1) Looks like you've got power and GND reversed on the display (I'm not familiar with the part, but that's what your schematic is telling me).
2) I assume you are relying on the RP2040 internal pull-ups for the buttons. Maybe your tracks are short enough to get away with that, but personally I'd be fitting pull-up resistors to guard against RF pickup etc.
3) Huge crystal! OK, you've got space for it here, but the 3.2x2.5mm crystals take much less space and so let you keep the connections shorter.
3) Layout: Did you adjust your GPIO selections to make layout easier? Obviously there are some constraints, but this design looks like it has everything coming out of the wrong side of the chip and having to make a long trip around the PCB, which could have been made easier by swapping GPIO assignments.
4) Layout: Your ground connectivity isn't too bad, but for example C13 and C8 aren't really doing anything for you: the + end is nice and close to the chip, but the negative end has an enormous distance to travel to reach the RP2040 ground pad. Current flows round loops, so both ends of the capacitor are equally important! Normally the best approach, particularly if you have plenty of space like you do here, is to keep almost everything on the top side and just use the bottom side for short links to cross things over - so that most of the bottom is GND and where you do have breaks in it the current only has a short detour go go round. If you really can't avoid a big break in the ground on the bottom side, stitch across it with tracks or fill on the top side.
For example, you've got a big group of tracks coming from the right-hand side of the chop and going round towards the LEDs/buttons. These could have stayed on the top side and gone round to the south of the crystal, keeping them away from the critical area around the RP2040. They could either go 'between the legs' of SW8 and leave the rest of your layout as-is, or keep them where they are and, at the point where that whole bunch of tracks currently goes to the bottom side to avoid a 3V3 track, put the 3V3 on the bottom side and keep it as short as possible just to cross over the LED/button tracks.
Inside the ring of pads of the RP2040, you've used up valuable routing space to connect pin 19 to GND. This is a fairly unimportant pin (no current flowing in it), so could easily have been connected to GND on the outside, letting you use that inner routing space for power. Or on the other hand, if you are going to connect it like you have now, at least connect it to the top-side groundplane: you've got a finger of groundplane that reaches towards it but doesn't touch.
5) Layout: Whether or not you do anything about my comment (4), you should add lots more via holes to stitch together the bits of groundplane you have on the top side to the groundplane on the other side.
2) I assume you are relying on the RP2040 internal pull-ups for the buttons. Maybe your tracks are short enough to get away with that, but personally I'd be fitting pull-up resistors to guard against RF pickup etc.
3) Huge crystal! OK, you've got space for it here, but the 3.2x2.5mm crystals take much less space and so let you keep the connections shorter.
3) Layout: Did you adjust your GPIO selections to make layout easier? Obviously there are some constraints, but this design looks like it has everything coming out of the wrong side of the chip and having to make a long trip around the PCB, which could have been made easier by swapping GPIO assignments.
4) Layout: Your ground connectivity isn't too bad, but for example C13 and C8 aren't really doing anything for you: the + end is nice and close to the chip, but the negative end has an enormous distance to travel to reach the RP2040 ground pad. Current flows round loops, so both ends of the capacitor are equally important! Normally the best approach, particularly if you have plenty of space like you do here, is to keep almost everything on the top side and just use the bottom side for short links to cross things over - so that most of the bottom is GND and where you do have breaks in it the current only has a short detour go go round. If you really can't avoid a big break in the ground on the bottom side, stitch across it with tracks or fill on the top side.
For example, you've got a big group of tracks coming from the right-hand side of the chop and going round towards the LEDs/buttons. These could have stayed on the top side and gone round to the south of the crystal, keeping them away from the critical area around the RP2040. They could either go 'between the legs' of SW8 and leave the rest of your layout as-is, or keep them where they are and, at the point where that whole bunch of tracks currently goes to the bottom side to avoid a 3V3 track, put the 3V3 on the bottom side and keep it as short as possible just to cross over the LED/button tracks.
Inside the ring of pads of the RP2040, you've used up valuable routing space to connect pin 19 to GND. This is a fairly unimportant pin (no current flowing in it), so could easily have been connected to GND on the outside, letting you use that inner routing space for power. Or on the other hand, if you are going to connect it like you have now, at least connect it to the top-side groundplane: you've got a finger of groundplane that reaches towards it but doesn't touch.
5) Layout: Whether or not you do anything about my comment (4), you should add lots more via holes to stitch together the bits of groundplane you have on the top side to the groundplane on the other side.
Statistics: Posted by arg001 — Fri Feb 02, 2024 9:42 am