1996 Jun 27
54
Philips Semiconductors
Product specication
8-bit microcontroller with on-chip CAN
P8xCE598
13.5.19.3 Time Segment 2 (TSEG2)
This time segment provides:
additional time at the sample point for calculation of the
subsequent bit levels (e.g. arbitration)
synchronization buffer segment directly after the sample
point.
TSEG2 is programmable from 1 to 8 system clock cycles
(see Section 13.5.10.
13.5.19.4 Synchronisation Jump Width (SJW)
SJW defines the maximum number of clock cycles (tSCL) a
period may be reduced or increased by one
resynchronization. SJW is programmable from 1 to 4
system clock cycles, see Section 13.5.2.
13.5.19.5 Propagation Delay Time (tprop)
The Propagation Delay Time is:
tprop is rounded up to the nearest multiple of tSCL.
13.5.19.6 Bit Timing Restrictions
Restrictions on the configuration of the bit timing are based
on internal processing. The restrictions are:
tTSEG2 ≥ 2tSCL
tTSEG2 ≥ tSJW
tTSEG1 ≥ tSEG2
tTSEG1 ≥ tSJW +tprop.
The three sample mode (SAM = HIGH) has the effect of
introducing a delay of one system clock cycle on the
bus-line. This must be taken into account for the correct
calculation of TSEG1 and TSEG2:
tTSEG1 ≥ tSJW +tprop +2tSCL
tTSEG2 ≥ 3tSCL.
13.5.20 SYNCHRONIZATION
Synchronization is performed by a state machine which
compares the incoming edge with its actual bit timing and
adapts the bit timing by hard synchronization or
resynchronization.
t
prop
2
physical bus delay
input comparator delay
output driver delay
+
(
)
×
.
=
This type of synchronization occurs only at the beginning
of a message.
The CAN-controller synchronizes on the first incoming
recessive-to-dominant edge of a message (being the
leading edge of a message's Start-Of-Frame bit;
see Section 13.6.2.
Resynchronization occurs during the transmission of a
message's bit stream to compensate for:
Variations in individual CAN-controller oscillator
frequencies
Changes introduced by switching from one transmitter
to another (e.g. during arbitration).
As a result of resynchronization either tTSEG1 may be
increased by up to a maximum of tSJW or tTSEG2 may be
decreased by up to a maximum of tSJW:
tTSEG1 ≤ tSCL [(TSEG1 + 1) + (SJW + 1)]
tTSEG2 ≥ tSCL [(TSEG2 + 1) (SJW + 1)].
TSEG1, TSEG2 and SJW are the programmed numerical
values.
The phase error (e) of an edge is given by the position of
the edge relative to SYNCSEG, measured in system clock
cycles (tSCL).
The value of the phase error is defined as:
e = 0, if the edge occurs within SYNCSEG
e > 0, if the edge occurs within TSEG1
e < 0, if the edge occurs within TSEG2.
The effect of resynchronization is:
The same as that of a hard synchronization, if the
magnitude of the phase error (e) is less or equal to the
programmed value of tSJW
To increase a bit period by the amount of tSJW,if the
phase error is positive and the magnitude of the phase
error is larger than tSJW
To decrease a bit period by the amount of tSJW if the
phase error is negative and the magnitude of the phase
error is larger than tSJW.