Hi guys,
We are on our 2nd generation product using CC8520s. We started our designs from CC85XXEM-RD (http://www.ti.com/tool/cc85xxem-rd). We had to make some changes due to the product size constraints, specifically:
1. We do not use the CC2590 chip -- we are driving directly from CC8520.
2. We have a master and slave designs (sometimes more than one slave)
3. On our slave design, we do not use an inverted F antenna -- instead we use the smaller size meandering F (AN043)
4. We have a lot of plastic and other elements around the antennas
Because of #4, we started looking at tuning the antenna by first measuring the impedance of the raw antenna (no tuning network) using a network analyzer. It turns out that both sides are tuned to a lower than desired frequency.
Long story short, for the slave, we have to shorten the antenna slightly to account for the environment. It is possible to tune it with passive components only, but the profile is much better if we adjust the layout.
It appears that we can get the desired antenna performance on both slave and master without the Low Pass T (the set of components between the balun and the antenna) used in the original reference design, or even if we have to, it would be a single pole only.
Questions:
1. Does this circuit (between balun and antenna) serve any other purpose other than antenna matching? I remember reading somewhere that you have a 4.89Ghz issue inside CC8520 -- so is this circuit also serves as a filter?
2. Based on CC85XXEM-RD -- is the output of balun truly 50+j0? Is this the impedance we are trying to match with?
Vlad
