Higher gain antenna resulting in inferior signal strength

[LontzroV]

Hello,

Our WiFi network is intended to cover large outdoor spaces. I understand the best way to increase coverage is to deploy more APs, but in our case the number of clients is relatively low and the requirements for throughput too. So I'm wanting to upgrade the existing access points with higher gain antennas to black holes between access points. I got 4 Alfa AOA-2458-79-AM antennas and I've tested them on an MR76 and on a MR86 AP. In both situations the signal levels with the Alfa antenna are worse than with the MA-ANT-20, and that's despite the Alfa being 7/9 dbi whilst the MA-ANT-20 is 4/7 dbi gain. I'm testing on a level plane with the antennas, so the decreased vertical beamwith should not be the issue. What am I missing? Does anyone have similar experience or found a way to boost the signal of MR access points?

[LontzroV]

To add more context: I'm mainly interested in boosting the 5GHz band, so as not to cause more CCI on the 2.4 GHZ band. I've created a SSID specific for testing the antennas, that is only available on the APs that test the antennas. I've also created a RF Profile that disables all SSIDs on the 2.4GHz band, and sets the 5GHz band to a TX power of 10 dbm. In the logs I can't see any changes to the TX power, so the AP does not seem to throttle the TZ power with the higher gain antennas present - or at least there are no logs of that happening. These antennas seem to be pretty high quality. To my knowledge Alfa is a reputable vendor, and each antenna shipped with a printout of its tested RF patterns (unique to each antenna, slight variations). I'm dumbfounded as to why the signal is worse, not better.

[UKDanJones]

If you want better coverage id always go for directional antennas rather than omnidirectional. I’d suggest you’re seeing the effect of the signal being squeezed ’flatter’ by the higher gain and you’re possibly not in the optimal position to get the benefit of that higher gain. 

did you use modelling software to validate the design? I’ve seen some pretty funky ‘omni’ antennas that definitely don’t distribute the signal evenly in 3D and they act more like a 360 directional antenna firing signal away from the horizontal plane. 

[UKDanJones]

This antenna looks like a nightmare… I dislike high gain omnidirectional antennas.

Could you not use directional ones?

[UKDanJones]

Something like this?

 

[rhbirkelund]

If I read that radiation pattern correctly, and the manufacturer has oriented is correctly, the only way you get any sort of gain is but mounting it directly perpendicular to the user. That is, if your antenna is located 2 m above the floor, your device also needs to be 2 meters above the floor.

You have almost 10 dB in the 90 degree angle, and 0 dB everywhere else.

Imo, you are better off with the MA-ANT-20.

[LontzroV]

Thanks for the input! @rhbirkelund I'm testing the signal exactly perpendicular to the antennas. Even then they're slightly worse than ANT-20. Maybe they're just junk after all. @UKDanJones I'm not sure how I can cover 20+ acres with directional antennas effectively. To cover all areas I'd need a few spots with many antennas pointing in 45 degree offset directions, requiring multiple APs to service the antennas. In that case I might as well spread the APs themselves out (at least according to my limited understanding). But I can appreciate how the signal pattern is much more predictable. Maybe I'll get one just to experiment with it.

[cmr]

@LontzroV - 20 acres!  I'm trying to work out how to cover the bits I need of 9 acres and that needs a fair few APs.  Out of interest how many were you thinking of needing?  I think I need 5 for just a strip of the 9 acre plot.

[LontzroV]

Yep, that's the pain. And that is just 1 of 4 sites (it is greenhouses and open growing spaces). We actually have 6 units on that property now (you can imagine the inconsistent experience), but before I started in this company there were just two...  I have no words. When I explain how far off we still are from an actual robust coverage, the numbers are always the impossible hurdle. We'd need at least 12 get things properly covered. So I was hoping to find a way to get more signal out of the existing APs, since the number of clients and throughput requirements are quite low. But it's not to be I guess.

[PhilipDAth · Accepted solution]

You can't get more signal out - you can only control how you spread the signal around.

An omni antenna spreads the signal in every direction.

A directional antenna sends all of the signal in one direction - so that signal will be stronger in that one direction.  The narrowed that signal beam the further it will travel.

[LontzroV]

I'm aware. Even if client and AP have to be within half a meter of height at 50m distance, if I get farther coverage I'm game for a flatter signal pattern (from a higher gain antenna). Hence I'm testing these. But it seems they just aren't working as they should. I posted here to see if I'm missing something obvious, but it seems you all know what I know, and surely these particular antennas aren't a way forward.

Thank you all for engaging though!