would splitting the ground plane on a PCB into 3 sections that are still commonly connected be a bad idea.
my idea is to split (shield) the power from the main board so there is no direct (line of sight connection) and at the same time split the input and output audio sections as well. all three planes are still connected at the top of the board i'm just making it so there is no direct connection so each section has to travel to the common ground but not interfere with each other.
my idea is to split (shield) the power from the main board so there is no direct (line of sight connection) and at the same time split the input and output audio sections as well. all three planes are still connected at the top of the board i'm just making it so there is no direct connection so each section has to travel to the common ground but not interfere with each other.
One important point is to avoid coupling the current pulses going through
the capacitors after the rectification diodes, into the audio ground.
there are only 2 1000u coupling caps right at the top of the board. the regulation is RJM's x-reg (with kind permission). the bridge rectifiers are dip4 50v 1.5a and on board.
i'm trying to accomplish making sure that each section does not interfere with each other and hopefully give a cleaner ground plane for each section.
i'm trying to accomplish making sure that each section does not interfere with each other and hopefully give a cleaner ground plane for each section.
The voltage regulator circuit will likely couple noise to the rails with this layout,
since the op amp and resistors won't have a quiet ground. I'd separate the rectifiers
and their DC smoothing capacitors into a power plane that is separate from all other grounds.
When that is done, all the regulator and audio grounds can then share the same ground plane.
This audio ground plane should connect to the power plane at a point brought out as a stub
from midway between the input electrolytics' ground leads. Also keep each rectifier and
filter capacitor pair close together.
That's more like it, and flip the rectifier filter caps so the grounds are nearer the audio circuit.
Keep both of the rectifier grounds on the power plane, that's important.
In fact, I'd consider swapping the relative positions of the rectifier and filter capacitor,
so the capacitor grounds are next to the stub.
Last edited:
A good place for inspiration is the my_ref-fe.
Input ground plane over the input traces. Power ground plane under the power traces. No power traces running over the input ground plane.
http://www.diyaudio.com/forums/group-buys/229358-my_ref-fremen-edition-interest-group-buy.html#post3356205
Input ground plane over the input traces. Power ground plane under the power traces. No power traces running over the input ground plane.
http://www.diyaudio.com/forums/group-buys/229358-my_ref-fremen-edition-interest-group-buy.html#post3356205
i had a good look at that but i'm little confused.
the ground planes look like they are for each stage. the input is on 0v which to me says that it relies on the input source ground which is no bad thing as it removed the duty on the rest of the circuit except that if the input source has a bad output grounding scheme (mostly very unlikely) then so does the input. i like this idea.
the input voltage ground is tied into what looks like a relay to feed the output ground to COM at switch on for 'anti rush in/thump control'.
the ground planes look like they are for each stage. the input is on 0v which to me says that it relies on the input source ground which is no bad thing as it removed the duty on the rest of the circuit except that if the input source has a bad output grounding scheme (mostly very unlikely) then so does the input. i like this idea.
the input voltage ground is tied into what looks like a relay to feed the output ground to COM at switch on for 'anti rush in/thump control'.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.