• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 64 subscribers
  • Views 1033 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LP8860 feedback, Pulse skipping, high ripple and noise issue

Orlando Grigoriadi
Prodigy 130 points

Hello to everybody, 

I am using LP8860 to drive 3 strings of LED in my design. The schematic can be seen in the attached figure.

The input voltage is an automobile battery (13.5V nominal but 4.5 - 24V are also possible).

My load is a fixed number LED, with a abs max voltage of 18.5V and abs max current 65mA per string.

My question is the following. When the input voltage was close to the output voltage for certain loads (for example 24mA per string) i heard audible sound. With the oscilloscope i captured the following signals.

1. The Output voltage (LED_A) ripple (AC coupled)

2. The driving GD signal.

When there was audible sound the output voltage ripple was high.

The driving GD signal was a train of pulses (which made the output voltage go high) and then a long time (about 80usec) there was no switching activity. Then the train of pulses was present once more.

First i would like to ask weather this is a feedback error of not. Could this be characterized as pulse skipping?

Then i changed the boost Pi compensation control, integral and proportional from 2 to 4 and the audible noise was lowered, the ripple too. The train of pulses was more frequent. The time between the train of pulses was lowered to about the half (40usec) for the same load at the same input voltage.

I would like to ask if this could cause a stability problem and how could i test the stability.

Then i was thinking of also lowering the capacity connected to the feedback pin. Could this cause also stability problems?

I have chosen the most short off time and blank time options.

Could you advise any change which could increase stability and decrease ripple? 

Thank you in advance,

Orlando Grigoriadis

  • 0
    TI__Genius 11020 points

    Hello,

    When the boost input voltage gets close to output voltage the boost enters PFM mode. You will see it as a pulse skipping when checking the GD voltage. These pulse bursts can cause increased ripple. It is not actually instability what you see, but it is just the way the PFM mode works.

    You can change the behavior with the boost parameters, you could try adjusting also the LLC comparator threshold (PFM entry point) to see if it lowers the audible noise. Increasing output capacitance helps reducing the output ripple.

    Measuring stability (phase margin)in PFM mode is not possible, only in PWM mode.

    Best Regards,

    Tomi Koskela

  • 0 in reply to Tomi Koskela
    Prodigy 130 points

    Thank you very much for your reply. Indeed, increasing the capacity decreases sound and ripple. I increased the input capacitor by 330uF and the output cap by 220uF, to see how the circuit behaves at high capacitor values.


    I would like to ask the following question: By monitoring the GD output, the current on the Rsense resistor (in my schematic R91) or the current on the inductor L3 i see that the signals are not periodic, even in lower input voltages (for example 11V), when PFM is not present.

    Is it some kind of sub-harmonic oscillations the reason for non periodic signals? The waveforms at a constant input and a constant load (65mA per channel) should be normally periodic, should they be not?

    By changing the artificial current peak value (BOOST_SEL_IRAMP and BOOST_SEL_IND bits) and the Rsense i could resolve this issue? If not could you please advise.

    Thank you very much