[TowerTalk] Cover antennas - the good news and the bad news

Sat Oct 31 09:15:47 EDT 2020

On 10/30/20 10:17 PM, David Gilbert wrote:
> 
> I would think that testing antennas today should be easier than it was 
> when the K7LXC/N0AX report was done, and I don't see why an antenna 
> range is even needed anymore.  A wideband noise source mounted on a 
> drone with an appropriately polarized short antenna could fly around the 
> antenna while a receiver on the ground connected to the antenna took 
> measurements.  Drones are very stable these days even in stiff winds, 
> and while I don't think GPS positioning would necessarily be accurate 
> enough to define near field location, various ways of doing so from the 
> ground (optical triangulation, tethers, etc) should be sufficient.  Far 
> field measurements out several wavelengths could most likely just rely 
> on the GPS coordinates.

It doesn't need to be a noise source - a CW tone works just fine.
or you can fly the receiver and radiate from the antenna.

This is a pretty standard technique these days - They use it for 
calibrating radio telescope arrays, for instance.

One would probably want to calibrate the probe's antennas.  Polarization 
purity is going to be a real challenge.  What you probably want to do is 
actually do three orthogonal "short dipoles" (or loops).  Or a "tripole" 
(three monopoles, with a somewhat more complex feed system), and then 
use an electronic switch (i.e. diodes) to switch among the three.

Then calibrate against something like a dipole and monopole, so you can 
get the antenna pattern of the "probe"

> 
> All of that assumes reciprocity, of course, but even if that was in 
> question I'd bet that it would be relatively simple and inexpensive to 
> design/build a small receiver for the drone that forwarded signal 
> strength data to the ground via an RF or optical link.

An RTL-SDR would do, or, if you want to fly the Tx, one of those SiLabs 
sources. It's a lot easier to fly the Tx, and just have it step through 
your test frequencies and antennas in a "recognizable" pattern.  For 
instance, do 1 second no antenna, 1 second axis A, 1.5 second axis B, 2 
seconds axis C, step to next frequency.

You'll need telemetry from the UAV that gives X,Y,Z position and 
attitude - but that usually comes down over the control link. Whether 
it's easily retrievable is another thing.

Let's talk required accuracies
GPS easily gives position within 10 meters in all three dimensions, and 
at a distance of 100 meters, that's about 6 degrees.  If you move out to 
1km, you're in "sub degree" territory.

Power measurement accuracy (assuming we're interested in relative 
measurements, not absolute)  is determined by:
1) How accurately do you know the distance (inverse square)
2) How stable is the Tx
3) How stable is the Rx
4) What's the SNR.

Let's assume that the SNR isn't limiting - it's easy to get 50 dB SNR  - 
Path loss for 1km for 10 MHz between isotropes is 52 dB. You're making a 
narrow band measurement (100 Hz?), and even a crummy receiver with a 34 
dB NF will have -120 dBm noise floor.  So radiating 1 mW (0dBM) will 
give you a 60-70 dB SNR, and 40-50 dB in the -20 dB nulls.  Yes, you're 
radiating from a "short dipole" and there's mismatch and all that, but 
if you had to radiate 10mW or 100mW that's not a deal breaker.

How stable is the Tx - this mostly a matter of power supply voltage and 
"mismatch" stability as the antenna moves. Let's say the power supply 
voltage is stable to 1% (easy with a 3 terminal regulator) - that's 
about 0.08 dB

How stable is the Rx - probably similar to the Tx, 1% voltage gain 
stability is plausible.

Distance uncertainty.. 10meters at 1km is 1%, so another 0.08 dB.

These are all random, so RSSing is suitable - and that comes up as 1.7% 
or around 0.2dB. for relative pattern measurements.

If you want absolute measurements, you'd need to either have a standard 
gain antenna (hard at HF) to compare against, or calibrate your Tx and 
Rx, which is a lot tougher.  Rx can be calibrated with a signal 
generator, but you have a mismatch issue (Rx input Z  is typically only 
"nominally 50 ohms" - you'd need to measure that somehow )
For Tx, getting an accurate "radiated power" is tough.

I've not looked at it, but you might be able to do something like they 
do for MIL-STD-461 Measurements - a "voltage probe" that is a 1 meter 
monopole or a bicone of similar dimensions.  You should be able to 
calibrate this to E-field levels.  Then measure your UAV transmitter 
against that.

The SNR won't be as good, but you can make longer measurements to drive 
the noise floor down.  A test antenna into a high Z receiver with 1 
meter effective height would be down 20 dB from a full size dipole for 
40m (heff=~10m), but if you can make 1 Hz bandwidth measurements, and 
maybe at 100 meter distance you can pick up 40 dB of SNR.

> 
> One big advantage of using the drone, of course, would be that the 
> measurements could also be taken at the actual operating site, thereby 
> including the effects of nearby structures.
> 
> I haven't seen any reports of anyone doing this yet, but I assume that 
> people already have

Pretty standard in the commercial and research worlds. A big 18 meter 
deployable boom for Europa Clipper is being tested this way, for instance.

"Calibration of the SKA-low antenna array using drones"
https://dial.uclouvain.be/memoire/ucl/fr/object/thesis%3A14813/datastream/PDF_01/view

"A new approach for in-situ antenna characterization, radome inspection 
and radar calibration, using an Unmanned Aircraft System (UAS)"
https://ieeexplore.ieee.org/document/7944287

UAV-Based Calibration for Polarimetric Phased Array Radar
http://www.caps.ou.edu/reu/reu16/finalpapers/Boyer-Paper.pdf
(see "Twitching Eye of Horus")

https://www.haystack.mit.edu/astronomy/astronomy-projects/edges-experiment-to-detect-the-global-eor-signature/
is using UAVs - there's probably something there in the hundreds of 
memos - look at the memo list link in the above.

Actually there's a fair amount of UAV stuff at Murchison Observatory in 
Australia.

Low frequency radio astronomers obsess about calibrating all the details 
out because the things they are looking for are deviations from the 
expected power spectral density.  So small variations in the 
"chromaticity" are important - Small, here, is parts in 1E4 or 1E6 - so 
things like slight mismatch causing ripple in the cable losses over 
frequency are important.

And then, those that are looking for "objects" are looking for faint 
sources near strong sources, and understanding the phase and amplitude 
of each antenna in the array, in every direction, is important, so you 
can do the "beam forming" implied in calculating the images.

For myself, I'm looking into calibration of large arrays of small 
satellites in space that cover 0.1 to 30 MHz (because that's blocked by 
the ionosphere, and can't be observed from Earth). My first array in 
space is looking at the Sun with 6 crossed dipoles, which is by far the 
brightest source in the sky, so my calibration doesn't have to be that good.

However, if you're looking for the remnants of the big bang, highly 
red-shifted down to HF and VHF from the original 1420 MHz emission 
frequency, you need to start worrying about measuring to 1 ppm kinds of 
uncertainties.

> 
> 73,
> Dave   AB7E
> 
> 
> 
> 
> On 10/30/2020 8:35 PM, chetmoore at cox.net wrote:
>> Not a FLAME. But You would likely benefit from reading the k7lxc N0ax  
>> tribander report. After raeading it I ordered the  C31XR.   Force 12 
>> is no more  but some of their antennas  are  mechanically improved and 
>> sold by JK and there are A lot newer tribanders  I would like to see 
>> them  test on their antenna range.  Hint,  tests of some of the mosely 
>> and hy gain antennas  did not fare all that well.  I had a TH6, a TH3 
>> , TH7 and  a classic ta33 all of which worked
>> Pretty well..........in their day.   My th-3 is still a good  FIELD 
>> DAY antenna  and great as a mult antenna to grab  south American 
>> mults  so I don’t have to  rotate the c31.
>>
>> N4fx
> 
> _______________________________________________
> 
> 
> 
> _______________________________________________
> TowerTalk mailing list
> TowerTalk at contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk