LM2623 - voltage ripple

Discussion in 'The Projects Forum' started by boochbrain, Apr 25, 2013.

  1. boochbrain

    Thread Starter New Member

    Apr 25, 2013

    I am using the LM2623 to provide a 5V supply on a PCB design. Voltage input comes from a 3.7V lithium ion battery. I'm experiencing an unexpected voltage ripple on the 5V output that is .5V pk-pk. According to the data sheet, I should only be seeing less than 30mV. I've used the same schematic from application note AN-1274. Link is below.


    L1 and D1 are different parts as suggested in the application note, but have nearly identical specs. The parts I've used are below.





    What are the potential sources of voltage ripple on on the output of the LM2623? What should my approach be for debugging this? Could my PCB layout be the source of the issue?

    I have a degree in electrical engineering, but the inner workings of a boost regulator is not my forte. Maybe an explanation would help me debug the voltage ripple issue.

    Much thanks!

    “I have not failed. I've just found 10,000 ways that won't work.”

    -Thomas Edison
  2. crutschow


    Mar 14, 2008
    Possible reasons for excess ripple:

    Incorrect inductor value
    Incorrect output capacitor value
    Incorrect frequency of operation
    Too high output current load

    Layout could be a concern. Did you build it on a PCB per the layout shown in the data sheet?
  3. JMac3108

    Active Member

    Aug 16, 2010
    Are you sure the .5Vp/p is output ripple and not oscillation?

    To verify that what you're seeing on the output is actually ripple, look at it with an oscilloscope and make sure its frequency is the switching frequency of your boost, and that its somewhat of a triangle wave.

    If the frequency is much lower and it looks more like a sine wave, your boost may be oscillating. Remember that a switching power supply is a closed-loop system that can oscillate.

    [Note: After writting above section I looked at the app note and see that the part you're using is a hysteretic controller. These don't generally oscillate. But they have other challenges. Read through the datasheet and make sure all of your parts are correct, and match what they used in the app note]

    In any case, the most common three problems with switching power supplies are layout, layout, and layout! I've had one with a 1V oscillation on the output simply because I placed the required ceramic bypass cap for the internal VCC a couple mm away from the device pin.

    Are you sure your output caps are the right ones? Did you check their ESR compared to the ones in the app note?
  4. boochbrain

    Thread Starter New Member

    Apr 25, 2013
    Thanks for the reply. First off, yes I know it is a ripple, as I see a triangle wave.

    I'm fairly certain that all of my circuit component selections are at the right specs. I matched almost exactly to the application note. I am beginning to believe it is the layout that is causing this. My layout if very far from the design in the application note. What factors in the layout can cause excess ripple?
  5. crutschow


    Mar 14, 2008
    What is the output load current?
    Are you using the battery for the input?
    What is the ripple frequency?