4
\$\begingroup\$

I'm designing a board with LAN8720 and LPC1768.

LPC's ADC and LAN's magnetics connections require analog 3.3V power.

Should I separate these two analog power planes with a ferrite as I separate it from digital 3.3V power plane? Or is it OK to use only one analog 3.3V power plane?

I edited this message and added below picture. Ethernet PHY(LAN8720) to magnetics connection and LPC1768 ADC is using 3.3V analog power. Is it better to keep them separate as I draw in the picture?

enter image description here

\$\endgroup\$
2
\$\begingroup\$

Form what I read: "The digital ground pins (GND), the analog ground pins (AVSS), and the GND_CORE pins on the LAN8720 QFN are all connected internally to the exposed die paddle ground. "

So use a separate analog ground for your ADC as you mentioned above. Maybe it will work if both share the same ground plane, but it's easier to omit a ferrite if you don't need it than add one in if you do.

\$\endgroup\$
  • \$\begingroup\$ Dirk, as suggested in this article hottconsultants.com/pdf_files/june2001pcd_mixedsignal.pdf I will use one ground plane, but I will keep analog and digital traces in a distance so they won't interfere. However my question is about 3.3V analog power plane. I edited my question and added a picture. \$\endgroup\$ – zenprogrammer Jul 28 '14 at 12:25
0
\$\begingroup\$

I'd do it like this: -

enter image description here

Note that I've shown 0V (ground) and 3.3V power going straight to the noisy object (the digital block) and from where the connections hit this block I would tee-off to separate power and ground planes for each of the ethernet block and analogue block.

Reason: I don't like digital currents in a ground plane getting any further than they need to and this means minimizing digital ground currents (or 3V3 currents) from upsetting the ethernet and ADC sections.

\$\endgroup\$
  • \$\begingroup\$ I general I find the power leads are much easier than the ground return. If the Analog part needs to be very low noise, then I'd run it's ground return straight back to the power supply. (The only times I've had issues with the supply voltages was when I mistakenly used the supply as a reference voltage in a design.) \$\endgroup\$ – George Herold Jul 28 '14 at 14:04
  • \$\begingroup\$ @GeorgeHerold the problem occurs with ADCs - a lot of ADCs have to have a digital ground split very close to the chip and there is a danger that tying 0V back to the power supply may cause problems. \$\endgroup\$ – Andy aka Jul 28 '14 at 14:12
  • \$\begingroup\$ Oh, OK I was just looking at that piece of wire going from all the ground connection back to the power supply and thinking that all the digital and ethernet currents flowing in that bit of wire are going to couple into the analog section. (Not a big effect.) \$\endgroup\$ – George Herold Jul 28 '14 at 14:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.