Impedance discontinuity on SMA connector

[davenn]

The idea was using the pillars fixing the board to the chassis as the return path. That is why I placed the mounting holes close to SMA's.
As the board hasn't components at the bottom, I guess I should discard pillars.
Am I right?

as we all have said several times, you cannot do that
Look again at that document provided earlier and at the very first figure 4.1 a

any deviation from this will cause impedance and loss problems


Dave

 

[D.A.(Tony)Stewart]

This is just one supplier of boards. where you can compare Dk and Df vs f

**broken link removed**

Selecting a laminate material is a balancing act
http://www.ee.sc.edu/classes/Fall12/elct865/Class_Notes/Isola Part 3.pdf

 

[jasensio]

Ok I will spend some time reading and checking all the info.
I'll be back

Thanks

 

[volker@muehlhaus]

This is just one supplier of boards. where you can compare Dk and Df vs f

Sure, but what's the point?

The line on FR4 will have some slope in insertion loss between 1GHz and 2GHz, and so does the coax cable used to feed the signal. Even the antenna gain varies across the frequency range. So what?

I've designed 1:8 splitters for that range, and indeed the output isolation had a very strict requirement, so FR4 variation had to be considered. But here it's all nice and relaxed, as far as the substrate is concerned.