Help needed with Friis Equation Calibration

[makif]

Hi all,

I will make some measurements for my work. I arranged this setup :

Pretty straightforward. I use signal generator to create 900 MHZ signal and use spectrum analyzer to measure it. The antennas are same for both TX and RX side and they are : https://www.pasternack.com/images/ProductPDF/PE51089-1.pdf

My input signal is +22 dbm. I directly connected generator to spectrum analyzer and measured +20.5 dbm. That's ok, i counted it as the loss of the cable of spectrum analyzer.

My antennas have 5m cables. They are RG-174/U coaxial cables. With some research on the web, i found out that loss at 900 MHZ for this type of cable is 90 db/100m. So, i have 4.5 db+4.5db = 9 db of antenna cable losses. Antenna datasheet says it has 3 dbi gain.

Now, i placed my antennas very close to each other ( 5 or 6 cm ).Using analyzer, i read output power as around -2 dbm . My initial thought was that for such an input power, this value was so low. According to my rough calculation, i should have had:

+20.5 dbm - 9 db + 6 db - 6 db = 11.5 dbm ( Final 6 is free space loss at 5 cm )

As you see i am quite far away from theoretical value. I tried it at 40 cm, 1 m and 2m. I have always big difference between theoretical and measured data.

I am not really very experienced on RF field. What i am missing here ?

Thanks in advance for all responses.

 

[jiripolivka]

Hi all,

I will make some measurements for my work. I arranged this setup :

Pretty straightforward. I use signal generator to create 900 MHZ signal and use spectrum analyzer to measure it. The antennas are same for both TX and RX side and they are : https://www.pasternack.com/images/ProductPDF/PE51089-1.pdf

My input signal is +22 dbm. I directly connected generator to spectrum analyzer and measured +20.5 dbm. That's ok, i counted it as the loss of the cable of spectrum analyzer.

My antennas have 5m cables. They are RG-174/U coaxial cables. With some research on the web, i found out that loss at 900 MHZ for this type of cable is 90 db/100m. So, i have 4.5 db+4.5db = 9 db of antenna cable losses. Antenna datasheet says it has 3 dbi gain.

Now, i placed my antennas very close to each other ( 5 or 6 cm ).Using analyzer, i read output power as around -2 dbm . My initial thought was that for such an input power, this value was so low. According to my rough calculation, i should have had:

+20.5 dbm - 9 db + 6 db - 6 db = 11.5 dbm ( Final 6 is free space loss at 5 cm )

As you see i am quite far away from theoretical value. I tried it at 40 cm, 1 m and 2m. I have always big difference between theoretical and measured data.

I am not really very experienced on RF field. What i am missing here ?

Thanks in advance for all responses.

Your antenna arrangement is plainly wrong. Please start first by connecting your transmtter through an attenuator to your receiver (spectrum analyzer). Then your power data should agree.

Next you may try to connect the antennas instead of the attenuator. At 900 MHz, I would suggest you separate the antennas at least by 10=-20 meters. Measure cable loss to be sure

Any conductive object like your test instruments located between antennas would contribute to a low or high loss due to not line-of-sight propagation .

 

[makif]

Your antenna arrangement is plainly wrong. Please start first by connecting your transmtter through an attenuator to your receiver (spectrum analyzer). Then your power data should agree.

Next you may try to connect the antennas instead of the attenuator. At 900 MHz, I would suggest you separate the antennas at least by 10=-20 meters. Measure cable loss to be sure

Any conductive object like your test instruments located between antennas would contribute to a low or high loss due to not line-of-sight propagation .

Thanks for the answer.

I didn't get attenuator thing, when i directly connected transmitter and receiver with a coaxial cable, i was able to measure correct result ( 20.5 dbm out of 22 dbm transmitter power ). I thought that 1.5 db loss as a cable loss. My problem is when i connect antennas. Do you suggest i should connect an attenuator between receiver antenna and analyzer ?

I need to get loss at 1m. There are several propagation models around using loss at 1m as a reference. I guess at 900 MHZ, Friis equation should hold at 1m ?

That picture was only for showing the setup. In real measurement, there is no obstacles between both antennas. So, i am pretty certain that there is LOS propagation.

 

[Dan Mills]

But given the physical size of those aerials you are still in the near field at 1m Friis only applies in the far field.

Further those look like 1/4 wave whips, which must act against a ground plane does your setup have one for each aerial?

Your -6db for free space loss would assume an aperture on the receving aerial encompassing 1/4 of the area of the sphere, seems unlikely even if you were not way into the near field, you need to be far enough apart that the field appears to have a locally spherical wavefront, not a cylinderical one (Which is what they approximate in the near field and is what gives them gain).

Make measurements at say 10, 20, 30 & 40m and plot the curve againt that expected from the maths, you will be much happier.

Any metal in the vicinity will tend to screw you up, do the work in an open field site.

I find "8dBi Omni Radiation Pattern" in the aerials data sheet to cause a certain amount of headscratching.

Regards, Dan.

 

[makif]

But given the physical size of those aerials you are still in the near field at 1m Friis only applies in the far field.

Further those look like 1/4 wave whips, which must act against a ground plane does your setup have one for each aerial?

Your -6db for free space loss would assume an aperture on the receving aerial encompassing 1/4 of the area of the sphere, seems unlikely even if you were not way into the near field, you need to be far enough apart that the field appears to have a locally spherical wavefront, not a cylinderical one (Which is what they approximate in the near field and is what gives them gain).

Make measurements at say 10, 20, 30 & 40m and plot the curve againt that expected from the maths, you will be much happier.

Any metal in the vicinity will tend to screw you up, do the work in an open field site.

I find "8dBi Omni Radiation Pattern" in the aerials data sheet to cause a certain amount of headscratching.

Regards, Dan.

Thanks for the answer Dan. I don't have ground planes beneath my antennas. Do you suggest me to use ground plates right under antennas ? Any ideas for their sizes ?

I will make measurements at 10 m and 20 m tomorrow and will compare it with expected theoretical value. How did you estimate near-field BTW ? Only formula i know requires antenna diameter which i don't know in this case.

What about 8 dbi omni radioation pattern ?

 

[jiripolivka]

Thanks for the answer Dan. I don't have ground planes beneath my antennas. Do you suggest me to use ground plates right under antennas ? Any ideas for their sizes ?

I will make measurements at 10 m and 20 m tomorrow and will compare it with expected theoretical value. How did you estimate near-field BTW ? Only formula i know requires antenna diameter which i don't know in this case.

What about 8 dbi omni radioation pattern ?

As I wrote, your antenna arrangement is plain wrong. Your antennas must be located over the "far zone", more than that is important for low-directivity antennas like yours.
Relatively good results may be achieved by locating the antennas in an anechoic chamber where other than LOS wave are rejected.

Start with 20-meter distance and draw a plot of power as a distance function. You will see how the power level abruptly varies for short distances.

 

[makif]

As I wrote, your antenna arrangement is plain wrong. Your antennas must be located over the "far zone", more than that is important for low-directivity antennas like yours.
Relatively good results may be achieved by locating the antennas in an anechoic chamber where other than LOS wave are rejected.

Start with 20-meter distance and draw a plot of power as a distance function. You will see how the power level abruptly varies for short distances.

Ok, i will try that but how do you estimate far-field distance ?

One final question : Datasheet says 3 dbi gain and 8 dbi omni radiation pattern. I should measure antenna gain as 3 db in the far-field zone right ?

Thanks for all your helpful posts.

 

[jiripolivka]

Ok, i will try that but how do you estimate far-field distance ?

One final question : Datasheet says 3 dbi gain and 8 dbi omni radiation pattern. I should measure antenna gain as 3 db in the far-field zone right ?

Thanks for all your helpful posts.

Some sources indicate the far-field distance as (1...10) D squared/ lambda, where D is aperture size (approx.monopole size) and lambda is wave length.
With aperture antennas like horns and dishes the antenna pattern forms quite far away from the antenna. Nevertheless your experience with the pair of monopoles shows also the difficulty.

Over short distances like you show one cannot calculate any real number due `to close-object dispersion and inductive coupling between the radiators.