• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 81 subscribers
  • Views 2759 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.

Bypass capacitor question on the MSP430F5529

BuffaloEngineer
Intellectual 680 points
Other Parts Discussed in Thread: MSP430F5529, MSP-EXP430F5529LP

On the schematic for the MSP430F5529 launchpad, they have these bypass capacitors: 

DVCC1: (10uF), 100nF - 10uF is DNP

DVCC2: (10uF), 100nF - 10uF is DNP

AVCC: 10uF, 100nF

What is the reason for the difference in the bypass caps, why the DVCCs have the 10uF caps as DNP?  I would think that they would keep the capacitance on all power lines.  

Didn't find a quick link to the schematic: 

http://www.ti.com/tool/msp-exp430f5529lp

  • 0
    TI__Mastermind 20945 points

    Hi,

    Which caps are you referring to?  Can you provide a component number?  The 10uF cap on DVCC1 is C30, which is populated on every 5529LP I checked.  From the schematic, I only see 100nF caps on DVCC2 (C14) and AVCC (C29).

    Mike

  • 0 in reply to Mike Pridgen
    Intellectual 680 points

    I must have mixed up my schematics, but the same question applies, why the difference in the bypass caps on DVCC1 vs. AVCC and DVCC2?  Is DVCC1 supposed to power up slower?  

  • 0 in reply to BuffaloEngineer
    Guru 25210 points

    The 10uF capacitor is the bulk decoupling capacitor for the +3.3V power rail. It supplies bulk current demands for all the VCC pins.

    Additionally, the 0.1uF capacitors at Dvcc1, Dvcc2, and Avcc are local high-speed noise decoupling capacitors, and need to be located close to the device pins with short (low inductance) traces.

    Internally to the MSP430, the Dvcc1 and Dvcc2 pins are connected together. There are multiple supply pins to the part to help with I/O supply demands (same with ground).

    Avcc is separate, and you will see has the option for a filter resistor or ferrite bead for decreasing digital noise in sensitive analog applications. The LP board populates it with a 0-Ohm resistor that you could change if needed.

  • 0 in reply to Brian Boorman
    Intellectual 680 points

    Thanks very much!  One follow up question - there's a 220nF cap on V18.  Is this considered a bypass cap as well for the internal 1.8V supply?  

  • 0 in reply to BuffaloEngineer
    Guru 25210 points

    BuffaloEngineer said:
     Is this considered a bypass cap as well for the internal 1.8V supply?

    That's exactly what it is. VCORE as well. Be advised of the note in the datasheet that you cannot externally load V1.8 or VCORE

  • 0 in reply to Brian Boorman
    Intellectual 680 points

    I'm interested in swapping the 470nF capacitor VCORE for 220nF to reduce the unique component count.  

    If my input capacitance to DVcc is 4.8uF, the ratio of Cdvcc/Cvcore > 10 is still preserved - would there be any stability problems using a lower capacitance value?  

  • 0 in reply to BuffaloEngineer
    Guru 12050 points

    As Mouser does not list NanoF, I hate that there is no standard to not use it.

    0.47uF vs 0.22uF, you should be OK. even better put two 0.22uF as your are halving the ESR.
    Only time you need to be spot on, is if it used in a RC filter, timing circuit or radio tuning.

    10uF vs 4.7uf, same here can you go with two 4.7uF instead?

    In the 1mhz+ range the mcu need small amount of power but really fast, so it's important that decoupling caps
    are really close as even at the near speed of light on a 1/4" trace is to long of a time.

    When the caps are above 1uF, they are most likely used as bulk cap (like a short term battery)
    And if you start having brownout, measurable voltage drops when put a few of them on varies spot on the pcb.

  • 0 in reply to Tony Philipsson
    Guru 227245 points

    Tony Philipsson said:
    0.47uF vs 0.22uF, you should be OK

    http://e2e.ti.com/support/microcontrollers/msp430/f/166/p/331278/1235744.aspx#1235744

    http://e2e.ti.com/support/microcontrollers/msp430/f/166/p/1235744/Reply.aspx

    The 470nF on VCORE (as well as the 220nF on V18) are the output capacitor of the switching regulator. Changing its value will decrease regulator efficiency (increase power consumption), decrease regulation speed and maybe even lead to oscillation.

    You read the datasheet wrong. 470nF is mandatory, and DVcc capacitance must be at least 4,7µ (as this is what the regulator needs at least on its input when ramping up) or more.
    If Cvcore can be 470nF+10% (due to cap tolerance), then you can’t use a 4.7µ cap on DVcc, as it might be 4.7-20% while it needs to be >5µ.

    Properly interpreting datasheets is an art. :)

    However, your suggestion of using two 220nF instead of 470nF sounds reasonable. It should be within the required range. Unless the lower ESR causes oscillation (which I've seen happening on a stand-alone switching regulator, where the datasheet indeed specified a minimum ESR).

**Attention** This is a public forum