Strange behavior of W25Q64JVSSIQ

[adhoc_user]

Normally sleep current of the system is 12uA. On some boards it reaches 20uA. After applying pressure to the  flash chip and resolder it, the consumption goes down to 12 uA. What would be the problem?

[SiliconWizard]

Normally sleep current of the system is 12uA. On some boards it reaches 20uA. After applying pressure to the  flash chip and resolder it, the consumption goes down to 12 uA. What would be the problem?

Bad solder joint?

Jokes aside, the difference in current draw seems to be roughly the difference between the standby current and power-down current of this flash chip. Which would suggest that on the problematic boards, the flash chips don't stay in power-down mode (which I assume is what you set the flash chip to when your system is in "sleep mode"?) If so, one possible cause given your symptoms could be the /CS pin, if otherwise other flash commands seem to work as expected.

[Salitronic]

Do you have a pull-up resistor on the CS pin? (the schematic snippet you shared does not include it)

If you refer to the power-up/down section of the datasheet you'll see that the CS must always follow VCC on power-up/down. Failure to do so can result in an unknown power-up state, and even though you can typical reset the flash memory though software reset, I've observed issues in the past when I forgot that pull-up myself.

Further assuming that you are putting the entire circuit in sleep mode, you need to make sure that the GPIO pin of the MCU is not floating that CS pin, it needs to be held high during the sleep mode.
If you don't have the pull-up, I'd suggest adding it and retesting the power consumption.

[adhoc_user]

recently found out when i put oscilloscope probe on SPI_MISO line consumption drops 12uA.

https://lowpowerlab.com/forum/moteino-m0/spi-miso-needs-pull-resistor/

After reading this entry, I added pullup resistor on MISO line. It seems worked.