FEDEVEL
Platform forum

HasanTheSyrian_ , 10-02-2025, 05:43 PM

I have a stripline trace, the top reference is gnd and the bottom is pwr. Is the impedance calculation affected if one of the planes is a power plane?

For example for my manufacturer, JLC, their calculator automatically chooses "top ref" and "bottom ref" but there is no accounting for a power plane

HasanTheSyrian_ , 10-02-2025, 05:44 PM

Also the power plane is close to the signal layer than the gnd plane

QDrives , 10-02-2025, 07:34 PM

A signal has a reference to both the layer above as well as below it.
The distance between the signal and reference(s) matters. The one closest to it, will have the most current.

HasanTheSyrian_ , 10-02-2025, 07:38 PM

what about signal integrity? most of the signal will use the power plane as return since its closer

HasanTheSyrian_ , 10-02-2025, 07:38 PM

and the power plane will have small cuts for every power pour

QDrives , 10-02-2025, 07:42 PM

What distances are we talking about?
"*the power plane will have small cuts for every power pour*" -- That would be problematic in EMC and SI.

HasanTheSyrian_ , 10-02-2025, 07:45 PM

but theres a gnd plane above at the same time

HasanTheSyrian_ , 10-02-2025, 07:45 PM

HasanTheSyrian_ , 10-02-2025, 07:45 PM

core is 0.5mm thick dielectric is 0.01

QDrives , 10-02-2025, 07:52 PM

Ok, that is just about only referencing power.
The stack-up that you show is for L1 for signals L3 for power and L5 or L6 for signals. L2, L4 and L6/L5 is Gnd.

If you need 4 signal / power layers, you should chose a different stack-up. Outer prepreg 2x1080 (~0.15mm), core 0.2mm, inner prepreg 4x7628 (~0.8mm).
That gives you something like 2x3 layers, so L2 and L5 are Gnd. L3 and L4 can be used for signals (perpendicular) or power.

"*but theres a gnd plane above at the same time*" -- The signal does not **"know"** that!

HasanTheSyrian_ , 10-02-2025, 07:55 PM

you mean i should swap L4 and L3?

HasanTheSyrian_ , 10-02-2025, 07:56 PM

maybe i can fill the power plane with the most common voltage and route other power nets with traces

HasanTheSyrian_ , 10-02-2025, 07:56 PM

-signal layer will still be 0.5mm away from the ground plane

HasanTheSyrian_ , 10-02-2025, 08:11 PM

or maybe i avoid cuts below the traces

QDrives , 10-02-2025, 10:01 PM

No, you should change the thicknesses of the dielectric layers.
"*or maybe i avoid cuts below the traces*" -- that is always wise!

Robert Feranec , 10-03-2025, 07:27 AM

I am working on a video to show this effects

Robert Feranec , 10-03-2025, 07:31 AM

as @QDrives explained, they are both reference planes. What the problems could be? For example the places where the planes are split (as you already mentioned) but also for example the places where signal is changing reference planes (in the places where VIAs are located)