Hi,
dinone, thanks for your reply!
I do indeed have a 2-layer board, there are some tracks one both sides, but both are filled with a ground plane. So, if I clear the ground plane from the top layer around the microstrip, and leave the bottom layer, it should be ok to use a microstrip?
It seems to be a bit of a 'black art', although I have been readingup on this subject for a while now, and I uderstand most of the theory.
1) you can change topology: you can design the track in CPW instead of Microstrip. So you can reduce the track wide and improve the isolation.
Interesting, I've been having trouble finding a clear definition of 'CPW', although the software utility I have 'TXLine' shows that the trackwidth must be absolutely tiny 0.09mm, and the gap between the track and ground plane, even smaller, 0.01mm. Alas, I don't think my PCB manufactuer can cope with this. Still, its introduced me to something new.
2) you can "not give a damn" of the track impedance and change it reducing the track wide: you will see that the performance of your transmitter/receiver not will be affect so much (if the track lenght is short).
This is what I was thinking, after all, the distance between the module's RF out connector and an SMA connector should be no more than 5mm, possibly 3-4mm.
After searchinng google for PCB designs involving 2.4GHz, it seems they get around the 'wide trackwidth' problem, by using a thinner board 0.8mm. This of course, halves the width that the microstrip has to be, down to rougly 1.5mm. Then again, I have seen designs that don't seem to obey any sort of rules. The have several 'impedence matching' tracks, but the actual connection to the SMA connector vary in width a great deal.
Anyway, thanks for your input, and if anyone else would like to add their two cents, I'd appreciate it.
BuriedCode.