Has the VNA calibration been properly performed? It is normal to get on the order of 10 dB down reflections from discontinuities, cable-waveguide transitions, etc. We exploit the VNA architecture by removing the effects of unwanted reflections via careful calibration. Only then can we achieve such apparently low values of reflection coefficient.

Thank you very much for your answer, John. Yes, the calibration was done carefully. I could not figure out where the unwanted signal come from. The whole system is placed in an anechoic room. What is the typical log.mag value that you got?

About what Martin quotes above, namely around -60 dB. We routinely achieve -90 dB for pure coaxial systems, but any transition to free-space launching antennas requires some form of mechanical calibration. While these method are quite good, they limit dynamic range to the above numbers. In any case you should be able to achieve better than -10 dB. I recommend performing a systematic check of all transmission line components starting with the test port cables. A connector may be damaged or something has been improperly mated.

I agree that without any reflection you should get -60dB to -80dB. If the calibration is not that good you can get -30dB maybe, -10dB sounds really high. Did you try to cover the receiving or the transmitting antennas to see if the signal changes? It will help you to figure out if the problem is in the connection to the VNA (cables, adaptors,etc...).

The given answers by colleagues have explained the causes of the relatively high received signal. As Mr Nam mentioned ( transmitter and receiver), I can point out another problem, which is the coupling between the transmitting and receiving antennas. To reduce coupling, the receiver can be put slightly behind transmitter, and increased distance between them will help.

In theory your result should be equal to the dynamic range of your VNA: -80dB. -120dB, or whatever you have. But in practice the mag of S11 will depend on how well you did the calibration of the instrument. What magnitude do you get when you put the matching load? Good luck with the calibration. regards J

OOps, I only now saw the earlier advice on the calibration and your response. Try what Khallil suggested, lower coupling shpuld help

Thank you all for invaluable comments and suggestions. I greatly appreciate that. We use transmitting and receiving antennas for measuring the reflection from a target. So we do not collect the S11 signal, but S21 instead. We tried to align the antennas at different angles and distances: low angles and long distances are just a little help. I even try to decouple the antennas by placing a beam blocker along side and in between them, but the improvement was just small. So it might, as many of you pointed out before, be the calibration that caused large reflection number. The measurements were at low frequency, below 18 GHz. If you have any further comments/suggestions, please let me know.

Is it the same set up that u are mention early? Perhaps u can post a VSWR picture or data for a blank waveguide

You have posted almost everything.... So what i could mention is --- noise. Maybe you should take care of your environment, and be sure that those unexpected are not from the noise or even the equipments

Hi,

Could you explain a little bit more your experience. In the first question, you speak about "free-space reflection measurements". So this means S11. then, after the very good and accurate answer from Martin Hudlicka, you said you don't "collect the S11 signal, but S21 instead". S21 is the transmission coefficient.

So for reflexion coefficient S11, as said Martin Hudlicka, -50db -60db is quiet good. And if you speak about S21, -10dB can be true: it depends on your antenna and the distance between your twos antennas. What kind of antennas do you use ?

Hi every body . I have samples that I need to measure its characteristics using free space method ..... Any body give me hints where can I do this