.5 = 15.5dB of power (approx) going into each antenna. This leaves you needing to make up 20.5 dB to get to 36dB EIRP. The highest gain 120 sector antenna you will find is a MTI Wireless at 13.5dB which will get you to 29dB EIRP on each antenna (7dB short).
18dB yagis would need to be used at the customer side to increase the link gain and make up 6dB of that 7dB deficit. The SM Tx power would need to be lowered to 18dB so that EIRP is not over 36dB.
In all it would be cost effective way to build a PoP that can scale to hundreds of users without having to go back and change out antennas. When the PoP hits 80 users or so, then take out the splitter and install 2 AP’s on the other non-overlapping channels and walk away.
One side note: as previously mentioned, omni antennas are prone to interference. A 3x120 sector omni is just as prone to interference so if you don’t have a really low noise floor it could be problematic.
Keep in mind that most antennas over 2db are multi-element arrays.
So that 11db omni is a stack of 1/4th wave antenna elements.
The most critical issue with combining elememts to make an antenna array is the spacing and positioning of the individual elements.
The wavefront from each element can help shape the overall pattern by combining where you need power, and cancelling where you don’t need power.
There lies the problem. If the array is rigid, so that you cannot change the position of the elements, then I would believe that the design is reasonable to use. If the elements can be tilted in a manner that the top and bottom of each unit is not equal distances from the others, then I would expect you to start to form a difficult to predict daisy pattern in your signal. That means there will be places where the antenna gain seems hither than the rating, and places where there is almost no signal at all. In other words, the tilt capability will ruin the omni pattern.
Not to mention the extra losses from the various connectors caused by the splitter.
You are better off with 3 radio sectors feeding three antennas, OR a single stick omni.
I have seen several of these rigs, and having experience in building antennas, I look at them and wonder how they can get away with calling those omni antennas. And the prices they charge…
Jerry Richardson wrote: Power is divided evenly to each leg (Tx -3dB) plus ~.5dB in connectors. The Rx gain stays the same less ~.5dB in connectors.
A Canopy 900 AP puts out 26dB. 26dB going into a 1x3 splitter will yield 26dB - 3.5 - 3.5 - 3.5 = 15.5dB of power (approx) going into each antenna. This leaves you needing to make up 20.5 dB to get to 36dB EIRP. The highest gain 120 sector antenna you will find is a MTI Wireless at 13.5dB which will get you to 29dB EIRP on each antenna (7dB short).
18dB yagis would need to be used at the customer side to increase the link gain and make up 6dB of that 7dB deficit. The SM Tx power would need to be lowered to 18dB so that EIRP is not over 36dB.
In all it would be cost effective way to build a PoP that can scale to hundreds of users without having to go back and change out antennas. When the PoP hits 80 users or so, then take out the splitter and install 2 AP's on the other non-overlapping channels and walk away.
One side note: as previously mentioned, omni antennas are prone to interference. A 3x120 sector omni is just as prone to interference so if you don't have a really low noise floor it could be problematic.
What's the math for 2.4 and 5.2 Ghz?
This is just math for RF.
George is correct in that there are certain intangibles that are very difficult to predict.
If the signal is being divided by three, wouldn’t it be 26dB / 3 for each sector? You mention -3db, where did you get this figure?
RF is logarithmic. 1/2 the power is 3dB.