Frequency Stability

Question

This question is related to this previous question. I am still trying to investigate the issues I am having on some boards. Section 7.4.2 of your data sheet (Crystal Characteristics) does not call out a frequency stability requirement, only frequency tolerance. What is the worst case frequency stability that you recommend using in designs. Thanks, Nick

jeongsk · Answer

Hello Sorry about the late reply. Tell your circuit has been modified points. FB4 0 ohms C29 1uF C32 1uF C34 1uF There are four changes. See the attached circuit. Thanks W5200E01-M3 REV115_20140625.pdf (157 KB)

newaters · Answer

Jeongsk, thank you for your reply. This reply seems to be for my other post, this thread should probably be merged with the other thread so that future readers can understand what we’re talking about. Can you still please confirm the frequency stability of the oscillator? I have some follow up questions to your reply: [ol] [li]What prompted the upsizing of some of the decoupling capacitors, why aren’t the .1uF decoupling capacitors adequate?[/*] []Right now the ferrite bead and capacitors form an LC low-pass filter on the 3.3V analog rail. If the ferrite bead is removed there is no longer a filter, the rails will be directly tied together and there will just be more decoupling capacitance. What’s the reason for removing this ferrite bead?[/] []Can you please confirm which pins the 1uF decoupling capacitors should be closest to? [ul] [] On my PCB, C34 is located closest to Pin 11, C36 is closest to pin 15, and C35 is closest to pin 23.[/] [] C29 is closest to pin 25, and C30 is closest to pin8.[/] [] C32 is closest to pin 27, and C33 is closest to pin 47.[/*] [/ul][/li] [li] What is the Maximum power supply ripple (peak to peak) allowed on the 3.3VA and 3.3VD rails? Your data sheet does not call this out.[/li][/ol]

jeongsk · Answer

Hi newaters. I will answer your questions. 1.What prompted the upsizing of some of the decoupling capacitors, why aren’t the .1uF decoupling capacitors adequate? Answer : This revision had on the inside of the W5200. So is that the amendments to fit inside the property. 2.Right now the ferrite bead and capacitors form an LC low-pass filter on the 3.3V analog rail. If the ferrite bead is removed there is no longer a filter, the rails will be directly tied together and there will just be more decoupling capacitance. What’s the reason for removing this ferrite bead? Answer : This question is also the same as the number one answer. This revision had on the inside of the W5200. So is that the amendments to fit inside the property. With this revision, the stable probably should not use the LC Filter. 3.Can you please confirm which pins the 1uF decoupling capacitors should be closest to? On my PCB, C34 is located closest to Pin 11, C36 is closest to pin 15, and C35 is closest to pin 23. C29 is closest to pin 25, and C30 is closest to pin8. C32 is closest to pin 27, and C33 is closest to pin 47. Answer: decoupling capacitor is to close to the power supply input pin. You are right. And decoupling capacitor is your choice. We guide the minimum decoupling capacitor. Make you want to add. 4.What is the Maximum power supply ripple (peak to peak) allowed on the 3.3VA and 3.3VD rails? Your data sheet does not call this out. Answer : What is the meaning of the power supply ripple? Would mean the noise ripple? Or are you referring to the power input range? Note the power input range is 3.3V +/- 10%.

newaters · Answer

[quote]Answer : What is the meaning of the power supply ripple? Would mean the noise ripple? Or are you referring to the power input range? Note the power input range is 3.3V +/- 10%.[/quote] I am using a buck converter that has a fixed switching frequency of around 450kHz. I am seeing about 3.7% ripple, 124mV peak to peak.