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.

TPS65217C AC Input Glitch Handling

Other Parts Discussed in Thread: AM3352, TPS65217CEVM

Hello,


I am evaluating TPS65217C for an AM3352 based design. It utilizes AC and BAT supply inputs. During testing I have seen instances where AC input is connected but the unit either keeps running off of battery or goes into OFF state. ACPWR bit in the STATUS register (0x0A) is set in this state. The only way to recover it to start running off of AC and charge the battery is to cycle the AC input. My design is supposed to run off of AC most of the time and switch to BAT if AC fails. The system is unattended and I can't afford to have AC input go into this state.


From what I have seen the conditions that would put the PMIC into this state include:

 * Slow AC ramp up (Function of the external power supply)

 * Inrush current while trying to switch from BAT to AC

Is there a tested solution that would make the TPS65217C robust against AC input fluctuations?

Thanks,

Samir

  • Samir,

    The behavior with a slow ramp is expected, and documented in the datasheet:

    Inrush current should not cause any problems. Are you testing this on your own board or one of our evaluation platforms?

    If you are physically cycling the AC input, this can also be done through software. A write of <00011101> to register 0x01h will disable the AC switch and a write of <00111101> will re-enable it.

  • Hi Greg,


    I know that a ramp up over 50ms will put the PMIC into this mode. Unfortunately I can not guarantee this parameter as it depends on customer's power supply. All I want to know id whether TI has a design that recovers the AC input regardless of what sort of glitch is applied to it.

    I am testing with the TPS65217CEVM.


    In this fault mode the CPU is not powered up so there is nothing that can modify PMIC registers. An interesting thing is that asserting PMIC nRESET line doesn't seem to help.Accirding to the state diagram, asserting nRESET with PWR_EN high should bring the device into ACTIVE state.


    Thanks,

    Samir