• State
  • Locked Locked
  • Replies 23 replies
  • Subscribers 65 subscribers
  • Views 3364 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.

AM5728: PMIC(TPS6590377ZWSR) POWER GOOD signal pull low

user5358338
Prodigy 150 points
Part Number: AM5728

Dears

    We find the PMIC (TPS6590377ZWSR) "POWER GOOD" signal pull low to 0.2V(the low pulse width about 2.6us) when testing in -40° temperature on our AM5728 board. It is developed base on AM572X Industrial Development Kit. So, we tested the "POWER GOOD" signal on the IDK board(The "POWER GOOD" signal had no connect on IDK board, In order to keep the same as our board, we add a 10K pull high resistor to VRTC_OUT), also the  "POWER GOOD" signal would be pulled low to 1.25V(the low pulse width about 2.6us) when testing in -40° temperature on IDK board. Now we are checking the SMPS output of the PMIC if there is anyone abnormal？ At the same time can you give us some suggestion?

  • 0
    TI__Mastermind 49151 points
    Dear customer,

    Thanks for your interest in TI device. your question has been assigned to the device expert.

    Thanks!
    Phil
  • 0
    TI__Genius 10000 points
    Hi,

    POWERGOOD by default only monitors SMPS12. Have you unmasked any other rails or are you only monitoring SMPS12? If you are only monitoring SMPS12, it is important to understand what the current and voltage of SMPS12 are doing at the time that the POWERGOOD signal is pulled low.

    Can you measure both the current and voltage of SMPS12 at the time POWERGOOD goes low?

    Thanks,
    Nastasha
  • 0 in reply to Nastasha
    Prodigy 150 points

    Hi, 

      1.  The POWERGOOD control mode we followed the IDK board. And the IDK board also would be pulled low when we tested in the -40°.

      2. We tested all the SMPS rails, there weren't  pulled low.

      3. We can't measure the current till now. But we tested in the condition only insert the SD-card(boot from SD-card, enter the login interface, there weren't any software and peripheral device were running). 

    We hope you can give us more suggestion,thank you.

  • 0 in reply to user5358338
    TI__Genius 10000 points
    Hi,

    Did you use an oscilloscope? The SMPS rails don't have to be off for POWERGOOD to go low, it could be a voltage drop (~12% typical) that triggers POWERGOOD.

    Is the processor in reset when the POWERGOOD toggles?

    Can you monitor VRTC, SMPS 1_SW, SMPS2_SW, and POWERGOOD at the same time on the oscilloscope and save a shot of then POWERGOOD goes low?

    Thanks,
    Nastasha
  • 0 in reply to Nastasha
    Prodigy 150 points

    Hi, 

      1.  Yes ,we used the oscilloscope.  If the SMPS1/2 was dropped or off the POWERGOOD will be go low, is it right? 

      2.  POWERGOOD go low(only go low about 2.63us in low temperature), the processor hadn't reset.

      3.  We had measured all the SMPS rails and VRTC separately. They seemed be fine. but there was about 200mV noise on SMPS6 (VDD_GPU for AM5728) in -40°， and it was about 100mV when in 25°.

    Now, we are monitor the VRTC and POWERGOOD at the same time, the next step we will monitor the SMPS1/2 and the POWERGOOD as your suggestion. 

  • 0 in reply to user5358338
    TI__Genius 10000 points
    Hi,

    1. The POWERGOOD will go low if SMPS12 voltage drops or turns off or if the SMPS12 current is too high.

    2. From my understanding, you are saying that the processor is operating during this time. Can you take a scope shot of the current during the time that POWERGOOD goes low (include SMPS1_SW, SMPS2_SW, POWERGOOD, load current)?

    3. It would be very helpful if you could share the scope shots that you take so I can better understand what is going on in your system.

    Thanks,
    Nastasha
  • 0 in reply to Nastasha
    Prodigy 150 points

    Hi, 

        We can't measure the current till now. But we tested in the condition only insert the SD-card(boot from SD-card, enter the login interface, there weren't any software and peripheral device were running). 

    So, I guess that the load current should be not too high.

       The scope shots as below. The channel 1(Yellow) is POWERGOOD signal, the channel 2 is VRTC.

  • 0 in reply to user5358338
    TI__Genius 10000 points
    Hi,

    It would be still very helpful to understand what is happening to the load during the time that POWERGOOD goes low.

    Can you take a scope shot of the current during the time that POWERGOOD goes low (include SMPS1_SW, SMPS2_SW, POWERGOOD, load current)?

    If you can't include the load current, can you include the SW pins? It would be very helpful to understand if there is a load transient behavior happening at the time of this event. Ideally, looking at the load current would give use the most information.

    Thanks,
    Nastasha
  • 0 in reply to Nastasha
    Prodigy 150 points

    Hi,
       The scope shots as below. The channel 1(Yellow) is POWERGOOD signal, the channel 2 is SMPS1/2(VDD_MPU).  We are measuring the SW pin now.

  • 0 in reply to user5358338
    Prodigy 150 points

    Hi, 

     The channel 1(Yellow) is POWERGOOD signal, the channel 2(Blue) is SW PIN

    SMPS1_SW

    SMPS2_SW

  • 0 in reply to user5358338
    TI__Genius 10000 points
    Hi,

    Thanks for these scope shots. I am working on reproducing this behavior in our lab. I will try to provide an update by the end of this week.

    Thanks,
    Nastasha
  • 0 in reply to Nastasha
    TI__Genius 10000 points

    Hi,

    Can you confirm - does this only happen at cold temp? Do you see this behavior at room temp? How long does it take from when you turn the device on for the POWERGOOD to have this behavior?

    Thanks,

    Nastasha

  • 0 in reply to Nastasha
    Prodigy 150 points

    Hi, 

       Yes, it happened at -40°，it  will not be happened at room temp, even it hadn't happened at -20°. It will be happened in four hours when turn on the device. Sometimes, it will be happened in one hour.

  • 0 in reply to user5358338
    Prodigy 150 points
    Hi,
    Any updated for the problem? Thank you!
  • 0 in reply to user5358338
    TI__Genius 10000 points
    Hi,

    I apologize for the delay. I am working to understand what is going on in your system to produce this POWERGOOD event at cold temperature. Do you have load on the PMIC? Is the processor running?

    Can you capture the inductor current of SMPS1 and SMPS2 when the POWERGOOD event occurs? You can put the inductor on its side and add a current loop to measure the current.

    Thanks,
    Nastasha
  • 0 in reply to Nastasha
    Prodigy 150 points

    Hi, 

      Sorry, we can not capture the current in our lab. The PMIC provide the source for the processor as IDK board. We only boot from micro-SD card, after booted,waiting for login interface on Linux , there weren't any other device running.

  • 0 in reply to user5358338
    TI__Genius 10000 points

    Hi,

    Can you confirm - is your processor running during these measurements or is it in reset? If it is possible to mount the inductor (on the output of the SMPS1 rail) on its side and add a current loop, it would provide more information as to what is happening in your system.

    Thanks,

    Nastasha

  • 0 in reply to Nastasha
    Prodigy 150 points

    Hi,

      Yes the processor is running, the system is ready. The processor had not be reset. But the POWERGOOG signal is pulled down. It would be also happened on your IDK board under the low temperature(-40°) , can you reduplicate it? We are sorry about that we can't monitor the current in our lab. Thank you very much.

  • 0 in reply to user5358338
    TI__Genius 10000 points
    Hi,

    Unfortunately, without current measurements, we cannot verify why the POWERGOOD event is occurring in your system. POWERGOOD always monitors the current limit. Without a measurement of current, we will not know for sure if your system is reaching current limit briefly at cold temperature.

    Current loading on the PMIC is very dependent on the software being used with the processor, along with the computations the processor is doing. I cannot reproduce this scenario without knowing the load profile that your system is observing.

    I recommend not using POWERGOOD, or masking the POWERGOOD_SMPS1 monitoring function if you don't want to have this POWERGOOD event in your system design.

    Thanks,
    Nastasha
  • 0 in reply to Nastasha
    Prodigy 150 points

    Hi,

     We tested the "POWER GOOD" signal on the IDK board(The "POWER GOOD" signal had no connect on IDK board, In order to keep the same as our board, we add a 10K pull high resistor to VRTC_OUT), also the  "POWER GOOD" signal would be pulled low to 1.25V(the low pulse width about 2.6us) when testing in -40° temperature on IDK board. Here is the  scope shots.

  • 0 in reply to user5358338
    TI__Genius 10000 points
    Hi,

    I understand that you see this behavior on the IDK board. However, you would still need to measure the current to understand if the SMPS12 rail is reaching current limit. The current is dependent on the activity the processor is doing and the software running on the board.

    Thanks,
    Nastasha
  • 0 in reply to Nastasha
    Prodigy 50 points

    Hello all,

    We believe we have seen this same issue in our application and would like confirmation. We are using a 3rd party SoM on our custom carrier board. We noticed during testing that we could not get the SoM to boot in cold temperatures (this was observed around 3°C). Lab testing confirmed this on multiple boards. Through extensive troubleshooting, we found that we could resolve the issue by lengthening the rise time of our 3.3V supply to the SoM. (As alluded to in this TI document: http://www.ti.com/lit/an/swca302/swca302.pdf). Our default/previous ramp time was 1ms, and we resolved it by extending ramp time to about 80ms (far longer than the 90mV/ms requirement in that document). FYI, this 3.3V rail is powering VCC1 on the PMIC.

    I would like to specifically ask - has TI seen any issue at all with cold weather performance on this SoM/PMIC? Are you aware of any limitations regarding power supply ramp time aside from what is mentioned in TI documentation?

    Thanks,

    Stephen

  • 0 in reply to Stephen Newberry
    TI__Genius 10000 points

    Stephen,

    Thanks for providing information about your observed behavior at cold temperature. This current thread discusses a POWERGOOD signal going low at cold temperature, but there are no trouble booting the device. I believe these are separate issues or behaviors. For further support or questions, please start a new thread.

    Thanks,

    Nastasha