K10 and External Memory Random Reading Erros

[antoniooliveira]

Hi Guys,

I have a custom board with a K10 and an external RAM memory of 1M bytes (1Mx8) from Cypress. I was able to setup and access the memory range starting in 0x60000000, I can write and read any address without problems, looked like fine. I am using the flexbus with processor expert for setup. I am using 8 bits byte width connected to FB_AD 31 thru 24 and addresses from FB_AD 0 thru 19, control pins FB_CS, FB_OE, FB_WR, Cypress memory dos not require more than this. It is just the simplest stuff.

So the problem starts when I try to access the addresses writing and reading in a sequence in a loop of 256 steps. The loop is sequential starting in 0 thru 255 with base memory of 0x60000000, with external bus clock of 6MHz. It writes and reads but with 3% of error, that means I write one thing and read another, and it is random too.

Did anyone have this issue before? Any help would be appreciated.

Regards,

Antonio Oliveira

[Hui_Ma · Accepted solution]

Hi,

Could you check the signal quality during Flexbus read/write operation? If there exists the EMI affect the FB_AD lines signal.

Please try to set the  PORTx_PCRn register's [DSE] bit to enable high drive strength during the Flexbus pins initialization code.

Wish it helps.

Have a great day,
Ma Hui

-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------

[antoniooliveira]

Hi Ma Hui,

I will check the PCR options for the pins. I have used the Processor Expert to do the setups and did not see any option for high driver.

I will check and let you know.

Regards,

Antonio

[antoniooliveira]

Hi Ma Hui,

It is working now. I set the DSE for the output pins in the MK10 and the signal is better now. I did set to the imputs too but I am not sure if works, according to the manual just outputs are affected.

There was also one input defective in the memory chip, the pin D4 was not responding. Afert I could see a clear signal I could figure out the problem.

The memory can be accessed at 48MHz without any problem, I did check it for a fast sequence of read and write, the CS cycle is 42 ns, and the real cycle is 200 ns, may due to other commands like FOR loop. The mcu clock was at 95MHz.

Regards,

Antonio Oliveira

[Hui_Ma]

Hi Antonio,

Thank you for the feedback.

That's a good news. Have a good day!

best regards,

Ma Hui