This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Which is my error in calibrating and/or designing this 868 MHz antenna?

Guru 18385 points
Other Parts Discussed in Thread: CC1101

Good morning,

I'm working with the TI Antenna DK. The kit is great. However, I cannot use the antennas in it as they come and I have to adapt to the geometry of our 868 MHz product, based on CC1101. I understand the boards #1, #2 and #9 in the kit are useful as SOL calibration standard, since all the boards share, more or less, the same shape.

So, my question is fourfold (does this word even exist in English?)  :

  1. After calibrating with these 3 boards, I assume the calibration plane is just at the point I indicate in red in the following picture, am I right? :)


2.- However, if my design does not follow this geometry, I'm wondering if I can simply calibrate at the end of the SMA cable attached to my VNA (this way the calibration point would be the green point in the figure) and then use the "port extension" feature to move along the blue trace up to the red point. Is my assumption right? Or this feature should only be used with cables and not PCB traces? How can I know the distance value for the port extension? This does not seem a good strategy to me, too many unknowns :)

3.- Finally, in my design, the blue trace is not this long and is shorter than λ / 10. It is 1 cm long. λ at 868 MHz is around 34 cm. According to theory, its contribution could be neglected. However, in the following Smith Chart simulation, I can clearly simulate how its contibution clearly affects the matching (note the indicated length of 1 cm = 0.01 m and minuts sign to indicate rotation Towards Load). What is your opinion about this?


4.- According to previous questions, would the best strategy be to generate to use calibration boards as in the example from TI? What if the blue trace is not 50 Ω as in the example from TI? I cannot get this clear area in real boards. Then this calibration would be useless, wouldn't it?


Ok, let me know all the things I'm doing wrong so I can quit considering RF as black magic! :)

  • 1) Yes
    2) Fully possible, we do it all the time. How practical it is depends on the VNA. Ours have a cal kit so calibration is done in seconds and it has a auto "port extension" feature. Basically you would ideally have to use this anyway since on your final board you don't have a SMA to connect to and a semi rigid cable is the alternative. When using a semi rigid you have to use port extension before soldering it to the board.
    3) A line that is shorter than 1/10 wave length will not behave like a transmission line but the signal still need some time to travel down the line (phase shift). A short line have to be taken into consideration but if this is the line toward the antenna (which is 50 ohm) you will not move far in the smith chart.
    4) Depends a bit on you VNA. With ours I have never used calibration board.