TMS570LC4357: EMIF interface with Asynchronous nvSRAM CYB104NA

Is any intended write/read operation followed by this unexpected read?  Is the read address same as the address used by the intended R/W operation?

Hi Wang,

There is no intended write/read operation followed by the unexpected read.

In code this is what I am doing:

Line 1: *(UINT32 *)0x60000000 = 0x12345678; 
Line 2: *(UINT32 *)0x60000004 = 0xABCD5678; 
Line 3: *(UINT32 *)0x60000008 = 0x89AB5678; 
Line 4: *(UINT32 *)0x6000000C = 0x9ABC5678; 

Line 5: Data1 = *(UINT32 *)0x60000000;
Line 6: Data2 = *(UINT32 *)0x60000004;
Line 7: Data3 = *(UINT32 *)0x60000008;
Line 8: Data4 = *(UINT32 *)0x6000000C;

I am keeping breakpoint at Line 1 and after Line 8. On executing Line 1 to Line 8, below is the waveform for this operation:

Pic1: 

Pic2:

Address lines A7 to A15 are same as A6.

Pic1: is as expected. 4 32-bit write followed by 4 32-bit read.

Pic2: is unexpected Address lines and chip select line transitions. Not sure why EMIF module is generating this?

Regards,

Vishwas HC

Hello,

I probed the nCS signal on LC43x HDK when running the code you posted, and didn't see the extra unexpected read operation.

BTW, did you enable the MPU by calling _mpuInit_()?

Hi Wang,

Yes, I am initializing the MPU before performing read/write.

PINMMR setting are also okay. I verified that all EMIF required signals are multiplexed properly.

I am observing unexpected CS transitions when I probe TMS570LC43x HDK's J10 "nCS2" pin.

Memory Region 5 setting is as below:

Regards,

Vishwas HC

Hi Wang,

Unexpected CS transitions are happening after 170+ms from the last read operation, with a breakpoint after the last read operation. Can you please check if you are observing the same?

I am not observing this unexpected transitions during power-on reset or power off.

I am observing this issue with a debugger connected. If I do not keep breakpoint after the last read, unexpected transitions occur when I suspend the program execution.

Can you please confirm if you are also observing the same?

Regards,

Vishwas HC

Hi Vishwas,

I can do one more test. Last time, I also checked the CS signal for more than hundreds ms.

Hi Vishwas,

The maximum window size I can capture using my TDS scope is 200ms. I did not see the extra transactions. In the picture below, the channel 1 is nCS signal, the channel 3 is nWe signal, and channel 4 is nOE signal.