• State TI Thinks Resolved
  • Locked Locked
  • Replies 13 replies
  • Answers 1 answer
  • Subscribers 66 subscribers
  • Views 5345 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.

TPS65217: Issue with power-up, only 0.7V @SYS

Eric Schmieder63
Intellectual 400 points
Part Number: TPS65217
Other Parts Discussed in Thread: AM3358, LM2596, , TPS5430

hello everyone,

I have a custom board similar to the BBB. When I try to power it up with 5V at the AC-port, it dont power up completly. I checked SYS(Pin 7, 8) but got only 0.7V.

PWR_EN is directly connected to the AM3358.

Thanks very much for your help.

Best regards

Eric

  • 0
    TI__Genius 10885 points

    Hi Eric,

    Are you seeing the rails turn on temporarily and shutting down, or just never turning on?

    The AM3358 should be asserting the PMIC_PWR_EN pin to keep the PMIC active. Can you measure the PMIC_PWR_EN voltage?

    If the PMIC is not responding at all, you may want to check the 5V slew rate, which needs to be under 50mS for the AC or USB inputs to recognize a rising edge.

    Best Regards,
    Rick S.

  • 0 in reply to Richard Solarski
    Intellectual 570 points

    Hi Rick,

    We have the same problem on our board. When we provide 5V to TPS65217C's AC pin, the sys is 0.6V, while if we provide 4.5V to AC pin, the sys's output is proper and everything works fine. 

    The PMIC worked with 5V input, but now it does not. What cause the PMIC's overvoltage protection's threshhold voltage dropped ?

    Regards,
    Shu

  • 0 in reply to SHU LIU
    TI__Genius 10885 points

    Hi Shu,

    I can't explain why the over-voltage threshold would decrease, how consistent is this behavior?

    When you slowly increase AC does the power-path switch at a specific voltage?

    The power-path also references the voltage on the BATT pins, which can cause unexpected behavior. Can you measure AC, BATT, SYS, and PWR_EN simultaneously with an oscilloscope?

    Best Regards,

    Rick S.

  • 0 in reply to Richard Solarski
    Intellectual 570 points

    Rick,

    Thanks for your reply.

    We double checked circuit and it was not because of the over-voltage threshhold.

    I checked your earlier post that talks about the slew rate. Will that be the problem ?

    We use a DCDC converter LM2596 (150KHz switching frequency) to provide 5V input to the AC pin of TPS65217C. We scoped the 5V output from LM2596 and its clean +5V and not with so much noise.

    We have around 10 sample boards. Some boards can be powered up from LM2596 to TPS65217C with no problem. Some boards can sometimes be powered up and sometimes cannot be powered up.

    PWR_EN pin of TPS65217C is connected to one IO of am335x.  

    Will the slew rate of 5V from LM2596 be the problem that causes TPS65217C sometimes not starting up and sometimes starting up?

    Regards,

    Shu

  • 0 in reply to SHU LIU
    TI__Genius 10885 points

    Hi Shu,

    Yes, T-rise must be less than 50mS or the device may not power up.

  • 0 in reply to Richard Solarski
    Intellectual 570 points

    Hi Rick,

    How to make sure that T-rise is within 50ms from LM2596? There is an input cap (10uF) at AC pin.

    Regards,

    Shu 

  • 0 in reply to SHU LIU
    TI__Genius 10885 points
    Hi Shu,

    I'm not familiar with the expected rise time of that device, if you measure with a scope is it significantly over 50ms? Reducing output capacitance should help, and if the input supply is current limited during startup it might help to delay enabling the LM2596.

    Best Regards,
    Rick S.
  • 0 in reply to Richard Solarski
    Intellectual 810 points

    HI Guys...

    I too have an AM335x / TPS65217 custom board, i'm getting 0.8v @SYS.  I have the PWR_EN input 10k pulled up to 1.8v VLDO1 (VDDS).  

    I'm confused about the 5v input slew rate being mentioned in this post, first is "...check the 5V slew rate, which needs to be under 50mS for the AC or USB inputs to recognize a rising edge."  later it's "...if you measure with a scope is it significantly over 50ms?"   I am supplying he AC input via a TPS5430 switcher set to 5.2v, the output slew rate for this supply is 8ms.  Is this good?

    Thanks,

    Brian Weir 

         

     

  • 0 in reply to Brian Weir
    TI__Genius 10885 points
    Hi Brian,

    Since the rise time is significantly less than 50ms it should be fine. Were you able to solve this issue?

    Thanks,
    Rick S.
  • 0 in reply to Richard Solarski
    Intellectual 810 points

    Richard,  thanks for the reply.

    Our Issued, I believe was due to ESD/EOS damage to the AM335x.   We replaced severeal TPS65217s' on different boards that had the issue.   I  don't believe changing the PMIC ever fixed anything for us.  The more antiquated I get with the TPS65217, the more I'm realizing how bullet proof it really is.  The only way we have been able to fix this problem is to replace the AM335x, and measuring power supply rails to GPIO pins reveal shorts.  We are assuming ESD damage

    Thanks, Brian 

    .        

  • 0 in reply to Brian Weir
    TI__Mastermind 30310 points
    ,

    Do you believe the Root Cause of your problem is also related to an AM335x EOS issue, as described by Brian Weir?
    Have you tried swapping out the TPS65217 PMIC with a new IC on a couple PCBs are re-testing?
  • 0 in reply to Brian Berner
    Intellectual 570 points

    Brian ,

    The issue that SYS is 0.6 or 0.7V is still not resolved on our side. It only happens when we use AC input to LM2596(5V) then TPS65217C. There is still a 50% chance to power up AM335X. However if we feed 5VDC directly to TPS65217C, AM335X powers up every time. Therefore, I do not think AM335X is ESD damaged on our side. 

    I don't think the rise time longer 50ms is the reason since 50ms is a very long period. We checked with oscilloscope.

    We did not try to use a different PMIC or regulator yet. 


    Regards,
    Shu

  • 0 in reply to SHU LIU
    TI__Mastermind 30310 points

    can you please probe the PWR_EN pin under both conditions: When you apply 5V directly to PMIC and when LM2596 applies 5V to PMIC?

    In the original reply to the thread, asked this question to someone else but you did not answer it for your issue:

    "The AM3358 should be asserting the PMIC_PWR_EN pin to keep the PMIC active. Can you measure the PMIC_PWR_EN voltage?"

    Can you also measure the voltage on the AC pin during the power-on sequence? Does the LM2596 voltage applied to the TPS65217 show any glitches or exhibit non-monotonic behavior around the UVLO trip point of the TPS65217 AC pin?

    An ideal power supply will always be monotonic unless the Current limit is reached, but a power supply from another DC-DC converter may go up and dip once or twice before the power supply rail is good. If LM2596 glitches or voltage dips during VAC ramp, then this could be the problem.

    We will need to see scope shots attached in your answer to assist you further with this issue.