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.

TPS62130 not working because soft start pin has low impedance between 100Ohm-15kOhm

Other Parts Discussed in Thread: TPS62130, DCP020515D

We are using the TPS62130 circuit for supplying four DCP020515D (external clocked) with +5.4V out of +12V computer power supply. The output is pre biased by a schottky diode from +5V of same computer power supply. This path is used for delivering the high startup current (>4A) of all DCP020515P. The TPS62130 will be delayed (10ms) enabled via RC circuit from external clock of DCP020515D and delivers the normal load current (<2A).

Generally it works, but up to 10% of TPS62130 did not start after an indefinite time. Reason: The soft start pin has low impedance (100Ohm-15kOhm), also if it is unsoldered and measured directly between soft start pin and GND. If we apply a 10k resistor between SS pin and output, most of the not working devices will start and are working normally!!! If so the power path is not damaged! The output draws no current, so internal switches are off if device is not working!
We made many tests with hot plug and unplug of input voltage, overloading, overheating (not soldered exposed pad), several 1000 start cycles but we cannot destroy it in that way!  Not all devices are affected but it is not acceptable when 10% stops working after indefinite time!

What is going wrong here?

Ciruit:

Layout:

Startup behavior with 10ms delay:
CH1: Switching node of TPS62130
CH2: Output current via 51mOhm shunt (instead L60)

  • Thanks for posting your schematic and layout. Don't ever install the resistors on the DEF and FSW pins as you pull them up to Vin. Those pins are not rated for the input voltage, so they need to be pulled up to Vout if you want to tie them logic high.

    It looks like you tried to post a waveform. Can you try and repost it?

    Could you also share a waveform of Vin, Vout, EN, and SS? Could you share these waveforms on a damaged unit and normal unit? If you power cycle Vin, does a failing unit restart?
  • Many thanks for fast reply! We didn´t install any pull up resistors on DEF and FSW, we know that!
    Sorry for missing waveform, here it is:

    Startup behavior with 10ms delay:
    CH1: Switching node of TPS62130
    CH2: Output current via 51mOhm shunt (instead L60)

    Start up at working sample:

    Start up at not working sample:

    Soft start pin has approximately 4.5kOhm and voltage is around 10mV!
    If we insert 10k resistor between outpout and soft start pin, it is working normally!

    A power cycle did not help, it has 4.5kOhm between SS and GND permanently, also at desoldered device!

  • Thanks for posting these.

    I understand that you can still measure a low resistance on SS on a bad unit, but it would still be a good idea to put this unit on an EVM and see if it behaves the same with the SS voltage not increasing.

    How have you detected the 10% failures? What type of testing is occurring when one of these devices fails?

    Could you post a waveform of Vin and Vout when the 12V input supply is removed?
  • Start up of working sample with SS pin connected to GND:

    Input impedance is 1MOhm on all channels!

    Remove of +12V input:

    CH1: +Vin (12V)
    CH2: +Vout (5.4V)
    CH3: Softstart
    CH4: Enable
    Measured directly at TPS62130, input impedance of measurement unit is 10MOhm on all channels!

    Connection of +12V input:

    Connection of +12V input, zoomed:

    We use TPS62130 in our measurement unit and 3 of 100 fails at first checking or calibration.
    If boards pass the calibration they are stored und build in later. At final checking up to 2 of 18 have same error!
    Most devices will start working with 10k resistor between SS and Vout, depends on SS impedance!
    All defective devices have high impedance on SW pin, if not supplied!

  • Hmm, it seems clear that you have damaged the SS pin somehow. Mounting a failing unit in a known good setup (on an EVM) would confirm that the IC is damaged and not something on the board.

    I would prefer that we remove the cause of the damage, rather than adding a resistor to Vout to overcome the damage.

    Is there any reason why you choose a 10 nF SS cap? It is rather large. You might try reducing it to 1 nF or so and seeing if you have more or less failures.
  • One more thing: is the 5V on the output ever removed? If so, can you show a picture of Vin and Vout when it is removed?
  • Yes the ICs are damaged and we do not know why!
    We do not add a resistor to overcome the damage, currently we exchange the IC!
    But we have 800 TPS62130 in use, most one are at the customer side and changing all and hoping everything is good is no solution!
    We want to know the cause of damage but if did not find we have to install the resistors. Is a dirty solution but even better than not working TPS62130! According to datasheet there is no limitation of SS capacitor value! Most applications in datasheet uses 3.3nF, we used 10nF because we had good expierences at prototypes and it already used on board. Are there maybe any troubles at discharging circuit inside IC?
    We measured maximum derating of 0.1V/us at SS pin with 10nF-> Maximum discharging current of 1mA, should be no problem, or?
    And we will test it with higher SS capacitor (100nF) and on/off cycles, then it should happen more often if so!
    Removing the output is not easily possible but we can measure hot off of measurement unit!
  • We need to find out what action is causing the device to fail and then capture the waveforms on a scope.

    Aside from ESD or mechanical damage, pulling Vin more than 0.3V below the SS pin voltage could damage the SS pin. This is what I would investigate as your apply/remove/add power to the system.
  • Wolfgang, have you ever solved the problem??? I'm highly interested as I'm currently facing the same problem.