Actions
Forum Thread Details
  • State Verified Answer
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 175 subscribers
  • Views 296 views
  • Users 0 members are here
Related

VCCIO_EN and PWR_EN Shutdown Sequencing for Custom Carrier Card

sgfuller31

Hello,

 

I am in the process of making a custom carrier card for MicroZed. I have a question on the proper shutdown sequencing order, the MicroZed Carrier Design Guide reads on pg. 17  "To enable power to the MicroZed, PWR_EN must be pulled high. PWR_EN is tied to JX1 pin 5 and is pulled up to Vin on the MicroZed. To shut down power to the MicroZed, PWR_EN and VCCIO_EN should be pulled low. VCCIO_EN should be pulled low first to maintain proper shutdown sequencing. See the Power Architecture and Sequencing Diagram for more detail."

 

From the MicroZed Breakout Carrier Card schematics I can see how this is readily implemented with an RC delay circuit going into an NMOS transistor for the PWR_EN signal while the VCCIO_EN has no such delay capacitor. (Sheet 3 of 5). My problem is that this circuit relies on a "5V_OFF" signal. This signal is driven to 5V when the switch is turned off. In my application however, the 5V supply will turn on with the rest of the system (an autonomous vehicle of sorts that I'm building), and then turn off when the overall system is powered down. So in other words, there will be no extra 5V line that receives power (5V_OFF) once the main 5V line is powered down.

 

Is there an example on how to implement proper shut down sequencing if all of the power to the board is cut? Likewise, I suppose I should ask the question then of what would happen with the MicroZed and the Breakout Carrier card, being powered off of USB, if the USB cable is suddenly unplugged? (Switch still turned "on"). That would seemingly prevent the proper shutdown sequencing since the "5V_OFF" signal would not be driven high.

 

Thanks for your time and help

  • 0

    If power is removed altogether you won't need to worry about the sequencing. We discovered a backfeed issue where in some cases the bank voltages weren't shutting down unless you pulled that VCCIO_EN signal hard low. Basically we had to make sure that customers who would still be powering the SOM asserted a hard "off" to that line as there were situations where if left floating the signal would not always shut down the regulators. For those situations we had an issue where pulling low PWR_EN would shut down the core and auxiliary rails, but if you didn't deliberately pull down VCCIO_EN as well that voltage could potentially float high enough where it did not shut down the bank voltages. Result was that the core and aux rails were powered down but power remained on the banks, that's bad.

     

    Long story short, removing power to the SOM will eliminate any potential issues with this sequencing. Power down sequencing is obviously much tougher since something must stayed powered to do a controlled shut down. As such, most of our designs assume total power shut down simultaneously.

    Thanks

    Chris

  • 0 in reply to ctammann

    Hi Chris,

     

    Thanks for your response. I think that makes good sense to me. I'm just going to repeat this back to make sure I understand it clearly.

     

    • The reason why those extra pieces of circuitry exist is to support the ability to leave the SOM powered on, but to be able to turn the bank voltages for the PL IO on and off. If the extra circuitry didn't exist the core and aux rails could be powered down while the bank voltages remained high which would be a problem. This was due to the backfeed issue you stated earlier.

     

    Since in my application I plan to always use the PL IO in conjuction with the SOM I can likely simplify the supporting power schematic as in the following image. (Taken from sheet 3 of the MicroZed Carrier Breakout Schematic). This should ensure that I still follow the proper startup sequence order, but will not need to use the shutdown circuitry. I suppose this will also mean I will not need to use the PWR_EN signal since it is pulled up on the MicroZed SOM.

     

    Let me know if you see any issues with this plan. Thanks,

     

    Sam

     

    image

  • 0 in reply to sgfuller31

    Here is a diagram i created when documenting the issue:

    image

    In your scenario if 5V goes away completely then there is no concern about a portion of the system remaining powered and causing a sequencing issue. The key pieces to keep in mind though are VCCIO_EN (1.8 PGOOD in this diagram) is used to enable the bank voltage supplies on the carrier. The carrier then releases the PGOOD when the carrier power supplies are stable. Does that make sense? If you follow the power up sequencing (power to SOM, use VCCIO_EN to enable bank voltages provided from the carrier, PGOOD to the SOM released when all carrier supplies are up) then removing power from the whole system you will not need to worry about power down sequencing.

  • 0 in reply to ctammann

    Yes, that makes sense to me. Thanks again for all of your help.

     

    Sam