pic18->m25p54

[AJ]

Hello,

I have been trying to connect a PIC18F452 at 20Mhz to the M25P64(Memory Chip from ST Microelectronics). I have read the previous forum posts, but i had no luck in establishing a connection. I tried both the hardware SPI using built-in functions and also software implemention. I can see with the help of an oscilloscope, the train of pulses going, but the SDO pin of the M25P64 always remains low. I have tried putting delays and using various speeds. Even if i run the program like

while (TRUE){
output_low(flash_select);
printf("%X",spi_read(0x9f));
delay_ms(1000);
output_high(flash_select);
}

still can see only the pulses going out through the SDO pin of the PIC. But nothing is seen on the SDO pin of the M25P64. I am using voltage-dividers for level converting and also tried using the LF version of the PIC.


Then I tried the software one as shown below:

void send_byte_flash(int8 data){
int i;
output_low(FLASH_SEL);
delay_ms(1);
output_low(FLASH_SCK);
for(i=0; i<8; ++i)
{
output_bit(FLASH_DI, shift_left(&data,1,0)); // Send a data bit
output_high(FLASH_SCK);
delay_ms(1); // Pulse the clock
output_low(FLASH_SCK);
}
output_high(FLASH_SEL);
}
//.............................................................................................................
BYTE get_byte_flash()
{int i;
BYTE flashData;
output_low(FLASH_SEL);
delay_ms(1);
output_high(FLASH_SCK);
for(i=0; i<8; ++i) // Get 8 bits of data
{ output_low(FLASH_SCK);
shift_left(&flashData, 1, input(FLASH_DO));
delay_ms(1);
output_high(FLASH_SCK);
}
output_high(FLASH_SEL);
return flashData;
}
//.................................................................................................................
BYTE read_id_flash()
{
BYTE status;
output_low(FLASH_SEL); // Enable select line
send_byte_flash(0x9f); // Send status command
status = get_byte_flash(); // Get the status
output_high(FLASH_SEL); // Disable select line

return status; // Return the status
}



But none of this worked.

Your help will be very much appreciated.Thanks.

[Ttelmah]

Looking at the data sheet, you need to clock three more transactions after sending the 'read identification' command. Basically, when you send this command, the chip then responds _on the next transaction_, with the first identification byte. So your test code would need to be something like:

[AJ]

Hello,

I really appreciate your help. But I'm sorry that the problem is not that simple.
I have tested all modes of built-in SPI functions in all modes.
I have used the modes like SPI_MASTER, SPI_SS_DISABLED, SPI_XMIT_L_TO_H, SPI_L_TO_HIGH, SPI_CLK_DIV_4, SPI_CLK_DIV_64, SPI_CLK_DIV_16, SPI_SAMPLE_AT_END. But no effect. I also tried the pull-up resistors on SDI,SDO,SCK outputs. but still the SDO_pin of the M25P64 is always low no matter what.
I tried using the sotware SPI using differnt delays for read and write operation. (I studied the SPI drivers supplied by the CCS and did something similar.) . Still the same.
I am not ruling out the possibility that the chip is not working. But the chances are very less, since i have changed a couple of units to test. Do you know of a way to test whether the chip is alright or not (other than the read_identification. something electrical) ? It's been quite a few days for me now. I have got other memory chips which use I2C working without any hiccups.

I know there must be some aspect of the M25P64 or the PIC's SPI that I don't take into consideration which may be the key to this. I need some more help.

[AJ]

Hello,

I really appreciate your help. But I'm sorry that the problem is not that simple.
I have tested all modes of built-in SPI functions in all modes.
I have used the modes like SPI_MASTER, SPI_SS_DISABLED, SPI_XMIT_L_TO_H, SPI_L_TO_HIGH, SPI_CLK_DIV_4, SPI_CLK_DIV_64, SPI_CLK_DIV_16, SPI_SAMPLE_AT_END. But no effect. I also tried the pull-up resistors on SDI,SDO,SCK outputs. but still the SDO_pin of the M25P64 is always low no matter what.
I tried using the sotware SPI using differnt delays for read and write operation. (I studied the SPI drivers supplied by the CCS and did something similar.) . Still the same.
I am not ruling out the possibility that the chip is not working. But the chances are very less, since i have changed a couple of units to test. Do you know of a way to test whether the chip is alright or not (other than the read_identification. something electrical) ? It's been quite a few days for me now. I have got other memory chips which use I2C working without any hiccups.

I know there must be some aspect of the M25P64 or the PIC's SPI that I don't take into consideration which may be the key to this. I need some more help.

[Ttelmah]

Have you actually tried what I suggested?.
Your manual 'test' code, won't work, because you deselect the chip between bytes. The 'point' is that the chip has to be selected, then clocked a command, and then _while still selected_, the response must be clocked back.
A variant of the test code (actually clocking an address, and reading the byte back), has worked OK for me in the past.

Best Wishes

[AJ]

Hello friend,

I have tried what you have said. But that did not help. In fact i had tried this earlier also where it failed.

Regarding the chip-select like what you say, I've just tried not making the chip-select high(Deselecting). But it ain't kicking.

Have you interfaced any serial flash memory modules from ST Microelectronics before ? I am really lost.

[Ttlemah]

Yes, I have used the smaller models of the M25P family, in the past, and had no problems.
I did have problems in the past with the CCS SSP functions, ad have in the past posted my own 'variants', which were what I actually used for this. However the newer compiler versions seem to have fixed these oddities.

Best Wishes

[AJ]

Hello,


I am using the PCWH compiler version 3.225. I have tried some of the codes found here in this forum before. Is there any setting in the M25P64 that i should make before writing commands to read the ID or write enable. I just can't figure it out. Please Help.

[Guest]

[AJ]

Anybody there to help me ??