Zybo-Z7 QSPI mode does not work.

[mhmt9851]

Hello,

While working with Zybo Z7 board, i accidently connected 5V to 3.3V power pin of Pmod port JE.
Similar to these threads;

https://forum.digilent.com/topic/17695-zybo-z7-10-qspi-flash-programming-mode-broken/
https://forum.digilent.com/topic/18202-zybo-z7-will-not-power-on/
https://forum.digilent.com/topic/21497-zybo-z7-pgood-led-is-not-on-no-devices-detected-on-target-in-vivado/

At first the pgood led was flashing, and Vivado did not detect the board. I did not use the board a couple of days.
Now in jtag mode pgood led stays on and Vivado detects the board.

I can program the device in jtag mode.
pgood led stays on in qspi mode but the oob demo program does not boot. Sometimes it works, sometimes it does not, and sometimes pgood led turns off and instantly turns on.
in sdcard mode the oob demo works. Taken from this thread : https://forum.digilent.com/topic/18148-zybo-z7-20-factory-demo-program/

I think the qspi flash works, because sometimes it boots correctly and the oob demo starts.
So the issue must be in the power regulator. Can you help me troubleshoot the issue. Maybe i can
replace some parts and fix it.

I measured the voltages of capacitors from the linked thread;

C219, C220, C221, C222 are 4.53V
C228, C229 are 3.26V
C237 is 1.34V
C240 is 1.79V

 

[JColvin]

Hi @mhmt9851,

Based on what you described, I'm personally leaning towards the QSPI flash acting up rather than power regulator (though it could be a combination of a number of things). The reason for this is because you are able to program the board successfully over JTAG and the SD card; if the power regulator was at fault, you would experience inconsistent configuration as many of the FPGA / SoC banks are also powered by the same 3.3 V supply that the flash memory (and FPGA configuration) uses.

The power good led status being inconsistent when jumper JP5 is set to QSPI is odd though. The jumper grounds the SPI_SQ2/MODE2 pin on the flash memory and otherwise doesn't affect the supply voltages which is what the pgood LED relies on (you can trace this back on the final page of the schematic, https://files.digilent.com/resources/programmable-logic/zybo-z7/zybo-z7-d1-sch.pdf, from LD13 being enabled on PG_ALL_EDGE to ultimately if the EN input to the power regulator is above 0.8 V (as per the ADP5052 datasheet).

Doing some math to check the worst case scenario with the R295 and R299 with your measured 4.53 V supply input (to see if it has a chance of dropping below the 0.725 V cuttoff) is:

4.53 V * (4.7k Ohm * 0.95) / (4.7k Ohm * 0.95 + 20k Ohm * 1.05) = 0.79 V

So in the worst case scenario where the 5% variance for the resistors was at the worst end, the ADP5052 would not power on because it did not meet the initial 0.806 V threshold (https://www.analog.com/media/en/technical-documentation/data-sheets/adp5052.pdf, Table 2, Precision Enabling). However, this shouldn't be the culprit as you would see the device not powering up on the other two modes.

Could you measure the voltage across C94 (next to the IC21 flash chip on the underside of the board)? I'm not expecting anything dramatic, but the only really plausible explanation I can think of is that the flash chip is damaged in such a way so that when it is in use it is causing an overcurrent condition which is then tripping the ADP5052 to enter its own hiccup / fault condition.

Thanks,
JColvin

[mhmt9851]

Hi @JColvin,

I measured C94 and its 3.25V. Also for it to work, after switching the power on i need to press the ps-srst button one time. After that, every time i press  ps-srst it programs itself successfully, until i switch the power back off. But it only works maybe one in ten. It used to program itself directly after switching the power on. This does not happen in sdcard mode. It programs itself immediately. Lastly i dont know if its related but when i switch the power on sometimes LD6 flashes red for a short time.

Thanks,

mhmt9851