NCV7520
www.onsemi.com
22
been detected and the fault data is latched into the channel’s
status register.
Thus, a state change of the inputs (POR, RSTB, ENB, IN
X
or G
X
) or a state change of an individual channel’s feedback
(DRN
X
) comparison must occur for a timer to be triggered
and a detected fault to be captured.
Fault Capture, SPI Communication, and SPI Frame
Error Detection
The NCV7520 latches a fault when it is detected, and
parity and frame error detection will not allow any register
to accept data if an invalid frame occurred.
The fault capture, parity, and frame error detection
strategies combine to ensure that intermittent faults can be
captured and identified, and that the device cannot be
inadvertently re−programmed by a communication error.
When a fault has been detected, status information is
latched into a channel’s status register if of higher priority
than current status, and the FLTB flag is set. The register
holds the status data and ignores subsequent lower priority
status data for that channel.
Current status information is transferred from the selected
status register into the SPI shift register at the start of the SPI
frame following the read status request. This ensures that
status updates continue during inter−frame latency between
the status request and delivery. The FLTB flag is reset when
CSB goes low.
The selected status register is cleared when CSB goes high
at the end of the SPI frame only if a valid frame has occurred;
otherwise the register retains status information until a valid
read frame occurs. The FLTB flag will be set if a fault is still
present.
Status registers and the FLTB flag can also be cleared by
toggling RSTB H→L→H. A full I/O truth table is given in
Table 18.
Status Priority Encoding
Shorted load (SCB) faults can be detected when a driver
is ON. Open load (OLF) or short to GND (SCG) faults can
be detected when a driver is OFF. Status memory is priority
encoded in a 3−bit per driver format (Table 12).
Status memory will be encoded “Diagnostic Not
Complete” during a blanking period unless a fault of higher
priority has previously been encoded. Status memory will be
encoded “Diagnostic Not Complete” (cleared) for the
selected status register at the end of a valid SPI frame.
“Diagnostic Complete − No Fault” will be encoded when
BOTH on−state AND off−state diagnostics have been
completed unless a fault of higher priority has previously
been encoded. A diagnostic cycle may start from either an
off−state or an on−state.
When a diagnostic cycle starts from an off−state and no
fault is detected, “No SCG/OLF Fault” will be encoded
unless a fault of higher priority has previously been encoded.
Otherwise, “OLF” or “SCG” will be encoded unless a fault
of higher priority has previously been encoded. If the cycle
continues to an on−state and no fault is detected, “Diagnostic
Complete − No Fault” will be encoded. Otherwise, “SCB”
will be encoded.
When a diagnostic cycle starts from an on−state and no
fault is detected, “No SCB Fault” will be encoded unless a
fault of higher priority has previously been encoded.
Otherwise, “SCB” will be encoded. If the cycle continues to
an off−state and no fault is detected, “Diagnostic Complete
− No Fault” will be encoded unless a fault of higher priority
has previously been encoded. Otherwise, “OLF” or “SCG”
will be encoded unless a fault of higher priority has
previously been encoded.
Status is latched for the currently higher priority fault and
is not demoted if a fault of lower priority occurs. Default
response after reset is D[8:0] = 1 (“Diagnostic Not
Complete”).
When a channel is configured for auto−retry mode, its
status register bits are not affected by ENB and are reset to
“Diagnostic Not Complete” after reading the register (see
Figure 18).
When a channel is configured for latch−off mode and no
“SCB” status is present, its register bits are reset to
“Diagnostic Not Complete” after reading the register. If
“SCB” status is present, its register bits are set to “GLO”
after reading the register. This status is maintained until the
channel is un−latched either by successfully executing the
un−latch sequence or by disabling then re−enabling the
device via the ENB input (Figure 18). The “GLO” status
allows the application to detect a latched−off channel in the
event the “SCB” status data is discarded by the controller
due to SPI transmission error.
A Statechart diagram of the diagnostic status encoding is
given in Figure 18 and additional clarification is given in
“Appendix A − Diagnostic and Protection Behavior
Tutorial”.
VLOAD Undervoltage Detection
Undervoltage detection is used to suppress off−state
diagnostics when VLOAD falls below the specified
VLDUV operating voltage. This ensures that potentially
incorrect diagnostic status is not captured. On−state
diagnostics continue to operate normally and status
information is updated appropriately for an off−to−on input
transition during undervoltage.
Previous status information and FLTB are unchanged
when entering or leaving undervoltage. Upon a read of the
status registers during undervoltage, the status is changed to
“Diagnostic Not Complete” and will remain as such for
channels in an off−state during the entire undervoltage
interval. Status information and FLTB are updated
appropriately if a channel changes from off to on during the
interval.
When VLOAD returns to its normal operating range, a
channel’s t
BL(OFF)
blanking timer is started if the channel
was in an off state. Status information is updated
appropriately after the t
BL(OFF)
blanking interval, or after
the t
FF(OFF)
filter interval if the filter has been activated.
