[SOLVED] skin depth clarity required.

[shashy.br]

Hello everyone,

Till now i though the RF wave travels on the top surface of the conductor ,
let us say a wave of frequency 6720MHz has a skin depth of 0.77um or so,
and i should give proper thickness of Immersion gold to accomodate for the skin depth,
but i found a confusing piece of explaination for skin depth sating that a minor part of the wave only travels at the surface,almost a majority of the wave travels closer to the surface nearer to the dielectric.8-O

please let me know i this is right or I simply do not understand what they are tryin to say!!!!!:-D

**broken link removed**

 

[biff44]

you seem to be confusing 2 concepts.

If you have a symmetrical transmission line cross-section, such as coaxial cable, then you skin depth understanding applies. The RF currents on the center conductor, for instance, are mostly on the surface, with just a small portion going deeper. So if you are fabrcating your own coaxial line, you can use either a solid center wire, or a hollow pipe for the center wire with NO difference in performance. No RF currents penetrate deeply, so the microwaves do not know that the center conductor is hollow.

If you instead are talking about microstrip, i.e. a narrow width line over a wide ground plane, then you are talking more about what the electromagnetic fields are doing above (in air dielectric) and below the strip (in the PCB dielectric). Since the PCB dielectric constant is higher than air, you will find the fields are more concentrated in the solid dielectric than in the air. Consequently, the RF currenct will be somewhat stronger in the bottom side of the narrow strip than the top side (assuming you are at a high enough frequency for skin depth to matter).

 

[shashy.br]

Thanks for the reply .....

So in case of microstrip lines and CPWG lines,
the RF current is maximum at the bottom of the strip which is nearer to the dielectric!!!
hence the finishing which is done at the top surface does not affect the loss in the transmission line to a maximum extent,
for example consider a copper strip plated with ENIG ,is the losses encountered at X band are mainly due to the copper conductor only??(in case of microstrip lines)

 

[biff44]

To some extent yes. If the bottom of the trace is pure copper, that will have maybe 65% of the RF currents in it. Then what you plate on the top of the trace has less effect....although it is still important! don't go plating it wil nickle and think you will get away with it!

 

[shashy.br]

Thanks for the info ......helped me a lot