• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 33 subscribers
  • Views 795 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.

TPS2552 reversion protection issue

Bernd Eifer
Prodigy 30 points
Other Parts Discussed in Thread: TPS2552

We use the attached USB power circuit with TPS2552DRV. An external 5.5V connected to the USB power rail while powered up and enabled creates a reverse current flow that is NOT blocked by the device and no FAULT being indicated. After a disable/enable sequence current remains zero and FAULT becomes active.

This behaviour appears to be in contradiction to the datasheet of TPS2552.

Any idea of what is wrong?

While shorted at the output and current being limited by the device as desired there is also no FAULT indication appearing.

Appreciate any feedback,

Bernd

  • 0
    TI__Mastermind 42795 points

    Can you measure the reverse current under this condition of output biased with 5.5V? Also, when 5.5V is applied to the output and reverse current flows, what is the measured voltage at the TPS2552 IN pin? The reverse current needed to trip the comparator is ~135mV / Rdson or about 1.35A. The input supply must be able to sink as well as source current (such as a synchronous buck would do).

  • 0 in reply to Eric Wright
    Prodigy 30 points
    Eric,
    5.38V vs. 5.35V. I've never thought about the input pin just following the output with the consequence of the trip voltage never be met. Silly me.
    MY OTHER QUESTION: Do you see a reason why a short should not be reported at FAULT?
    Many thanks,
    Bernd
  • 0 in reply to Bernd Eifer
    TI__Mastermind 42795 points

    Hi Bernd,

    No worries, achieving the reverse current fault can be a difficult system level event.

    Regarding the FAULT reporting, is it possible to capture some waveforms showing IN voltage, IN (or OUT) current, and FAULTx pin? There can be cases where even if the IN voltage is stable that the ripple increase (the source dc/dc converter ripple) can cause a behavior of this sort. How do you have the EN pin tied? Make sure it is at a stable low level. If the root cause is ripple, you may be able to address it by changing the IN pin bulk capacitance.

  • 0 in reply to Eric Wright
    Prodigy 30 points

    Eric,

    now I succeded in pasting our schematic for reference. Your input has made me think more about inductance and resistance in our supply path, and tiny "IN bulk capacitors". The testbed has been our CPU board with supply through pretty long (> 1', AWG 26) wires.

    Having said this, I found both negative and positive spikes on the IN supply rail but no ripple. And there's not much IN capacitance (refer schematic).

    After this, I moved the board onto our power supply, i.e. it's final destination. Here we have just a B2B connector only and 1000uF bulk capacity on the PSU's side.

    Now I receive FAULT in both cases a) shortcut b) resistive load with 8R2.

    My conclusion: TPS2552 needs a "sufficiently" capacitor-stabilized IN power rail for both over-current signaling and reversion protection. It is definitely not sufficient to place 100nF and there is a good reason for a 10uF cap in Figure 5 page 8 of the datasheet ;-) .

    Thanks a lot for great, inspiring input.

    Is it possible for you to tell what amplitudes of short-term noise/spikes TPS2552 can tolerate on the IN pin without becoming confused? This is about economics and about being on the safe side.

    -Bernd