• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 719 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.

TPS61021A: "Feed Forward" capacitor/ Light Load instability

Bret Herscher
Prodigy 50 points
Part Number: TPS61021A


I am using the TPS61021A in a dynamic application (line and load changes) and have been experiencing output instability at very light loads. This doesn't seem to be related to the layout as I have replicated the condition on the EVM. It appears to be related to the "feed-forward" capacitor. (why they call this a feed-forward cap is a mystery to me btw). My total output capacitance is 70 uF (nominal) so, following the data sheet ... "For large output capacitance more than 40 μF, it is recommended to set the zero frequency (fFFZ) to 5 kHz." I diligently set the zero to 5kHz which happens to be 100pF and 320k in my design.

This seems to work well with moderate and heavy loads, but at light loads (1k Ohm at 3.3 V  = 3.3 mA) I can make the regulator produce all kinds of oscillatory behaviour. For example if I make the input voltage a triangle wave ramping from 3 Volts to 1.8 Volts and back again with a 100 Hz periodicity the regulator rings like a bell at around 3kHz  with a peak to peak amplitude of about a volt centered at the 3.3 Volt set point.

Just removing the "feed-forward" capacitor eliminates this.

Decreasing the load resistor (to 100 Ohms / 33mA) eliminates this.

Is the data sheet correct? Web Bench suggests a much smaller capacitor (20 pF) and the EVM has smaller still (10pF and 316k). There really isn't enough information in the data sheet for me to do my own phase margin analysis and I don't really trust web bench so I am stuck here, especially since the behaviour does not completely disappear with the Web Bench recommended value.

Has anyone else experienced anything like this? Is there any way to get more details of the insides of this part?

Any help appreciated,

Bret Herscher

  • 0
    TI__Mastermind 46905 points
    could you share the waveform you observed? with and without Cff?
    what is part number of the output capacitor? we may also try to reproduce the behavior in the EVM.
  • 0
    TI__Mastermind 40080 points

    Thanks Jasper.

    Hi Bret,

    Your output current is very small, so the TPS61021A works in the PFM mdoe at this load condition. please send the schematic and the SW waveforms to us so that we check what happened.

  • 0 in reply to Jasper Li
    Prodigy 50 points

    Jasper (and Helen),

    Many thanks for your replies ... I have found and fixed my problem and it is somewhat embarrassing ....

    Turns out that my technician did not do a good job of connecting the exposed pad to ground. Since this is the only power connection this was interfering with the regulator .... Amazingly enough the part does a good job of regulating at high currents with only the Analog Ground (AGND : pin 1) connected in PWM mode. In PFM mode the part also regulates but the behavior is oscillatory at some input voltages.

    The reason it happens on the EVM also was because I had my tech replace the part on the EVM and he made the same mistake on that.

    Once the Power Ground is connected the part is rock solid in both PWM and PFM modes of operation.

    Many Thanks,

    Bret

  • 0 in reply to Bret Herscher
    TI__Mastermind 40080 points
    well done. thanks for the update.