Hope it help.....you will require an noise free environment (anechoic chamber) ,then need to have an rotating platform which will move source antenna with the require degree....finally an receiver antenna which will connected to an VNA (measure the signal strength and calibrate it with degrees )...

(position of antenna depends that rather you want to use antenna for transmitter or receiver)

Agree.

There are some issues there though. You need to look at the dynamic range of your system, i.e. you want you received power higher than the noise floor of your vna, but you also need your receive antenna to be in the far-field of your transmit antenna.

Usually it is not automatic, which is downside of the approach. Yet it has some advantages as you can use time-gating either in the vna itself or externally with say a matlab script to get rid of the external reflections which can occur if you anechoic chamber is not very good or the fixture you use for the measurement somehow influence your pattern. This is very important for bi-static measurements.

And if you want to measure gain of your antenna you need to put a reference antenna with known gain (e.g. standard horn) in the same place and measure it and then compare the gains.

Not easy to get high quality measurements, especially for lower frequency and physically larger antennas. What frequency and now large is your RX antenna? You need as much separation as you can get so that the TX signal approximates a plane wave. But separating them causes transmission line loss from the VNA to the TX antenna. Using a TX amplifier at the antenna helps with that. Also, take a close look at the minimum RX signal that y our VNA can detect. Many of them (depending on the sensor etc) can only reliably receive down to -50 dBm. If you want FB ratios and side lobes, you need to go down to those levels. But that assumes you can get RX signals of 0 dBm or higher. Also take a look at the maximum signal level of the sensor. Dmitry spoke to the dynamic range. Spot on. You need as much linear dynamic range as you can get. RX preamps can help but they are not terribly linear. How much absolute accuracty do you want. If you are working 10 GHz and above with small antennas you can get fairly decent results by putting the whole system on a long conference table. You will not get results good enough for regulatory certification purpose but you can see the effects of changes you have made in the antenna.

As for the dynamic range ans sensitivity of a vector netowkr analzer:

One an get easily down to -160 dbm /Hz and apractical number (with 100 hz resution bandwidth) is about -140 dBm. This is largely enoughin most cases.

The -50 dBm quote above are valid for simple detector diode or if you want scalar networkanalzyer (which should NOT be used in this case) the instantaneous dynamic range (without range switching) of a VNA exceeds normally 100 dB.One possible problem is to find a good anechoic chamber for lower frequencies.In this case near field range techniques can be applied or time domain gating with the synthetic pulse method.

[email protected]

Not an easy subject. First of all the VNA is today the main tool used for antenna measurements. The idea is to transform the antenna measurement into a series S12/S21 measurement. So you need a second antenna and long enough cables. You also need one sample for each direction in which you would like to know the pattern, if you can arrange your measurement in such a way to have the two antennas (Device Under Test -DUT- and probe) sufficiently far apart. That is at a distance of at least 2*D^2/wavelength one from the other. You will then need to be able to rotate your DUT to cover all the directions you need. This usually implies two rotation axes, plus a third to adjust polarisation. You may also need to adjust the polarisation of your probe to make sure you get the right answer. This depends on your requirements in terms of antenna pattern and the way you arranged the two rotation axes. You also have the option to place the probe closer to the DUT, but then you need to convert the filed values you measure into the corresponding values you would have far away and this is quite involved. On the other side it may simplify other things. E.g. allow to measure relatively low frequency antennas (still over a spherical surface) or move the probe on a plane if the DUT antenna is directive, i.e. be at least a few wavelength is size (>~5) and radiate a collimated beam.

The next thing you would like is to avoid the presence of scattering objects all around else you will not be measuring the radiation pattern of your antenna but the combination of it with all the echoes coming from the environment. So you need an anechoic chamber or at least to place absorbing material (for the frequency range you are interested in) all around and at a sufficient distance. As an alternative you can go outside, bat scattering will also be present there and if you need precise measurements or you work at low frequencies (

I fully agree on what what said above ! often people use optical fiber links in this context to avoid problems related to cladding modes on coax cables in such a measurement test stand . optical fibers can also be used forproviding DC power to head amplifiers (i.e. strong LEDs on one side and and array of photo cells on the other side).Commercial systems for this can produce 5 Watt DC power.

F.Caspers