• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 3 answers
  • Subscribers 62 subscribers
  • Views 313 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.

LMR16010: Does slope compensation depend on inductance?

David Dean
Expert 2305 points
Part Number: LMR16010
Other Parts Discussed in Thread: LMR16030, LM63610-Q1

Does the slope compensation of the LMR16010 depend at all on the switching frequency and inductor value?  I have seen some other buck regulators that specify a range of recommended inductor values based on the switching frequency, output voltage, etc.  The datasheet only calculates the inductor based on ripple current.  In my application I may use an inductor that is relatively large or relatively small.  Is there any risk of problems with sub-harmonic oscillations or other instability if the inductance is either too high or too low?

  • 0
    TI__Mastermind 30990 points

    Hello David,

    In this device, slope comp is not dynamic. You are correct in slope comp determining the inductance required in the case that duty cycle exceeds 50%, in addition to the power conditions.

    Often calculating inductance for 30% ripple ensures stability in regards to SHO. It will be tough for me to give you an analytical equation to describe the inductance needed to avoid SHO, but, I can give you my review on your selected inductance and consult the responsible IC designer in the case I find your selection alarming.

  • 0
    Expert 2305 points

    At the moment it's looking like Fsw = 700KHz and L = 4.7uH or 6.8uH.  I could potentially go down to L = 2.2uH if safe to do so.

    Designing for a particular ripple is difficult because my load current can vary.  So if I design for 30% ripple @ 1A, the ripple will become much higher (as a ratio) for a lower output current.  How does that end up factoring into the 30% target?

    How much of a problem is SHO?  Can it lead to total instability and loss of control?  Or will the LMR16010 maintain a constant Vout, maybe with some additional ripple due to the SHO?

    Let me know your (or the IC designer's) thoughts on this.

  • 0 in reply to David Dean
    TI__Mastermind 30990 points

    Hello David,

    Either 4.7uH or 6.8uH should be satisfactory.

    The corresponding inductance calculation is often recommended to be done at the max rated DC load current for the device, not the application rated current.

    SHO can most certainly be a reliability issue.

  • 0 in reply to Marshall_Beck
    Expert 2305 points

    Ok, so in this case the 30% ripple would be based on the 1A DC rating, or 300mA?  

    Does that then follow for the higher-rated versions of this regulator?  If I instead used the LMR16030 would I select an inductor for a ripple of (3A * 30%) = 900mA?  

    Do you have a rule of thumb for how small of an inductor (or how large of a ripple) would start to run into SHO?  

  • 0 in reply to David Dean
    TI__Mastermind 30990 points

    Hello David, yes 300mA for 1A regulator. 600mA for 2A regulator and so on.

    I have yet to contact the designer on minimum inductance. I am not optimistic they will be able to give me a firm equation to follow, but I will see.

    I believe it is best to follow the conservative calculation in the datasheet (select inductor for 30% ripple of DC rated load current).

  • 0 in reply to Marshall_Beck
    Expert 2305 points

    My max Vin is 30V and Vout is 6V.  L=4.7uH results in a ripple current of 1.4A, much more than 30%.  In terms of stability and SHO do you still think that inductor will be ok?  

    It gets interesting - while the ripple current may be very high in this situation, the duty cycle is less than 50%, so SHO is not a concern, correct?  The duty cycle won't be > 50% until Vin is 12V, which results in a ripple current of 900mA.  Still higher than 30%, but at least reduced from the peak value.

    Can you give me more details on the risks of SHO?  Many datasheets (including LMR16010) don't explicitly mention it at all.  While there are some application notes available that talk about selecting an inductor to prevent SHO, not many discuss what would happen if a circuit did exhibit SHO - the one I was able to find says

    "This effect is not harmful to the regulator and in many cases it does not even affect the output regulation. Its most annoying effect is to produce a high pitched squeal from power components".

    Is this true?  If I were to run into SHO would it just be a matter of some audible noise?  Or is it a bigger problem than this application note is implying?

  • 0 in reply to David Dean
    TI__Mastermind 30990 points

    Hello David,

    I dont want to lead you down the wrong path so thinking about this more, I am going to err on the side of caution.

    For minimum inductance calculation, I advise you calculating for it at the device's nominal vin for the device's max rated DC load current.

    I wish I could give more insightful guidance on this, but many of the original team members on this project cannot be contacted.

    LM63610-Q1 specifies min inductance per slop comp. 6.8uH would be acceptable for your power condition.

    __

    Yes, > 50% duty cycle. Here is an article from one of my colleagues at my site on this topic. See slope comp seciton https://www.ti.com.cn/cn/lit/an/snva555/snva555.pdf

    Often, it results in irregular pulsing at a fraction of the switching frequency. I believe that to be the definition. Theoretically, it can fall in the audible range. Nevertheless, it is an unstable control loop which is a reliability concern.

  • 0 in reply to Marshall_Beck
    Expert 2305 points

    Thanks for the details.

    It seems like, no matter where I end up for inductance of frequency, it would be a good idea to test for SHO in the actual application.  Is that just a matter of monitoring the inductor current while sweeping through Vin?  Should I set the output current to the maximum, or would I need to sweep through that too?

  • 0 in reply to David Dean
    TI__Mastermind 30990 points

    Hello David,

    I would sweep across the recommending operation condition for the device, be that vin range and iout.

    Full Iout potentially may derate the inductance as well, temperature. This can be "simulated" without extensive test setup perhaps just by selecting a lower inductance to account for derating as well, tolerance.