Jim Lux 2020/12/12 07:48
On 12/11/20 9:05 PM, C.M. wrote:
>
> [Edited Message Follows]
>
> I got a similar dilemma. I own quite a number of portable VNA's. I
> tried to measure a 433MHz rod monopole and here are the results:
>
> Tindie v2: 428.000MHz VSWR: 1.24
> Tindie v2 plus: 431.000MHz VSWR: 1.22
> Tindie v2 plus 4: 418.500MHz VSWR: 1.25
> N1201SA: 428.000MHz VSWR: 1.42
> Arinst VNA: 414.398MHz VSWR: 1.20
> NanoVNA-Fv2: 424.000MHz VSWR: 1.08
>
Welcome to the wild world of antenna metrology !
That's not a huge difference, converting to voltage reflection
coefficient magnitude
1.24 -> 0.107
1.22 -> 0.099
1.25 -> 0.111
1.42 -> 0.174
1.20 -> 0.091
1.08 -> 0.038
With the exception of the two noticeable outliers, you're looking at 5%
sort of variation, which isn't uncommon.
This is a rod monopole, you say? Is it 1/4 wavelength mounted on top of
a large (> 1 lambda in all directions, i.e. about 1.4x1.4 meters, but
preferably more like 2 lambda) metal surface, and then the VNA is
underneath it? You're reading the VNA from a distance (more than 2
meters away)? There's no other power cables, USB cables, etc. The
environment around the antenna and the ground plane is strictly
controlled - there's nothing within several meters (or at least nothing
that moves)
If you really want to get a handle on measurement uncertainty, you'll
need to make multiple measurements with each device (with a
disconnection and reconnection - since that is often one of sources of
measurement uncertainty. You taking some pains to torque the connector
the same each time to the VNA?
Are some SMA and some N? Was that factored into the cal strategy?
In real terms, what does 5% mean - these are all around 20 dB RL - that
means the test set is seeing about 1% of the RF power reflecting back at
it. The antenna is radiating the other 99%. So if there's something in
the vicinity of the antenna that reflects back another 1%, then the
reflection coefficient changes about 10%. It doesn't take much to
reflect (or scatter, is the usual term) 1%
A person has a scattering cross section of about a 0.5-1 square meter
back towards the source (monostatic cross section) - At a distance of 2
meters, your monopole's field is spread out over about 12 square meters
(really back of the envelope, and 2m is probably actually in the near
field still), so a 1 square meter reflector will send back 1/12th of the
energy (i.e. about -10dB) - yeah the reflector's not a corner cube
(although that's taken into account in the monostatic cross section), so
not all of the energy comes back to the source. And, you don't know what
phase it is. And it's physically large, so the phase of the reflection
from a person's middle comes back at a different phase than a person's
head or arms.
But the point is - it's *really* easy to get a contamination of a
measurement that will show up as a 1 dB change in a reflection magnitude
of -20dB.
For those wanting to set up home antenna ranges (a *great* use of a
nanoVNA) there's a bunch of things you can do. - get the measurement
equipment away from the AUT - a cable, with lots of the *correct*
ferrite around it (don't just assume any old ferrites are right). -
have a consistent ground plane and other environment - think in terms of
multiple wavelengths - a good "open air test site" is really open . Be
able to adjust the height or position of the AUT and its ground plane,
and take a bunch of measurements, and look for a cyclical variation as
you move it (the sure sign of a scattering source somewhere). Move your
receiving probe (the other port on the VNA) for the same reason.
Good luck.
5-10% on a "backyard range" is doing quite well.
> All devices were calibrated against the same standards at port 1 and
> each device is placed on a styrofoam support. The antenna was directly
> connected to port 1 for each measurement.
>