Resistor matching Vs RF Matching
My question is with series terminations/matching on a typical high speed PCB bus design, versus matching to say, transmission lines, like when you transition between impedances on a microstrip circuit...
--------------------------here is my question-------------
On occasion it is neccesary to use matching resistors to transition between impedances, wether it be in the RF world or in the Digital world.
Can someone please explain to me, why in the RF world, you calculate this value using;
R = SQRT(Rsource*Rtline)
But in the PCB digital world, you use;
R = Rtline - Rsource
thanks
Hi,
I think there are a lot of things confused in your question. I am not sure that R = SQRT(Rsource*Rtline) is used for any matching?!
It is probably
Zmatching=SQRT(Zsource*Zload)
for quarterwavelength transmission line transformer matching between source haveing Zsource and load having Zload.
flyhigh
I think you are confusing "quarter-wave matching" and "source termination" (or "series termination", if you prefer).
First, a terminology issue: It may just be me, but when you refer to R of a transmission line, I think of R from an RLGC model of the line. From context, I can tell you are referring to the real part of Zo. But unless you are dealing with non-TEM modes in cutoff, Zo is mostly real so most people just use Zo not Rxxx. At least this is my experience, others may disagree.
Getting back to your original question, there are many different types of matching. For RF, there is also single and double stub, and for PCB's there is parallel or load terminations. In source termination, the source is matched to the transmission line, i.e., output resistance + R = Zo. Waves are reflected at the load, travel back to source, and are absorbed in the source resistor(s). In an RF application we don't want energy absorbed by the source, we want it absorbed by the load. A quarter-wave transformer maximizes the power absorbed by the load at a given frequency.