• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 1904 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

DAC38RF82EVM: ferrite beads vs three terminal capacitors

vic zap
Intellectual 740 points
Part Number: DAC38RF82EVM
Other Parts Discussed in Thread: DAC38RF83, LMK04828

Hi!

Is it possible to get schematics of the previous revisions of the DAC38RF8xEVM? Also, can you tell me please, were the EVM tested with ferrite beads on FB22, FB24, FB26 and so forth... instead of three terminal capacitors? Maybe previous revisions had ferrite beads on that places which were subsequently replaced by the three terminal capacitors due to inadequate performance or something else?

Regards, Vic.

  • 0
    TI__Mastermind 18065 points
    Vic:

    The 3-Terminal caps are actually a ferrite/cap filter. The device we use provides good selectivity at around 2 MHz. They are specifically useful for power rails that are driven from DC-DC converters where the switching spurs are around 2 MHz. We tend to propagate those all over the place, but normal ferrite beads are sufficient in most cases. I do recommend the 3-terminal device on rails driven directly from DC-DC converters.

    --RJH
  • 0 in reply to RJ Hopper
    Intellectual 740 points
    Hi, RJ Hopper!

    Thanks for the reply.

    Nevertheless, I'd still like to dwell on 3-term caps you've used to decouple the DAC38RF83 supply rails.
    Consider 1.0V supply powered by U12. As we can see, here we have a common supply, which powers separate DAC rails (+0.9V_VDDL1, +0.9V_VDDL2, ... etc). As I understand, these rails are implied to be isolated from each other by means of the 3-term caps (FB22, FB24, ... etc), which essentially are improved caps. This isolation is needed to prevent noise generated by one rail from coupling into another. But, as the common supply is a low impedance net, such a capacitive isolation of the separate rails (looking from the DAC) has not as much effect as it could be with ferrite beads instead of 3-term caps (see www.murata.com/.../c39e.ashx, paragraf 3.4, p. 17). So, can you give some comments concerning this?

    Also, I'd like to ask about decoupling the LMK04828 (U2). Its pins VCC7_OSCout, VCC8_OSCin, VCC9CP2 are throttled by FB7, FB8, FB9 respectively and are not provided any capacitance to compensate for the throttling. Any change in the current demand on these pins will lead to a voltage drop on the FBs. It absolutely contradicts to common decoupling approach, doesn't it? Can you explain such a decoupling, please?

    Regards, Vic.
  • 0 in reply to vic zap
    TI__Mastermind 18065 points
    The 3-term device is a not just a cap...it is a ferrite bead with a cap. Schematically I think the equivalent circuit is ferrite bead -> shunt cap -> ferrite bead. That said, there is no data to define absolute isolation requirements between those (and similar) rails. The use of the 3-term devices was considered good design practice. I think it is also ok to replace with traditional ferrite bead and cap barring the DC-DC switching spur issue that I mentioned previously.

    On the LMK, there is a ferrite plus caps feeding the net that is ultimately split to other device power pins. An additional ferrite was placed on each of those nets that feeds the specific pin. This approach is more of a contingency / debug tool. There is no issue with combining those nets directly; however, for debug purposes we liked to have a way to break the connection to feed in an external supply or to measure current. We could have simply used a 0 ohm jumper but throwing in a ferrite seemed like it wouldn't hurt.

    --RJH
  • 0 in reply to RJ Hopper
    Intellectual 740 points

    Thanks a lot for the fast replies! It's clear.

    Regards, Vic.