[HamWAN PSDR] More sector antenna results, and high gain antenna battle heats up!

[Bart Kus]

Hello,

Did some more antenna measurements tonight.  The Ubiquiti sectors were 
the new contenders for the "official sector" title.

The 5G-16-120 
<https://www.hamwan.org/t/tiki-index.php?page=Ubiquiti+5G-16-120&structure=HamWAN> 
didn't do so well.  Gain is dreadful and the radiation pattern is nearly 
omnidirectional.  Blah!

The 5G-19-120 
<https://www.hamwan.org/t/tiki-index.php?page=Ubiquiti+5G-19-120&structure=HamWAN> 
did a lot better.  It looks to be a 90 degree sector though, if we're 
honest.  -3dB points were measured just 81 degrees apart.  The +/- 60 
degree points look to be -8dB, not the -6dB I've come to expect from 
some manufacturer beamwidth claims.  The radiation pattern is also a bit 
weird in its shape, but not awful.  The frequency response is pretty 
good though.  The final drawback of this sector is price.  It's about 
$150 a piece.  Although that is cheaper than the Laird sector which is 
$180 a piece.  This Ubiquiti sector is also dual polarity, so if we ever 
do MIMO, antennas won't have to change.

More contenders have arrived though!  I'll get them tested soon. 
Hopefully they blow all these expensive sectors out of the water. :)

Now, on the "official high gain antenna" front, I've measured two new 
contenders as well.

The ARC-DA5834SD1 
<https://www.hamwan.org/t/tiki-index.php?page=ARC+ARC-DA5834SD1&structure=HamWAN> 
does not disappoint.  It is well constructed and weighs less than you'd 
think.  The frequency response is the flattest of ANY of the antennas 
tested so far.  The radiation pattern is THE most directional of any of 
the antennas tested so far as well.  Just look at that gorgeous 
rejection, all at least -30dB down.  This is the #1 choice if you've got 
the $300 (shipped) to spend on it.  It is also a dual polarity antenna 
with a fair amount of isolation between the polarities.  On the rear are 
N-connector to waveguide adapters. They combine to feed one solid 
circular waveguide that protrudes from the center of the dish.  At the 
end of the waveguide structure the metal turns to plastic and the waves 
are allowed to radiate through the plastic shield/support onto the 
subreflector at the very tip.  This subreflector has a really cool wavey 
pattern milled into it on the underneath to collimate all the waves 
properly to/from the main reflector dish.  I found the N connectors 
somewhat hard to start threading for whatever reason.  Was not getting 
alignment. The mounting system is well thought out and a pleasure to 
use, especially to adjust the tilt of the antenna!  You just turn one 
nut to raise/lower the tilt.  To rotate you can loosen both U-bolts. 
Now, normally this would make the antenna slide down.  But they're 
thought of that and provide you with a stopper you install on the mast 
before hand.  The weight of the antenna rotates on this base. When 
angled right, you just tighten the antenna's u-bolts.  There's also a 
built-in mini-mast section to bolt a radio to.  Very well thought out.  
A very nice antenna.  But still, $300 shipped from Streakwave isn't 
cheap.  WLANmall looks to have it 
<http://www.wlanmall.com/arc-34dbi-dual-pol-4-94-5875ghz-dp-dish.html> 
for much less shipping.  Streakwave tried to jack the shipping on this 
antenna by another $50 *after* I placed the order.  It would have been 
$350 if I let them have their way.

The Poynting K-GRID-003-06 
<https://www.hamwan.org/t/tiki-index.php?page=Poynting+K-GRID-003-06&structure=HamWAN>, 
henceforth known as "Die Antwoord 
<http://en.wikipedia.org/wiki/Die_Antwoord>", is a surprisingly well 
performing high gain antenna.  It is designed and manufactured in South 
Africa 
<https://maps.google.com/maps?hl=en&ll=-26.10403,28.079563&spn=0.012689,0.009495&sll=-26.103982,28.079655&layer=c&cid=11252272406233210136&cbp=13,3.3,,0,2.79&panoid=inPysFz8CF29sKnrYx1ziw&gl=US&t=h&z=17&cbll=-26.104026,28.079467>, 
just like its musical namesake.  It may indeed be "the answer" for 
HamWAN's directional antenna needs.  While I do not have official 
trustworthy absolute gain comparison figures yet for any of the 
antennas, the preliminary numbers make it look like Die Antwoord gets 
about 5dB over its similar-in-size cousin, the Laird GD5W-28P 
<https://www.hamwan.org/t/tiki-index.php?page=Laird+GD5W-28P&structure=HamWAN>.  
It also looks like the $300 dish is only 1dB better.  Die Antwoord also 
features a much flatter frequency response, and a far better radiation 
pattern than the Laird.  You know you're doing something right when the 
manufacturer's published H-plane pattern 
<http://cdn.entelectonline.co.za/wm-43814-cmsimages/A-K-GD-003-06%20Datasheet.pdf> 
strongly resembles your own results.  Notice the distinctive side-lobes 
touching -20dB on either side of the main lobe.  Notice the containment 
of all other lobes within the -30dB limits.  This is a FINE antenna!  
And for not lying about the patterns like so many others do, Poynting 
gets my "fine manufacturer" seal of approval.  Poynting and I disagree 
on the frequency response, but the difference is 1dB, so I'll let it go. 
:)  The gain is really quite flat.  A surprising achievement for such a 
broadband antenna.  In fact, the gain flatness is only bested by the 
$300 ARC dish mentioned in the previous paragraph.  What does all this 
wonderful performance cost?  Would you believe a mere $95 shipped 
<http://www.titanwirelessonline.com/Poynting-5-8GHz-31dBi-Grid-p/at-30gr-58.htm>?  
NOTE: Do *NOT* follow the advice of the one gentleman who commented on 
this product at the link I just gave you.  The man is a fool who ruined 
his antenna.  :)  The radiating element is meant to be unconventionally 
oriented.  Follow the assembly instructions.

The physical size difference between the $95 Die Antwoord (bottom) and 
the $300 ARC dish (top) is small.  Here they are both enjoying my couch, 
just yearning to RF @ each other:



To truly measure the real long distance gain of these antennas, I've got 
to test them over a longer range.  Something like 100m should be enough 
to make the EM waves mostly perpendicular to the DUT antennas instead of 
hitting them at weird angles.  This straightening of geometry should 
allow for real-world gain measurements and comparison of absolute gain 
performance.  A 3ft dish on a 20ft range may not get the chance to focus 
correctly, and it may be losing gain because of this.  I've taken steps 
so that I can take the measurement lab mobile and deploy in a large field.

I also screwed up the measurement process enough times that I finally 
sat down and documented it, step-by-step, at the bottom of the Antenna 
Analysis 
<https://www.hamwan.org/t/tiki-index.php?page=Antenna+Analysis&structure=HamWAN> 
page.  Who could possibly screw up executing a 57 step sequence from 
memory? :)

Finally, I also had a discussion with a microwave antenna measuring 
friend of mine, and figured out how to measure the absolute gain of each 
antenna.  This is different from measuring "comparison of absolute 
gain", aka "absolute gain delta between antennas".  This would be 
deriving that actual dBi you'd print on a box of a product and you'd 
have to stand by it.  I'm not sure if the process is worth it though, 
since it's only the deltas which lead to antenna selection decisions.  
We'll see.

Now, where are those new sectors ...

--Bart

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.hamwan.net/pipermail/psdr/attachments/20130217/edb390de/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: jebbbbhb.png
Type: image/png
Size: 1909352 bytes
Desc: not available
URL: <http://mail.hamwan.net/pipermail/psdr/attachments/20130217/edb390de/attachment.png>