• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 62 subscribers
  • Views 748 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.

LM27762: PFM and constant switching frequency for light load

Alexandre Lavanchy
Prodigy 10 points
Part Number: LM27762


Dear Sir,

My design uses the LM27762 with a 5.0V input and +4.5V / -4.5V outputs. According to the data sheet, I have 1uF (0603 X7R) for C1 and 4.7uF (0805 X7R) for Ccp. The load current on OUT- is varying between 20mA and 80mA. 

The circuit works fine so far. My only concern is that the LM27762 is always in PFM mode, with a cycle frequency around 60 to 100kHz depending on the load current. The transition between PFM mode and constant switching produces a large ripple, which is a problem for my circuit. Furthermore, I need to shift the switching frequency higher, far away from the bandwidth of a sensitive signal.

How can I make the LM27762 increase its switching frequency and stay in constant switching mode? I would be glad to use Webench to simulate the circuit under my specific load conditions, but the LM27762 has been removed from the simulation tool. So, could you help me to find the correct parts values in order to stabilize the charge pump for constant switching under these conditions of light load?

Thank you, best regards,

Alexandre

  • 0
    TI__Mastermind 21580 points

    Hi Alexandre,

    LM27762 is not supported by Webench as it is an older device. However a PSPICE transient model is available on the product folder.

    The only option I see to increase the switching frequency is to add an extra resistive load on the output. Would this work for you?

    Best regards,

    Varun

  • 0 in reply to Varun John
    Prodigy 10 points

    Dear Varun,

    Thanks for your reply. Adding resistive load is possible, although I would prefer to avoid it in order to save power. I did some tests, and I have to raise the output current from 20mA up to 100mA in order for the LM27762 to reach continus switching. What is your opinion on lowering C1 and Ccp values, is this risky regarding stability or transcient response?

    Best regards,

    Alexandre

  • 0 in reply to Alexandre Lavanchy
    TI__Mastermind 21580 points

    Hi Alexandre,

    Reducing the C1 and Ccp maybe by a factor of 2 can help reduce the amount of resistive load you require. I don't see it going into continuous switching without the extra load. The transient performance could be a bit worse though with lower C1 and Ccp values.

    Also you can add the resistive load to Ccp instead of at the LDO output. This will help slightly by saving on the losses through the LDO.

    Can you let me know what your end equipment is and what is the bandwidth of your sensitive signal? Would be helpful for us to understand your needs so that we can feedback this information to our new product development.

    Best regards,

    Varun

  • +1 in reply to Varun John
    Prodigy 10 points

    Hi Varun,

    Many thanks for your feedback. Adding extra load directly to Ccp in order to preserve the LDO is a very good idea, and I will try it. I will also try reducing C1 and Ccp by two.

    This is for an audio product. The measured bandwidth is 22kHz. For now, the noise content produced by the LM27762 in the 60kHz - 100kHz is too high and produces aliasing issues. Measuring the signal with an audio analyser shows this noise back in the low frequency bandwidth, which gives poor measurements figures.

    Best regards

    Alexandre

  • 0 in reply to Alexandre Lavanchy
    TI__Mastermind 21580 points

    Hi Alexandre,

    Thanks for the information.

    Let me know if you need any further support.

    Best regards,

    Varun