[PCB_FORUM] Re: Power plane -vs- gnd path signal return path

Power plane -vs- gnd path signal return pathDear Sir

Do you see any signal "return path" or integrity issues with routing a critical 
signal trace on the bottom layer as long as the driver of the signal was 
powered by 3.3V. 
No issues in theory.  However it would be better to use layer 12 (inside of the 
power plane layer) instead of layer 14 (outer most layer) since the impedance 
control (no change in impedance) will be better on an internal layer than 
outermost layer.  However this is a second order effect and from a first order 
point of view return path is ok with the power plane layer 13 for all signals

Can we assume that the power and gnd planes would be perceived as a short to 
hi-frequency signals, so that the return path would not have to get to a ground 
plane, but would just stay on the 3.3V plane directly beneath the trace until 
it returned to the driver.
Yes, the planes are considered shorted for high frequency signals. Any way a 
lot of capacitors are connected between the two planes, called decoupling 
capacitors, so the planes are actually shorted for high frequency signals 
(capacitors allow high frequency signals to pass throgh)

Regards
Aditya

  ----- Original Message ----- 
  From: Jim Wages 
  To: icu-pcb-forum@xxxxxxxxxxxxx 
  Sent: Thursday, March 13, 2008 6:50 PM
  Subject: [PCB_FORUM] Power plane -vs- gnd path signal return path


  Ok, folks. Let me know if my thinking is wrong here.

  I have a 14 layer stack up. External etch layers are adjacent to power 
planes, which are adjacent to ground planes. Lyr 2 is a split power plane. 
Lyr13 is a solid 3.3V plane.

  Do you see any signal "return path" or integrity issues with routing a 
critical signal trace on the bottom layer as long as the driver of the signal 
was powered by 3.3V. What about a differential pair?

  I'm assuming that the power and gnd planes would be perceived as a short to 
hi-frequency signals, so that the return path would not have to get to a ground 
plane, but would just stay on the 3.3V plane directly beneath the trace until 
it returned to the driver.

  Thoughts?



  Jim S. Wages

  SR. PCB Layout Designer

Other related posts: