• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 3 answers
  • Subscribers 64 subscribers
  • Views 440 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LM2678: Datasheet clarifications

Jonathan Geronga
Expert 7930 points
Part Number: LM2678
Other Parts Discussed in Thread: LMR33620, LMR33630

Hi Team,

I would like to get your comments regarding the customer clarification with the datasheet of LM2678 as the customer seems seeing conflict information regarding his design:

The customer has 4 questions below:

The customer's design is a 5V output with a load that may vary between 1.3A to 2.0A max. The input voltage may also vary from about 9V to 12V. Figure 18 on datasheet gives a value of 22uH as the region for the majority of the variation in his design, but also touches into the area of a 33uH. Is it better to use a 22uH or a 33uH to cover the varying specifications?

Table 5 on the datasheet for a 5V output voltage shows C2, C7, and C4, for a 22uH inductor and C2, C3, and C4, for a 33uH inductor, however, in Table 1, the values for these capacitors differ between companies. Why are two different value capacitors listed?

Table 7 on the datasheet for a 5V output voltage shows C12 and C13 for both 22uH and 33uH, however, in Table 1, the values for C12 and C13 differ between companies. Why are two different value capacitors listed?

Also, Table 7 on the datasheet lists values for ‘Input Capacitors’, however, column two requires the inductance value. From their understanding, only the output of the regulator has an inductor. Is the table asking for the value of the output inductor or does a requirement to place an inductor on the input exist too?

Let me know if you have questions with the customer.

Thanks,

Jonathan

  • 0
    TI__Mastermind 28020 points

    Hi Jonathan,

    1. I would go with the 22uH. However, either the 33uH or the 22uH would be OK. Just be sure it can handle the current limit.

    2. The table gives different manufacturers for the convenience of the user. Each capacitor listed will have a different combination of capacitance and ESR. The range of those values/ESR combinations will work with this device.

    3. The different capacitors will have different ESR. You should choose a capacitor value in the range and with the lowest ESR.

    4. The table is asking for the output inductance.  This will set the input ripple current as well as the output ripple current.

    Have you created a design in Webench yet? This will give you the recommended values for each component.

    Regards, Jason

  • 0
    TI__Guru 72972 points

    Hi Jonathan,

    If the output current spec is only 2A, then the LM2678 5A converter is non-optimized.

    Please consider the LMR33620 36V/2A synchronous buck converter as an alternative that provides higher efficiency, better EMI, smaller size, WSON/SOIC package options, etc. This is footprint compatible with the LMR33630 3A version if the output current spec increases.

    Regards,

    Tim

  • 0 in reply to Timothy Hegarty
    TI__Expert 7930 points

    Hi Tim and Jason,

    Thank you very much for your support and comments.

    I will share this with our customer and get back to you for any response coming from the customer.

    Happy holidays!

    Regards,

    Jonathan

  • 0 in reply to Jonathan Geronga
    TI__Mastermind 28020 points

    Sounds good. Happy New Year!

    Thanks, Jason

  • 0 in reply to J Arrigo
    TI__Expert 7930 points

    Hi Team,

    Here are the follow-up questions from the customer upon sharing your answers:

    You stated the LM2678 will not be optimized at 2A. According to Figure 4 in 6-10, the graph shows the efficiency would be ideal at 5V, 2A, and then begin dropping as it approaches 5A. Am I missing something in the datasheet that specifies the regulator would be non-optimized?

    The LMR33620 has two packages, the HSOIC that has a thermal pad underneath, and the VQFN that doesn't have a thermal pad. If I understand the datasheet correctly, the VQFN dissipates all its heat through air, is this correct?

    Please let me know your comment.

    Regards,

    Jonathan


  • 0 in reply to Jonathan Geronga
    TI__Guru 72972 points

    Jonathan,

    The converter isn't optimized in terms of the inductor sat current is much higher than the nominal 2A requirement (it must be sized for the current limit level).

    The VQFN package puts most of its heat into the PCB. See the thermal performance metrics in the datasheet as a comparison. Efficiency will be must better with the recommended parts as the diode loss is avoided.

    Regards,

    Tim

  • 0 in reply to Timothy Hegarty
    TI__Expert 7930 points

    Hello Tim,

    Upon sharing your answer with our customer, here are some additional question below:

    Based on your response, regarding the optimization, you're stating that if the load is 2A, and I use the LM2678 with a 5A rated inductor, the regulator will not be optimized?

    What if the inductor was rated for 2A instead with the LM2678?

    Please let me know your comment.

    Regards,

    Jonathan

  • 0 in reply to Jonathan Geronga
    TI__Guru 72972 points

    Jonathan,

    The comment regarding optimization relates to using a lower current converter that matches the application requirement. This enables a smaller footprint inductor with inductance value to provide a ripple current of 30-40% full-load current.

    In general, the inductor sat current must exceed the current limit of the converter, hence a 2A inductor is not suitable for the LM2678. The switch current limit of the LM2678 is 7A typical at room temperature (see Figure 5 in the datasheet) -- the sat current should be well above this to provide adequate margin over temperature.

    Regards,

    Tim