I am designing a 12 GHz LNA on a Regoers 5880 substarte. When i cary the thickness of the substarte for a micorstrip implementation my noise figure varies.
What is the reason behind this.Does this happen in all substartes or am i doing anything wrong.
Please help.
Also , I need to add coupling caps and bias circuit. How do i design the coupling caps....
As you vary the substrate thickness the impedance of printed traces/transmission lines will change. This in turn will change the matching for the devices and therefore the Gain and NF of the LNA will also change.
hmmm... not quite right RealAEL. I am trying to match to a the same load with the same impedance. So i dont change the electrical parameterslike Zo and effective electrical length. I am only changing the physical parameters of the line which yield me the same impedance but yield different noise figure.
What was the first line you were looking at CPW has less interference and there for has a lower noise figure where microstrip has a higher noise figure.
I use decoupling caps on many of my designs as long as the electrical length of the cap is short it will not affect your performance (ie DC does not care about impedance). Just try to keep the cap about the same size as your microstrip/CPW line and try to keep is short (ie small gap between cap sides) you will find if you design it like this you should not run into any problems. (try not to make sharp corners anywhere on your board even at the cap solder edges)