how to set OCTOSPI1 of STM32L4P5G-DK as 1.8V instead of 3.3V?

[JKim.2]

Hello,

 

Testing 3.3V Octo-SPI PSRAM is fine on the STM32L4P5G-DK board. I'd like to test another 1.8V Octo-SPI PSRAM now. The solder bridge options SB17/SB18 for PSRAM and SB13/SB14 for MCU are modified for the purpose. However, a certain DQ bit is still failing. What else do I have to configure accordingly?

 

Thanks.

[Mike_ST]

Hello,

>> The solder bridge options SB17/SB18 for PSRAM and SB13/SB14 for MCU are modified for the purpose. What else do I have to configure accordingly?

That should be it, unless you're connecting a chip that requires different connections than the ones available on the STM32L4P5G-DK.

>> However, a certain DQ bit is still failing.

You means only one data line is failing ? Could it be a bad solder ?

 

 

 

 

[JKim.2]

Hello @Mike_ST ,

 

The 1.8V Octo-SPI PSRAM has the exact same 24-ball configuration as the original 3.3V Octo-SPI PSRAM. 

only the bit17 in 32bits is failing. other 31bits are fine. The OCTOSPI1 speed was set as 60MHz. (SYSCLK 120MHz, prescaler 2) as AN5050 introduced. Do I still need to configure something in the .ioc file?

 

Thanks.

[Mike_ST]

Hello,

Hard to tell, maybe timing related.

Please have a look at the following application note:

AN5050 

[EZHON.1]

what's do you mean to test 1.8V Octo SPI device?

for my understanding, most of the Octo SPI only support 3.3V or 1.8V range. if you want to test them, I think you will solder a new Octo SPI memory which support 1.8V only?

 

[JKim.2]

Hello @EZHON.1 ,

 

Yes. the device codes of Octo-SPI PSRAM are separated for 1.8V and 3.3V Vdd. I replaced the existing 3.3V PSRAM by 1.8V PSRAM. Do you think I need to configure another thing except changing the two solder bridge options?

 

Thanks.

[EZHON.1]

Personally, it should be OK. you have selected 1.8V for OctoSPI and MCU IOs.

could you check the soldering firstly?

 

[EZHON.1]

Hello, @JKim.2 

OctoSPI is 8 bits data. If you can get the first 2x8 bits without issue, it's no Hardward issue. do you test "0" and "1" in bit 17 is the same issue? maybe related to FW configuration?

 

[JKim.2]

Hello @EZHON.1 ,

 

As I mentioned at the beginning, 3V PSRAM has no issue. Then the FW configuration is fine I guess. Do I need to change any FW configuration for 1.8V?

The expected '2' turned to '0' in the failure log.

              FAILURE: mem_addr = 0x903408c4 antipat = fff2fdcd output = fff0fdcd

              FAILURE: mem_addr = 0x903408c8 antipat = fff2fdcc output = fff0fdcc

Thanks.

[EZHON.1]

It's strange. the bit1 of the 8 bits date is right for bit 1, bit 9 and bit 25, only fails at bit 7. can you write others liking: "fff2fdcd", "55AA55AA" or "AA55AA55"?