• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 64 subscribers
  • Views 753 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.

Original question:

LM5118: Possible load regulation issues at high end of transition mode

LM5118: Possible load regulation issues at high end of transition mode

John Vranes
Intellectual 590 points
Part Number: LM5118

Hello,

My original thread has been locked, but I wanted to ask a related question.  I've had a few more devices exhibit the behavior again as described in my original thread, where in response to a load transient (going from about 100mA to 7 ohms), the output voltage dips a few percent and stays there until the load starts to decay, at which point the output drifts back up before spiking by several volts. 

It's occurring under similar operating conditions as before, with a Vin of about 28V (towards the upper end of the transition region of operation (i.e., just below full buck mode)), Vout of about 21.4V, Iout of about 100mA and 7 ohms (about 3A), with no other particular actions applied to the circuit before the failure other than simple power ups at about 100 to 200mA of loading.

I've yet to replace the ICs and see if the behavior goes away again.  I expect that it will, but in the meantime, are there any other ideas as to how this type of behavior may be occurring?  Like before, I'm wondering if it's just a bad IC, or if there's some sort of tolerance in the IC that my design might be susceptible to.

Thanks,

John

  • 0
    TI__Guru* 78270 points
    I don't remember ever seeing your schematic. Could you share it so we can have a review.

    Thanks,
    Youhao
  • 0 in reply to Youhao Xi
    Intellectual 590 points

    Sure, Youhao.  I've attached it to this reply.  I forget what exactly I did last time to make it so it was visible to you, but I'll see if this works.  Let me know if it doesn't work and I'll try a different way of loading the graphic.  A couple notes on the schematic:

    1)  As specified in the datasheet guidelines, SGND and GND are connected together at a single point.  Their connection is not shown in the schematic but is implemented in the layout.

    2)  The high-resistance feedback connection to EN is implemented in order to prevent the circuit from shutting down during brief under-voltage transients.  I should note that the circuits exhibiting the behavior in question were not subjected to under-voltage transients.

    Thanks,

    John

  • 0 in reply to Youhao Xi
    Intellectual 590 points
    Hi Youhao,

    Have you had a chance to review the schematic? Are you able to view it?

    Thanks,
    John
  • 0 in reply to John Vranes
    TI__Mastermind 21720 points
    Hey John,

    Is it only at VIN = 28V that you see this issue? Reviewing the schematic I don;t really see anything that would cause an issue as you describe. Can you please post the picture of the load transient? It would be good to also probe the COMP pin and the inductor current if possible.

    Thanks,

    Garrett
  • 0 in reply to Garrett Roecker
    Intellectual 590 points
    Hi Garrett,

    Thanks for the reply. This behavior has occurred at input voltages ranging from about 27.4V to 28.2V. Granted, at the highest end of that range, the voltage doesn't dip the few percent, but it still has an excessive spike when removing a load, compared to normal behavior. The original thread has some more information and waveforms of the regulator output. However, I don't have the load transient itself captured. I can say that it's a fast on transition (in my benchtop tests, the heavier load is turned on by me connecting its return lead to ground while the regulator was running, even intentionally causing come bounce to see if the regulator handles it normally without excessive spikes, which it does at input voltages higher or lower than the range specified above) and a slow off transition (taking a few milliseconds to decay to zero)

    I'll work on getting a waveform of the load transient and perhaps the COMP pin. I don't have a way to get the inductor current. May not be until next week sometime. What kind of sampling rate are you looking for on the COMP pin? This test is normally over several seconds, so a great sampling rate isn't as feasible, but I can try capturing snippets at a time if needed to increase that.

    Thanks,
    John
  • 0 in reply to Garrett Roecker
    Intellectual 590 points

    Hi Garrett,

    I wasn't able to get specific data from a failing circuit, as it turned out that I no longer had access to one that had failed.  But I was able to still capture the load transient that leads to this behavior.  It's a fast on and a slow off transition.  A slow off transition scope capture would probably have had a short spike in its current due to the spike in regulator output voltage, but again, I was not able to capture that.  Anyways, here are the rising and falling load transients.  Vertical axis units are Amps.

    Thanks,

    John