Information Memory Sector Erased after power recycled

Hello Deepak,

there are many different possibilities where the trouble can result from.

Have you already checked the Understanding MSP430 Flash Data Retention  App Note?

How do you wirte in to the Flash? Can you provide minimized sample code so for test purposes to see if this behaviour is reproducible?

What is the Vcc supply voltage? Minimum processor frequency is defined by system clock. Flash program or erase operations on the G2302 device require a minimum VCC of 2.2 V. Please check the datasheet at page 17.

Deepak Pathania

... to store calibration data i am using information memory

Why storing Calibration data? As I can see it in the datasheet, Segment A already contains factory preprogrammed calibration data! Please check the datasheet. After reset, segment A is protected against programming and erasing. It can be unlocked, but care should be taken not to erase this segment if the device-specific calibration data is required.

Deepak Pathania

added startup delay after selecting the clock switch (16Mhz)

Deepak Pathania

Added start up delay at the beaning of the code -> solves the problem

Deepak Pathania

I want to know how reliable is this solution?

This should be always considered , specially with slow rising Vcc supply at power-up.

Deepak Pathania

i wanted to enable brown out protection but could not find registers on this device to do that! but datasheet front page tell it has brown out!

moreover is this memory dependable? ( application is not life saving or life threatening) 

The G2302 includes a Brownout reset that is by default activated and will not let the MSP430 operate unless the voltage is acceptable. However, powering the MSP430 device with an unstable power supply could be also problematic. would also highly suggest that you ensure that the power rails are clean and that any noise is kept to a minimum.

BR,

Mo.

Thanks Mo, for your early reply

Erase and Write Code

void write_EEPROM(unsigned char value, unsigned char Address)

{
char *Flash_ptr; // Flash pointer
Flash_ptr = (char *)0x1000 + Address; // Initialize Flash pointer
FCTL3 = FWKEY; // Clear Lock bit

FCTL1 = FWKEY + WRT; // Set WRT bit for write operation
*Flash_ptr = value; // Write value to flash
FCTL1 = FWKEY; // Clear WRT bit
FCTL3 = FWKEY + LOCK; // Set LOCK bit
}

void erase_EEPROM(unsigned char Address)
{
char *Flash_ptr; // Flash pointer
Flash_ptr = (char *)0x1000 + Address; // Initialize Flash pointer
FCTL3 = FWKEY; // Clear Lock bit
FCTL1 = FWKEY + ERASE;// + EEI; // Set Erase bit, allow interrupts
*Flash_ptr = 0; // Dummy write to erase Flash seg
FCTL1 = FWKEY; // Clear WRT bit
FCTL3 = FWKEY + LOCK; // Set LOCK bit

}

Calibration data of micro is not touched but we want to write calibration data for our sensor so we arite to these location 

page containinig micro calibration data starts from 0x10C0 not 0x1000 so i am writing my own calibration values used in sensor at 0x1000

data retention AN is good will read it again for better understanding but we are working at normal temperature 

How do you define slow rising VCC? we have 10uF cap on supply for making supply rails smooth! is that causing trouble?

do u suggest some other solution? say if i reduce 10uF cap to 1uF but this will introduce more noise(anticipated) and will not filter glitches in supply at startup and power down phases

regards

Deepak 

Deepak Pathania

Clock speed at this moment is 16 MHz, reset cap is 1.5nF

Changed the clock speed to 1MHz -> Same failuer

The fFTG timing might be the problem here.

Can you please implement the below code just before writing/erasing to the Flash?

  DCOCTL = 0x0;
  BCSCTL1 = CALBC1_1MHZ;                    // Set DCO to 1MHz
  DCOCTL = CALDCO_1MHZ;
  FCTL2 = FWKEY + FSSEL0 + FN1;             // MCLK/3 for Flash Timing Generator

BR,

Mo.

Mo.

Why storing Calibration data? As I can see it in the datasheet, Segment A already contains factory preprogrammed calibration data!

There may be several reasons for own calibration data.
The factory-stored data is only for certain temperatures and still has some tolerance. Adn it doesn't take external circuitry into account.
If you have e.g. an external voltage divider, then tolerance of the individual resistors of the divider won't be covered by the factory calibration data. Same for offset voltages of an OpAmp that works as input impedance changer.
Doing an own calibration run will eliminate all of this and maybe provide the 'calibration data' in a way it only has to be multiplied with the ADC reading. (simple formula to be executed at runtime).

The factory-stored values are a good starting point to get the ADC functions running. But for the final applicaiton, own calibration might be necessary if you don't want to risk a rather large tolerance.