• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 641 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.

TPS54626: Start up issues

Brett Gidge
Intellectual 370 points
Part Number: TPS54626
Other Parts Discussed in Thread: TPS54540

I'm having start up problems with the TPS54626. I used Webench to design the switcher supply. The layout closely resembles that recommended in the datasheet. The load has a pretty big range. It ranges from 10mA to 3.5A. I just had 4 prototype boards built. All exhibit the same issue but two of them require a higher load current to see the issue. I believe that my issue is an output undervoltage lockout problem (I don't know why because I'm operating the chip within its parameters). The chip latches when this condition occurs which is really bad. I monitored the feedback voltage (Vfb) and the Soft Start voltage. The feedback voltage seems to ramp up to around 530mV and then slowly climb towards the typical 765mV level. When the board fails it simply shuts down at about 1.7 times the soft start time which is why I think it is the output undervoltage lockout kicking in. Note, if I reset the switcher via its enable pin then it comes right up. After an enable pin reset, the feedback voltage ramps straight up to the 765mV (it doesn't go to the 530mV and then slowly ramp to 765mV). What should I look at to try and fix this issue?

Here is an o'scope capture of the voltages (Yellow = Vfb, Blue = Soft Start pin) when it fails:

Here is an o'scope capture of the voltages (Yellow = Vfb, Blue = Soft Start pin) when I reset the switcher using its enable pin (note no power cycle):

Here is an o'scope capture of the voltages (Yellow = Vfb, Blue = Soft Start pin) when the board doesn't fail (lighter load):

Note, the bottom horizontal cursor is at 530mV and the top horizontal cursor is at 770mV.

I appreciate any help!!

  • 0
    Intellectual 370 points

    Oh here is my switcher circuit:

  • 0 in reply to Brett Gidge
    Intellectual 370 points

    Ok the schematic didn't post for some reason. Let's try it again.

  • 0 in reply to Brett Gidge
    Intellectual 370 points

    Another piece of data...I removed my zero ohm jumper that is on the output of the switcher (R223 in the schematic I provided). I then added a purely resistive load (about 2.5 ohms so about 2A) to the switcher output. I'm basically removing my circuit board load to make sure it isn't doing anything weird. Well, the switcher behaved the same way. It wouldn't come up. Then if I toggle the Enable pin it comes up fine. This switcher is useless if it can't start up with its rated load.

  • 0 in reply to Brett Gidge
    Intellectual 370 points

    Yet another piece of data...I think the problem is the 12V rail. The 5V rail starts switching when the 12V rail feeding it gets to right around 5V. It pulls the 12V rail down some which I think is causing a type of brown out condition but one that isn't low enough for the switcher to reset. Here is an o'scope capture. The yellow trace is the 12V rail and the blue trace is the 5V switcher feedback pin, Vfb. Ideas of how to fix this? Note, the TI 12V switcher I'm using, TPS54540, doesn't have a power good output.

  • 0 in reply to Brett Gidge
    TI__Mastermind 18510 points

    Hi, From your description, output UVP should be triggered during 12Vin rising after soft-start end.

    Suggestions:

    1, as you said, after 12Vin ready, toggling EN, output will be normally. add divider for EN pin, after Vin ramp up higher value, then EN will be enabled.

    2, make soft-start time longer

    3, also suggest to double check 54540 output 12V performance, such as transient response, power up ability with loading.

    Yuchang

  • 0 in reply to Yuchang Zhang
    Intellectual 370 points

    Hi Yuchang,

    I ended up adding a divider on the enable pin that consisted of a zener and a resistor. That worked!! It delays the 5V switcher startup by like 40msec which is plenty. The switcher can now start up with the heavier load. Thanks!

    Brett