TEST RESULTS, SPIDER
(1) Spider with horizontal radials, Impedance = 26 ohms, SWR=1.9
(2) Spider with slanting radials 45 degrees, Impedance = 67 ohms, SWR=1.3
(3) Spider with vertical radials, Impedance = 73 ohms, SWR=2.2
Hi ab cd,
A 75 ohm spider (with horizontal radials) on top of a piece of 75 ohm coax will show 75 ohm on your analyzer if the antenna is really tuned to the measured 1090 MHz --> marker must be at the deepest point of the measuring curve. (It is true with a 65.3 mm pin in the middle and mostly measured at the feedpoint. See the explanation below...)
Avoiding the problems caused by the impedance differences of the antenna and coax, this latter shall be half-wavelength long or its multiplied value (VF is used) - so that we can see the real phase and impedance values on the antenna feedpoint. In all other cases the cable will have effect on the measured impedance. The cable can be of any length, only if the impedance of the antenna and the cable match, but here comes a new problem: Forcible insertion of radially routed wires into the coax cable can distort the shape and dimensions of the dielectric - the impedance of the cable.
Fully bent radials theoretically give 32.5 ohm, but take care of the proper lenght here as well.
Would you check the parameters and measure the antenna again, please?
Bending the radials may detune the antenna slightly, so a minor correction of the pin may be necessary.
(keep the frequency tuned in measurements)
VSWR measurements with a too wide frequency band will cheat - since the linearity is not provided by the device. Use 2-300 MHz bandwidth instead the full range. Please calibrate the analyzer to the measured band only - if it is possible on your device. Calibration points are not infinite, thus between two of them you can easily find anomalies.
Sorry for pointing at so many possible mistakes in antenna measurement, but just bending and measuring again will not always give the true answer about the antenna itself.