[Canberrauav] radio test results

Andrew Tridgell tridge at samba.org
Sun Dec 4 20:28:30 EST 2011


We got some great results testing Johns cloverleaf antennas today!

What we tested was:

 - ubiquity 5.8GHz bullet radios
 - 20dBi patch antenna at RedHill
 - 2 cloverleaf and 2 'rubber duck' omni antennas at Mt Ainslie
 - tested the cloverleaf with and without Amitays amplifier

The range of the test is approximately 8km, which is more than enough
for the OBC competition.

All of the tests were performed with the same setup at RedHill. Jack
kept the 20dBi patch antenna pointed at Mt Ainslie. Both radios were set
to minimum bandwidth (a bit over 1MBit) for maximum range.

  http://photos.tridgell.net/v/CanberraUAV/RadioTesting/IMG_20111204_180302.jpg.html

At the Mt Ainslie end, John, Brett and Tony operated the other radio and
swapped between plane antennas. They had the Axis IP camera attached to
the radio so Jack and I were able to watch a video stream from the
camera at the other end of the link. The video test allowed us to judge
the quality of the network link quite well.

We monitored the signal strength at each end using a script talking to
the SNMP server on each radio.

For the tests with the amplifier enabled, we configured the radio at Mt
Ainslie to assume 20dB TX gain from the amplifier, plus 1dB of cable
loss. We assumed the cloverleaf antennas had a notional gain of
3dB. Thus the radio was configured to transmit at 14dBm, keeping the
EIRP at 36dBm (ie. 4W).

---------------------------
test1: 4 element cloverleaf

amplifier was enabled at Mt Ainslie

antenna in 'normal orientation':
   - sig level at Mt Ainslie: -82 to -84dB
   - sig level at Red Hill:   -79dB
   - good video quality

antenna at 90 degrees roll:
   - lost approx 2dB of signal

antenna at 45 degrees roll:
   - lost approx 3dB of signal

antenna pointing directly away from RedHill
   - lost approx 8dB of signal, link became unstable, very poor video


---------------------------
test2: 3 element cloverleaf

amplifier was enabled at Mt Ainslie

antenna in 'normal orientation':
   - sig level at Mt Ainslie: -83dB
   - sig level at Red Hill:   -81dB
   - good video quality

antenna at 90 degrees roll:
   - lost approx 1dB of signal

antenna at 45 degrees roll:
   - lost approx 1dB of signal

antenna pointing directly away from RedHill
   - lost approx 4dB of signal, link still good




---------------------------
test3: 5dBi 'round' omni

amplifier was enabled at Mt Ainslie

antenna in 'normal orientation':
   - sig level at Mt Ainslie: -79dB
   - sig level at Red Hill:   -72dB
   - good video quality

antenna at 90 degrees roll:
   - lost approx 9dB of signal at RedHill
   - lost link completely at Mt Ainslie

antenna at 45 degrees roll:
   - lost approx 9dB of signal at Mt Ainslie
   - lost approx 9dB of signal at RedHill
   - poor video

antenna pointing directly away from RedHill
   - lost approx 13dB of signal at RedHill
   - lost link completely at Mt Ainslie


---------------------------
test4: 5dBi 'flat' omni

amplifier was enabled at Mt Ainslie

antenna in 'normal orientation':
   - sig level at Mt Ainslie: -86dB
   - sig level at Red Hill:   -80dB
   - OK video quality

(antenna roll not tested)

---------------------------
test5: 3 element cloverleaf

amplifier was disabled and removed at Mt Ainslie. Radio at Mt Ainslie
was reconfigured for TX power of 27dBm (max available on radio)

antenna in 'normal orientation':
   - sig level at Mt Ainslie: -86dB
   - sig level at Red Hill:   -87dB
   - patchy video quality

antenna at 90 degrees roll:
   - kept link, but very poor video quality

   
The clear winner is the 3 element cloverleaf. While it doesn't have as
much gain in the optimal orientation as the round omni does, it is much
more robust when we have roll in the plane. We still need the amplifier
for video link at 8km.

Next steps:

   - find out where we can get a couple of these 3 element cloverleaf
     antennas
   - build a housing for the cloverleaf so it doesn't get damaged/bent (at
     least in transport)
   - the the 900MHz cloverleaf as well (we didn't test it as we didn't
     have a SMA -> RP-SMA adapter)
   - see if we can source a 5.8GHz amplifier without the heavy inline
     power supply. Amitay, can you tell us the make/model of this
     amplifier?
   - build a better antenna mount for the 20dBi patch

Thanks to everyone who helped today!

Cheers, Tridge



More information about the Canberrauav mailing list