• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 63 subscribers
  • Views 1385 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.

My LM5010 is having a quazi unstable behaviour

John Lambert
Prodigy 50 points
Other Parts Discussed in Thread: LM5010

I am working on a LM5010 design taken primarily from Webbench and I am getting odd behaviour:

Yellow: Vout AC coupled

Green: Feedback pin AC coupled

The desired output is 12V and the behaviour changes depending on the input voltage:

@15V: around 10Vout@700kHz (toff ~270ns) - no issues

@20V: around 12.3Vout@1MHz (toff ~270ns) - no issues

@25V: around 12.3Vout@1MHz (toff ~350ns) - odd "stopping" behaviour shows up every millisecond or so and lasts for ~100 us

@40V: around 12.3Vout@1MHz (toff ~ 700ns) - stopping behaviour happens around every 200-300us.

A few things I have been investigating: I originally started with a non-shottky diode and based upon info from the forums, I switched to a fast schottky, though maybe not optimal (STMICRO STPS3H100U).  

The main things that would seem to cause this behavour in my mind are:

-tripping the 2.9V cuttoff (but the feedback voltage barely changes...)

-Overcurrent?  There is only around 100mA of draw on the circuit.

-Basic instability?

-outside of the shown circuit there is over 20uF of ceramic capacitance through the shown ferrite beads.  The beads may poorly combine with the caps to create a bad harmonic frequency?  Maybe the beads don't provide enough DC resistance to keep the caps from oscillating the output...

-The inductor is up against the power plane (12V) rather than a ground plane.  Could this cause issues?

-Maybe my layout itself is bad - I have a hard time assessing this completely.

Thanks for your help-

  • 0
    TI__Genius 11490 points

    It may be useful to look at the inductor current waveform. Is it possible that the circuit is in DCM? Does it happen at high load as well.

  • 0 in reply to Vijay86929
    Prodigy 50 points

    Yes, I was able to verify that the circuit in those cases switches to DCM (discontinuous current mode).  The voltage output (yellow - AC) and the SW pin voltage (green - AC) are shown below.  I modified the circuit to so that C64 (to couple the ripple into the feedback pin) was increased from 0.82nF to 4nF.  This increased the rolloff frequency from ~100kHz to ~25kHz.

    My understanding of ideal "discontinuous current mode" is that it is like a rock skipping - single pulses that spread out a little bit more when current decreases.  Is that correct or is the behaviour seen below of a few pulses separated by a longer wait what is to be expected.  Is the amount or ripple @ 2MHz when DCM is entered expected?

    Thanks,

     

  • 0 in reply to John Lambert
    TI__Genius 11490 points

    Hi John,

    The DCM behavior can be graceful degradation of frequency or burst mode depending on design. We think LM5010 has burst mode behavior.

  • 0 in reply to Vijay86929
    Prodigy 50 points

    Thank you for the input.  I was able to place a 150Ohm resistor on the output (~80mA) and the DCM mode went away.  As it is right now with or without the load, it is very usable.  Thank you very much.