2 Replies Latest reply on Jul 24, 2013 6:35 AM by strangd

    Ground Vs. Power Planes & Impedance

    dupre

      I note that Hyperlynx LineSim does  not differentiate between power and ground planes in a layer stackup.   Does this mean there is no difference between power and ground planes  when defining a controlled impedance stack-up?  I was always under the  impression that you reference a controlled impedance trace to adjacent  ground plane(s).  Does a power plane do the same thing???

       

      - Joe

        • 1. Re: Ground Vs. Power Planes & Impedance
          yu.yanfeng

          Yes. there is no difference in charactor impedence whatever TL's  reference  plane be.  However you should keep the impedence between power plane and ground plane low enough,usually you have to add Decaps to implement this lowe impedence purpose.

          yanfeng

          • 2. Re: Ground Vs. Power Planes & Impedance
            strangd

            The signal wave front propagates from the source to the load(s)

             

            The high speed signal wave front is Electro-Magnetic and therefore it is a wave of photons moving through a conductor on the PCB from the signal source towards some load(s).  The photons which are intercepted by the atoms in the conductor, transfer energy to the atoms’ electrons and finally allow the electrons to move from atom to atom.  This would be the movement of charge in the conductor if the were not a ‘Return’ in the circuit.

             

            The signal wave front’s ‘Return’ wave front propagates from the source to the load(s)

             

            The ‘Return’ wave front is not created in the source.  The ‘Return’ wave front is not created in the load.  The signal wave front generates the ‘Return’ wave front.

            Because the signal wave front generates the ‘Return’ wave front, both the signal and the ‘Return’ wave fronts start at the source and propagated towards the load.

             

            The signal wave front generates the ‘Return’ wave front

             

            As the signal wave front propagate down a trace with say, a positive signal to it, there is a field generated at 90° to the direction of travel of the wave front; the right hand thumb rule.  When this field passes through another piece of conductive material, this field causes electrons in the other piece of conductor to start moving but in the opposite direction from the original signal.  There is a new wave front generated in the conductor being affected by the first field but it will in the opposite polarity.  Since the first wave front’s field is generating the ‘Return’ wave front, both are propagating across the board in the same direction.  Due to the interaction of the positive signal’s field with the second conductor generating an opposite field in the second conductor, the electron movement in the second conductor is in the opposite direction as the electron movement in the signal’s conductor.

             

            Fields generated by wave fronts

            If all the conductor paths are duplicated in the signal conductor and the ‘Return’ conductor, there will be a good, completed electromagnetic circuit from source to load.  One should not think of the ‘Return’ current starting at the load end of the trace at the same time as the signal starts at the source end of the trace.  To do so, the atoms at the load end would have to know what the atoms at the source end of the trace were doing before the light form the source end could reach them.  This is not a probable situation.  No attempt has been made to Entangle [2] the electrons of these distant atoms.  The electron flow in the signal carrying conductor and the ‘Return’ conductor are caused by the fields generated by the wave fronts.  The faster the rise time of the wave fronts and therefore the faster change in the fields, the better the coupling to the conductors and then smaller geometries of the conductor can produce other, possibly unwanted, fields and currents. Examples in planes would be resonance nodes and eddy currents.

             

            I have attached a U2U presentation that I did in 2005.  Have fun.