2018-08-09 01:15 AM
we used a SDRAM (ISSI 42-45S16160G 256MBit 16M*16b) on Bank 1 and a 512MB NAND Flash (IS34_35ML04G081) connected to Bank 2 and 3 on an STM32F429BIT6. Sporadically we get an 0 bit on a flash page, but this is no problem of the NAND (ECC did not detect any error).
I want to write data into an erased page (2048 byte). Mostly this is ok but somtimes, I detect a 0 bit on a position where I expected a 1. I used the standard HAL-NAND Routine to write.
It seems that the Hal NAND driver is disturbed by interrupts (specially the 2D DMA) who is active working in the background.
A test without the running 2D-DMA runs without any problem.
2018-08-09 09:34 PM
For commercial projects contact your local sales/support office and work with your FAE
Going to be super hard for anyone here to replicate your situation absent hardware, and demo of stress conditions, hoping for me-too stories.