• State TI Thinks Resolved
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 31 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

LMR36503-Q1: Efficient buck converter for very low load

Greg Marton
Prodigy 10 points
Part Number: LMR36503-Q1
Other Parts Discussed in Thread: LMR36500

Tool/software:

I am designing a DC/DC converter to go from 30-60V input to 3.3V to supply about 20 mA to a microcontroller and some other chips. Efficiency is key, because it needs to run for a very long time from a battery.

I choose the LRM36503-Q1 family, and in particular the LMR36503MSCQRPERQ1. I now realize that a LMR36500P3RPE would have been a better choice, the next one will be so. Anyway, I already have it built with the LMR36503MSCQRPERQ1.

It works, but I only get 45% efficiency when I would expect something over 60%.

I tried to calculate with WEBENCH, but I get bad results. The calculation says that the switching frequency will be 8.74 kHz, running in PFM mode. I don’t think it will actually lower the frequency that much, indeed, I measured 54 kHz. Also the calculation expects 90.6% efficiency, a wee bit too optimistic I think in any case.

Here is my design: LINK TO WEBENCH

What would be the ideal choice the input and output caps and inductor for this application? Is the LMR36500P3RPE indeed the best choice, or is there something that would be even better? 

Also, I get this voltage oscillation on the switch voltage. Why could that be? Is that a problem?

Thank you!

  • 0
    TI__Mastermind 19380 points

    Hi Greg, 

    The LMR36500 would probably be the better choice. 

    For a typical application, I would suggest a 2x 2.2uF capacitor, 100V rating (to account for derating), with a 100nF capacitor at the input. An output capacitor around 4.7uF would probably also be good enough, there is a table of external component values on the LMR36500 DS (table 8-1).

    The oscillation is the result of the device being in light load. After the low-side switch turns off, there is resonance between the parasitic inductance and capacitances that occur until the start of the next switching cycle. 

    Thanks,

    Richard