• TI thinks resolved

TPS65070: power on fail

Part Number: TPS65070

I'm using the TPS65070 as Power-Manager of an IIoT-System.
As power-supply I use LiPo or/and AC. After booting the TSP EN_DCDC1 is high and in the CON_CTRL1-Register there is the default setting “01111111”
This means after pulling down PB_IN the DCDC1-Converter should start when EN_DCDC1 is high. (While pulling down PB_IN, POWER_ON is set high)
The first time after attaching a power supply the TCP boots properly and the DCDC1 supplies the configured voltage.
After shutting down with pulling POWER_ON to low a reboot of the TPS doesn't start the DCDC1.

How can I fix this problem?

Yours sincerely,

  • Hi Kamal,

    The device expert for this is out of the office for the holidays, please expect some delay.

    How are you rebooting the TPS65070? To confirm, the sequence is PB_IN pulses low, TPS65070 starts up, POWER_ON is set high by the processor (or pulled up to one of the TPS65070 rails) before PB_IN goes high again within 15 s. Then POWER_ON is set low again to shut the device OFF. When PB_IN is pressed a second time, the device does not start? Are there any I2C writes sent to the TPS65070?


    Kevin LaRosa

  • Kamal,
    Below is a brief description for the POWER_ON. The TPS will enter POWER_ON 2 state and start the rails according to the programmed sequence.

    Note: if the USB or AC inputs are removed you will have to restart with PB_IN. POWER_ON
    This pin is an input to the PMU which needs to be pulled HIGH for the PMU to stay in POWER ON_2-state once
    PB_IN is released. Once this pin is pulled LOW while PB_IN=LOW, the PMU is shutting down without delay,
    turning off the DCDC converters and the LDOs. If POWER_ON is pulled HIGH while there is power at USB or
    AC, the TPS6507x will enter POWER ON_2-state and start the DC-DC converters and LDOs according to the
    sequence programmed. See Figure 34.

    If this does not clarify or help your condition, please explain your startup sequence or restart in greater detail so that we may better understand the problem.


    Gordon Varney

  • In reply to Gordon Varney:

    hi, please find below feedback from customer:

    I will describe the problem again. I use the TPS65070 as power manager for an IIOT-System. Only one of the three Step-Down Converters is used (DCDC1). For the chip the default settings of power sequencing is 011, this means that DCDC1 should be enabled by pulling EN_DCDC1 high. When AC or USB are used as power supply, pulling PB_IN to GND, POWER_ON to high and EN_DCDC1 to high should start the DCDC1. But this doesn’t happen. DCDC1 is not starting. Only when a battery is used as single power supply this sequence leads to a starting DCDC1, but only for the first time. By a second starting sequence, the same problem occurs as for AC or USB power supplies.

    Can you tell me why DCDC1 does not start, even when all starting conditions are set.




  • In reply to Gordon Varney:

    again a more detailed explanation:
    That`s no problem.
    Unfortunately, the answer of your expert does not help my condition.
    I will explain my startup sequence more detailed.

    If only AC or USB voltage is connected:
    Pulling PB_IN to low  Pulling POWER_ON to high  release PB_IN to high  Set EN_DCDC1 to high. The DCDC1-Converter is not starting.
    Pulling POWER_ON to low.

    If only battery is connected:
    First startup after connecting battery-voltage:
    Pulling PB_IN to low  Pulling POWER_ON to high  release PB_IN to high  Set EN_DCDC1 to high. The DCDC1-Converter is starting.
    First shutdown:
    Pulling POWER_ON to low.
    Second startup:
    Pulling PB_IN to low  Pulling POWER_ON to high  release PB_IN to high  Set EN_DCDC1 to high. The DCDC1-Converter is not starting.
    Second shutdown:
    Pulling POWER_ON to low.

    The startup- and shutdown-sequence I use is the same as described in the datasheet of TPS650710.
  • In reply to Kamal Najmi:


      Can you send me a copy of your schematic for review. As you describe it the DC-DC1 should be starting. Feel free to send the schematic directly to me.


    Gordon Varney

  • In reply to Gordon Varney:


      I have been looking at your schematics. I don't see anything obvious, but I do have a few things to check. I colored them blue below.

    10.3.11 Enable

    To start up each converter independently, the device has a separate enable pin for each of the DC-DC

    converters. In order to enable any converter with its enable pins, the TPS6507x devices need to be in ON-state

    by pulling PB_IN=LOW or POWER_ON=HIGH. The sequencing option programmed needs to be DCDC_SQ[2..0]

    = 101.

    If EN_DCDC1, EN_DCDC2, EN_DCDC3 are set to high, the corresponding converter starts up with soft start as

    previously described.

    Pulling the enable pin low forces the device into shutdown, with a shutdown quiescent current as defined in the

    electrical characteristics. In this mode, the high side and low side MOSFETs are turned-off, and the entire

    internal control circuitry is switched-off. If disabled, the outputs of the DC-DC converters are pulled low by

    internal 250-Ω resistors, actively discharging the output capacitor. For proper operation the enable pins must be

    terminated and must not be left floating.

    Optionally, there is internal sequencing for the DC-DC converters and both LDOs available. Bits DCDC_SQ[0..2]

    in register CON_CTRL1 define the start-up and shut-down sequence for the DC-DC converters. Depending on

    the sequencing option, the signal at EN_DCDC1, EN_DCDC2 and EN_DCDC3 are ignored. For automatic

    internal sequencing, the enable signals which are not used should be connected to GND.

    Note: If there is a load on the SYS pin the DC-DC's wont go through a soft start.

    I have also seen a problem if PB_OUT and POWER_ON are not pulled up with SYS. I see that you are using the +3.3V_FPGA. As a test you might try to pull them up to SYS.

    Please take a look at these ideas and let me know what you find. I will continue to look over your information for other suggestions.


    Gordon Varney