• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 64 subscribers
  • Views 927 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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: How to Troubleshoot and Determine if there is a Problem with our AM335x Design?

Brian Weir
Intellectual 810 points
Part Number: TPS65217
Other Parts Discussed in Thread: TL5209

We are working on our second AM335x board design using a PMIC TPS65217.  The first design used a TPS65217B for DDR2 and we used RTC functionality for PMIC Power Enable.  Our new design has DDR3 so we are using a  TPS65217C and do NOT use RTC functionality or RTC-only mode per the "AM335x Schematic Checklist" RTC section (RTC feature disabled), the  TPS65217C has pin 9 PWR_EN pulled to 1.8v via 10k.

for the most part we have had good luck with these designs.  The new design is very similar to the BBB (except for RTC disabled).  We have had some TPS65217 modules go bad with VSYS (pins 7&8) usually outputting .8 to 1v.  If we isolate the VSYS by disconnecting the VIN loads the output does not change.  Also, after this happens with the VSYS and VIN isolated, if I tie 5v to the VIN connections the outputs do not come up.  It's like the module no longer responds.  We have six 10uf caps on the VIN inputs located around the TPS65217 close to each of the input pins.  I noticed that BBB uses only four 10uf caps.  I've noticed that if a board is repetitively power cycled, thats when the problem seems to appear.  I am curious if I have to much capacitance on the VSYS output?  Also, same as the BBB, there is a TL5209 regulator being supplyed by the VSYS output. This schematic design was reviewed by TI engineering.  So far, we have had four of the new board prototypes and on two of them the TPS65217C has appeared to have died.

As this design is still a prototype, I'm looking for whatever I can to make it more robust, any ideas?

Thanks,

Brian Weir       

  • 0
    TI__Genius 10885 points

    Hi Brian,

    I'm assuming there is no battery in this system, is that correct?

    How are the USB, BATT and TS pins configured?

    If there is no battery or USB, can you re-produce the issue with a pull-down on BATT, leaving TS floating?

    Best Regards,

    Rick S.

  • 0 in reply to Richard Solarski
    Intellectual 810 points
    Rick, Thanks for the response.

    No battery in either our older or new design.

    The Old (TPS65217B) design had no connection to USB or TS and the BAT1 & 2 lines are tied together with a 10uf cap to ground.

    The New (TPS65217C) design USB input is the same as the BBB "USB PC Connector" only nothing is populated, so other than being an antenna, the USB is open (floating), do you think that's a problem? The TS pin is not connected and BAT1 & 2 lines are tied together with the cap not populated. Basically nothing tied to all three.

    I can only try pulling down the BAT1&2 pins on an Old board (as the software team has 2 boards, the other 2 boards are on their way back to the assembler to swap out the TPS65217 chips).

    The BAT1&2 pins have -0.26v on them, when I tie them to ground with the power ON the SYS output does not change (5v). If I tie them to ground and then power ON the SYS output, and the rest of the PMIC, comes up normally. So, no luck recreating the issue on the Old board. I will be able to try this on the New board next week when they are returned.

    Do you think the extra 2 10uf capacitors (over the BBB design) could be causing excessive inrush that could be damaging the PMIC SYS output FETs? I don't really think it needs the extra capacitance, I was considering removing 2 down to 4 like BBB. Something is causing these TPS65217 chips to fail.

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

    Hi Brian,

    AC has current limiting on the input, so if the inrush is not exceeding this current limit I wouldn't be worried about damaging the SYS FET.

    I'd be more suspicious of the power-path becoming stuck due to the absence of a battery. The TPS65217x was designed for systems operating off of a battery, which is why it defaults to BATT when the power-path is disabled, and even determines if a source connected to AC or USB is valid by comparing to the voltages on the BATT pins. If AC is removed, without a battery we are technically reaching a UVLO fault condition, and if it is still in this fault state when AC is re-applied we can't guarantee how the power-path will react.

    When power cycling, AC may have to start below 100mV and rise to a valid voltage (>4.3V OR 190mV > VBATT) within 50ms to be detected. This is a combination of the T_rise parameter and Vin(DT) parameters in the power-path detection limits section of the datasheet.

    Because of this, if AC is not fully discharged before re-applying the AC supply there is a chance it might not be detected as a valid rising edge - leading to this stuck position.