• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 534 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.

TPS40211 SEPIC Compensation - Switching Noise on COMP

Chris Davies
Prodigy 60 points
Other Parts Discussed in Thread: TPS40211

We have designed a 10 - 20V input 14 V 1A LED driver using the TPS40211 in a SEPIC configuration. Similar to the TI application note. We have found poor stabilisation of the output current, drifting with time and temperature ( 10% variation ). On investigation this has been due to a large component of switching frequency ( 1V at 600 kHz ) on the compensation pin. The output current has low ripple ( < 50 mv ). Changing the compensation components and lowering the crossover frequency by 10 or even 100 times has no effect at all. The only changes that seem to have any effect are the FET gate series resistor or the output diode type.

What mechansims are there for getting switching frequency components on the output of the error amp other than from the feedback voltage itself?

 

  • 0
    TI__Guru 56775 points

    Chris,

     

    The TPS40211 controller does not have any intentional method for forcing the output ripple on the error amplifier output (COMP pin)

    When you adjusted the crossover frequency, what components did you change?

    The typical control loop contains 2 capacitors and a resistor between the FB and COMP pins.  Generally, the larger capacitor is placed in series with the resistor and the smaller capacitor is directly across FB and COMP.  At the switching frequency, it is generally this smaller capacitor that is setting the gain.  If you are not changing the ratio of the Vout to FB impedance to this smaller capacitor value, you're not changing the error amplifier gain at the switching frequency.

  • 0 in reply to Peter James Miller
    Prodigy 60 points

    Thanks

    Yes I changed both capacitors, by a factor of x10 ( and even X100 ) and it had no effect at all on the COMP ripple. The 2 things that did make a large change ( possibly removed COMP ripple component completely ) were :-

    1) Increasing the series gate resistor from 0R to 10R. Why does this effect stability? Coupling gate edges back into Isns pin?

    2) Changing the output diode. Both Schottkys, one higher power than the other. The low power diode 3A is fine but the 6A diode generates large switching components at the COMP pin. Why should the output diode effect switching frequency component on COMP pin? For space reasons i need to use the 6A device as its lower profile.

    Chris

     

  • 0 in reply to Chris Davies
    TI__Guru 56775 points

    Chris,

     

    These now sound like noise issues with switch edge noise coupling into the error amplifier output.

    Can you share an oscilloscope waveform of the Gate Drive and COMP voltages?

    What are the values of feedback components are you using?  Please include everything from Vout back to COMP.