Do nothing, its not high speed and it doesn't require length matching between the traces.
The worse thing you could do is cut the ground plane under the other two traces, cuts in ground planes is sin number one in digital layout.
Your power traces on that design need making wider...And some of the component pads look larger than the others? The via next to the pad will act as a solder thieve!!!!!
For most digital at speeds below DDR memory interfaces the actual impedance of the trace dose not really matter and is not calculated or catered for during PCB manufacture. For this design it wont matter or make any difference to the functionality. USB interface, Ethernet DDR memory yes it can be critical for this the actual figure does not matter, what does matter is preventing discontinuities' of the traces impedance from source to destination.
A starpoint is the physical point where separate branches (planes on a PCB or wires in a box) join together, it allows the 0Vs to be separated to cater say for a high current path, but still be at the same voltage (or as near as possible).
http://www.x2y.com/filters/TechDay0...log_Designs_Demand_GoodPCBLayouts _JohnWu.pdf
http://www.ultracad.com/articles/slots.pdf
Slots in Ground Planes
And finally a couple of designs (double sided) that I have played with for fun, just about everything else I touch I cannot publish due to either NDA's or security reasons. Both need finalising one day, I laid them out as a bit of light relief from my day to day work, time approx. 4 hours each.
The worse thing you could do is cut the ground plane under the other two traces, cuts in ground planes is sin number one in digital layout.
Your power traces on that design need making wider...And some of the component pads look larger than the others? The via next to the pad will act as a solder thieve!!!!!
For most digital at speeds below DDR memory interfaces the actual impedance of the trace dose not really matter and is not calculated or catered for during PCB manufacture. For this design it wont matter or make any difference to the functionality. USB interface, Ethernet DDR memory yes it can be critical for this the actual figure does not matter, what does matter is preventing discontinuities' of the traces impedance from source to destination.
A starpoint is the physical point where separate branches (planes on a PCB or wires in a box) join together, it allows the 0Vs to be separated to cater say for a high current path, but still be at the same voltage (or as near as possible).
http://www.x2y.com/filters/TechDay0...log_Designs_Demand_GoodPCBLayouts _JohnWu.pdf
http://www.ultracad.com/articles/slots.pdf
Slots in Ground Planes
And finally a couple of designs (double sided) that I have played with for fun, just about everything else I touch I cannot publish due to either NDA's or security reasons. Both need finalising one day, I laid them out as a bit of light relief from my day to day work, time approx. 4 hours each.
There are 4 traces between 2 IC and the current flow from left to the right by them. The return pathes are under that traces (background is solid ground).
Trace width is 0.33mm and gap width is 0.32mm (IC pin step is 0.65mm).
I thought that make distance between traces twice more than trace width will be better than reduce it as maximum as possible (avoid capacitance).
If we have solid ground plane the return path will be under trace in any case for digital signal and strait line to the starpoint for analog signal. Right?
Let see this example
There are digital signal(green) to the IC and return signal (red) from the pin on opposite side of IC. Will the return signal not flow under another digital trace in spite of it is shortest way (like blue)?
And about analog signals. Return path (red) will make loop on solid ground plane.
Shall I make wide return trace instead solid plane?
Let see this example
An externally hosted image should be here but it was not working when we last tested it.
There are digital signal(green) to the IC and return signal (red) from the pin on opposite side of IC. Will the return signal not flow under another digital trace in spite of it is shortest way (like blue)?
And about analog signals. Return path (red) will make loop on solid ground plane.
Shall I make wide return trace instead solid plane?
An externally hosted image should be here but it was not working when we last tested it.
Not quite, how close the signal return tracks the signal depends on frequency as this first link shows:
Successful PCB grounding with mixed-signal chips - Part 1: Principles of current flow | EDN
For audio signals that cover 20-20,000Hz this can cause fun as the lower frequency signals follow the path of least resistance but when they get higher a certain proportion of the return current wants to also follow the path of least inductance, this is what often gives rise to ground loop problems and why it is often the lower frequency buzzes that occur with ground loops.
This is why ground planes are good for all frequencies and types of signals, they provide a direct low impedance path for the inductive return under the positive signal and provide a low resistance path for the resistive return current path.
This is just some visuals of return paths.
Visible Return Current
And this is some very interesting stuff on the Ground Myth.
http://sites.ieee.org/ctx-emcs/files/2010/09/Archambeault-Ground-Myth.pdf
This applies for low level signals, when looking at the outputs of power amps the high current returns require special treatment and usually a separate return, but that's another subject.
Successful PCB grounding with mixed-signal chips - Part 1: Principles of current flow | EDN
For audio signals that cover 20-20,000Hz this can cause fun as the lower frequency signals follow the path of least resistance but when they get higher a certain proportion of the return current wants to also follow the path of least inductance, this is what often gives rise to ground loop problems and why it is often the lower frequency buzzes that occur with ground loops.
This is why ground planes are good for all frequencies and types of signals, they provide a direct low impedance path for the inductive return under the positive signal and provide a low resistance path for the resistive return current path.
This is just some visuals of return paths.
Visible Return Current
And this is some very interesting stuff on the Ground Myth.
http://sites.ieee.org/ctx-emcs/files/2010/09/Archambeault-Ground-Myth.pdf
This applies for low level signals, when looking at the outputs of power amps the high current returns require special treatment and usually a separate return, but that's another subject.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.