We use STM32F767NIH and see random QSPI interface failure accessing a flash.

[Flavio1]

We use STM32F767NIH and see random QSPI interface failure accessing a micron flash. The failure happens when we run diagnostic sw in a loop that does a write, read back and verify function. Using a protocol analyzer, we notice that when the QSPI fails (during read), CS pin going high a few times. Is this normal?

[Tesla DeLorean]

Looks to be sending commands, which your analyzer isn't decoding for you. QPI Mode

You'd need to look at what's going on in your BSP.

Which Micron part?

[Flavio1]

We use MT25QL128ABA1EW7-0SIT. I captured a better view of the read cycle. After the dummy cycle, we should start reading, but not sure why CS goes high in the middle of the read. I think the protocol analyzer sees the toggle on the CS and interprets it as a command?

[waclawek.jan]

It appears that the memory does not "see" those pulses on CS, as it keeps outputting what appears to be reasonable data on all four data pins.

This, together with similar suspiciously short pulses on data, might indicate problematic LA attachment.

JW

PS. Please change your username to a normal nick.

[Tesla DeLorean]

Definitely some weirdness, D3 being LOW is odd. Would check that the memory/CPU have a common understanding about being in QPI mode.

I tend to check in 1-bit, 2-bit and then 4-bit mode. Write patterns that will exercise all 4-bits. Perhaps pick some where the high/low nibble have different states.

If it is doing a long burst 0x6B read, I wouldn't expect the CS to glitch unless there was some cross-talk, or cross-wiring.

You see this on ST boards, or just the custom board?

[Flavio1]

It's custom board.

[Tesla DeLorean]

Do you see similar issues on ST boards using same part? I think the F746G-DISCO​ has one. Stuck at zero for D3 very odd.

[waclawek.jan]

> D3 being LOW is odd.

Indeed, and I did not notice that. Also, on the second screenshot, the pulses on CS appear roughly where D3 ought to pulse, if the data pattern is simply a sequence of 0-1-2-....

D3 shorted to a neighbouring pin/track?

JW

[Flavio1]

Here is another capture with Saleae analyzer. QSPI test failed after 9 loops. Not sure if DQ3 being high by default is correct? both Q2 and Q3 have pull ups to 3.3V.

[Tesla DeLorean]

Certainly not stuck-at-zero, not seen any explanation about if you found a short, or what was happening with those last traces?