Your USB would be better routed as a differential pair and with SMD resistors. The digital devices need small value SMD X7R decoupling capacitors next to the power pins (the PTH capacitors are useless for local digital decoupling).
I would also move the route that runs next to the cap pins connected to the crystal, and again the crystal would be better with SMD devices to minimise loop area, as would all the digital.
You have slots in the GND plane which is not going to help signal integrity, especially where tracks cross the slots.
For the TPS60501 I would refer to page 18 of the data sheet for the layout and again change the caps to SMD devices as recommended, for best performance (minimise the noise).
Using SMD devices for the digital side will have many benefits and also free up space to allow you more chance to optimise the layout.
I would also move the route that runs next to the cap pins connected to the crystal, and again the crystal would be better with SMD devices to minimise loop area, as would all the digital.
You have slots in the GND plane which is not going to help signal integrity, especially where tracks cross the slots.
For the TPS60501 I would refer to page 18 of the data sheet for the layout and again change the caps to SMD devices as recommended, for best performance (minimise the noise).
Using SMD devices for the digital side will have many benefits and also free up space to allow you more chance to optimise the layout.
Looks nice!
I sent my PCBs out about a week ago, still waiting for them back. I decided to switch to SMD caps and a SOIC-8 op amp (both because of being much smaller and I wanted to use OPA1612) and I was able to get the board size down even further and was able to mount the caps closer to the pins.
Thanks a lot for checking and for good suggestions! Here's the new version with most of the suggestions implemented. All power decoupling caps are now in SM1210 X7R and also C5 and C6 which were supposed to be 1uf ceramic according to TI datasheet.
Attachments
Last edited:
Hi Marce,
what is the less Noisy : a Sussumu low noise resistor or a RN Dale throughol one ? (both for serie or shunt position to Gnd) ?
Could we have some advantage to mix through hole résistors with smd caps in a design despite the bad looking ? Fo example length of resistor which can avoid a copper trace or make 3 D design to cross at 180° a signal trace ? Is it noise with bad EMC with smd resistor VS the opposite with through hole resistor ?
what is the less Noisy : a Sussumu low noise resistor or a RN Dale throughol one ? (both for serie or shunt position to Gnd) ?
Could we have some advantage to mix through hole résistors with smd caps in a design despite the bad looking ? Fo example length of resistor which can avoid a copper trace or make 3 D design to cross at 180° a signal trace ? Is it noise with bad EMC with smd resistor VS the opposite with through hole resistor ?
For the digital side it doesn't matter what resistor you use, for the analogue there is a wealth of info on resistors, I believe the Blowtorch thread has just had a few pages of very good info regarding resistors and noise. I would have a look there, better qualified people than I have put forth there views, views which I will follow.
I'm actually in the process of redesigning a power supply for this project, so it'll run on external power which will result in a much cleaner power supply. It should be done in a couple of weeks, if you're interested, and the TPS60501 won't be needed, negating the difficulty you're experiencing sourcing the part. Hopefully that's helpful!
Hi! I tried to make my own version of one-sided pcb in eagle and here are results(still unfinished). I changed all decoupling capacitors on digital side of board to 0805 smd. I`m new to all of this so there may be some mistakes that I didn't see myself. Would be grateful for suggestions. EAGLE files are included.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.