MicroZed randomly resetting

Could it be that your design exceeds the power available through the USB port? 

By USB specification a single USB port can only supply a maximum of 500 mA @ 5V. By the time you power the PS section of the Zynq device, the DDR memory and the other components on the MicroZed board there is a limited amount of current left to power the PL section of the Zynq, any loads on the USB-Host interface, and anything plugged into the PMOD connector.

If this is a possible source of your resets there are several ways to provide additional power to the MicroZed.

Use a USB Y-cable, with two connections to the Host for a 1A source.

Populating a DC barrel jack (J4) and providing 5V there.

Plug the MicroZed into a carrier card to provide additional power.

-Gary

I am seeing the same results on 2 microZed Rev C boards that I have.  It appears stable when I run one with the default linux image connected to the usb port of my pc.  If I load aa fpga image in this configuration, it seems less stable (restart happens about the frequency llj describes. 
I have built a carrier board and when I run my microZed on this card, it happens VERY frequently (every 20s to couple of minutes).  I'm not really sure how to go about diagnosing this problem to fix it.  Are there any other reasons the microZed would reset besides power fluctuations?

Does your carrier board supply power to the MicroZed board? Have you estimated the power requirements of the MicroZed including the FPGA image you are loading. Have you monitored the power rails during the restart event to see if they are staying stable?

There are a couple of signals on the microheaders to the carrier card that can cause a reset. Take a look at the MicroZed Carrier Design Guide here for details: http://www.zedboard.org/documentation/1519

If you are able to share your design I can try it on a MicroZed powered by our IO Carrier Card to see if we can observe the restart event in this configuration.

-Gary

Yes, my carrier card supplies power to the microzed.  I have a 24V->5V 3A Texas Instruments power supply that I supply to the Vin headers.  I have not done a formal power estimation yet, as informally I calculated 3A as overkill.  (Note that the resets I'm seeing also occur when booting from QSPI with the manufacturer-installed boot image.)  I have monitored the power rails, but not when I've seen a reset.  What levels of fluctuation could cause a reset? 

I've looked through the documentation.  It seems that PG_MODULE, PWR_ENABLE, and CARRIER_SRST# and possibly INIT# could cause a reset.  (Am I missing any?) PG_MODULE, INIT# and CARRIER_SRST# are floating.  PWR_ENABLE is coming off my 5V power quality monitor.  I just noticed that my VCCO34 and VCCO35 are not enabled by PG_1V8 as recommended.  Could that cause this problem? 

Having the VCCO34 or VCCO35 supplied before the PG_1V8 may cause a situation where the Zynq device can be damaged over repeated power cycles, but I don't think it is likely to be the caues of the restarts you are seeing.

I will contact you offline to discuss your specfic MicroZed + carrier board.

-Gary

Has anybody found a solution? Please share!

Hello Olebon,

Are you experiencing the same issue? Are you seeing the issue on multiple boards? Under what conditions do you see this issue?

--Josh