ADIS16305

Rev. 0 | Page 16 of 20

DIAGNOSTICS

Self-Test

The self-test function allows the user to verify the mechanical

integrity of each MEMS sensor. It applies an electrostatic force to

each sensor element, which results in mechanical displacement

that simulates a response to actual motion. Table 1 lists the

expected response for each sensor, which provides pass/fail

criteria. Set MSC_CTRL[10] = 1 (DIN = 0xB504) to run the

internal self-test routine, which exercises all inertial sensors,

measures each response, makes pass/fail decisions, and reports

them to error flags in the DIAG_STAT register. This process takes

35 ms to complete and report the results to DIAG_STAT[5],

DIAG_STAT[10], and DIAG_STAT[15:13]. MSC_CTRL[10]

resets itself to 0 after completing the routine. The MSC_CTRL[9:8]

bits provide manual control over the self-test function for

investigation of potential failures. Table 27 outlines an example

test flow for using this option to verify the gyroscope function.

Table 27. Manual Self-Test Example Sequence

DIN

Description

0xB601

SMPL_PRD[7:0] = 0x01, sample rate = 819.2 SPS

0xB904

SENS_AVG[15:8] = 0x04, gyro range = ±300°/sec

0xB802

SENS_AVG[7:0] = 0x02, four-tap averaging filter

Delay = 50 ms

0x0400

Read GYRO_OUT

0xB502

MSC_CTRL[9] = 1, gyroscope negative self-test

Delay = 50 ms

0x0400

Read GYRO_OUT

Determine whether the bias in the gyroscope

output changed according to the self-test

response specified in Table 1

0xB501

MSC_CTRL[9:8] = 01, gyroscope/accelerometer

positive self-test

Delay = 50 ms

0x0400

Read GYRO_OUT

Determine whether the bias in the gyroscope

output changed according to the self-test

response specified in Table 1

0xB500

MSC_CTRL[15:8] = 0x00

While the self-test still functions when the device is in motion,

zero motion typically produces the most reliable results. The

settings in Table 27 are flexible and allow for optimization

around speed and noise influence. For example, using fewer

filtering taps decreases delay times but increases the possibility

of noise influence.

Flash Memory Management

The FLASH_CNT register (see Table 28) provides a tool for

managing the flash memory’s endurance. The FLASH_CNT

register increments every time there is a write to the flash

memory. Figure 18 quantifies the relationship between data

retention and junction temperature.

Table 28. FLASH_CNT Bit Descriptions

Bits

Description

[15:0]

Binary counter for writing to flash memory

600

450

300

150

0

30

40

JUNCTION TEMPERATURE (°C)

55

70

85

100

125

135

150

Figure 18. Flash/EE Memory Data Retention

Checksum Test

Set MSC_CTRL[11] = 1 (DIN = 0x9D08) to verify the flash

memory integrity against the factory check sum and read

DIAG_STAT[6] to check the results 20 ms after the command.

Status

The error flags provide indicator functions for common system

level issues. All of the flags are cleared (set to 0) after each

DIAG_STAT register read cycle. If an error condition remains,

the error flag returns to 1 during the next sample cycle. The

DIAG_STAT[1:0] bits do not require a read of this register to

return to 0. If the power supply voltage goes back into range,

these two flags are cleared automatically.

Table 29. DIAG_STAT Bit Descriptions

Bit(s)

Description

[15]

Z-axis accelerometer self-test failure (1 = fail, 0 = pass)

[14]

Y-axis accelerometer self-test failure (1 = fail, 0 = pass)

[13]

X-axis accelerometer self-test failure (1 = fail, 0 = pass)

[12:11]

Not used

[10]

Gyroscope self-test failure (1 = fail, 0 = pass)

[9]

Alarm 2 status (1 = active, 0 = inactive)

[8]

Alarm 1 status (1 = active, 0 = inactive)

[7]

Not used

[6]

Flash test, checksum flag (1 = fail, 0 = pass)

[5]

Self-test diagnostic error flag (1 = fail, 0 = pass)

[4]

Sensor overrange (1 = fail, 0 = pass)

[3]

SPI communication failure (1 = fail, 0 = pass)1

[2]

Flash update failure (1 = fail, 0 = pass)

[1]

Power supply > 5.25 V (1 = power supply > 5.25 V,

0 = power supply ≤ 5.25 V)

[0]

Power supply < 4.75 V (1 = power supply < 4.75 V,

0 = power supply ≥ 4.75 V)

not equal to an integer multiple of 16.