Title: 4'4 Optical Media
14.4 Optical Media
- Optical fibres support the CAN bus. The dominant
bit is characterized by the presence of light
emission and the recessive state is characterized
by its absence. - As regards the propagation time of an optical
fibre, the calculation process for speeds,
distances, propagation time, nominal bit time,
etc. is identical to that described above. - The main advantage of optical fibres is to resist
electromagnetic parasitic signals. - the problems arising with optical fibres are
mainly due to their insertion losses as a
function of distance and their poor capacity to
support the 'bus' topological configuration. - 'point-to-point' connections provides a CAN
optical fibre gateway between two CAN wire
systems
2a 'tractor and trailer' unit
3the implementation of a CAN network on optical
fibres
4the implementation of a CAN network on optical
fibres
54.5.1 Radio-frequency waves
- transport CAN frames by a radio-frequency wave
medium - amplitude modulation with total carrier
suppression (ASK), where the dominant level is
considered to be the physical state of the
presence of the HF carrier and the recessive
level is considered to be its absence - frequency modulation with frequency hopping, in
such a way that, for example, Fl is considered
recessive and detected as such and F2 is called
dominant - The system constructed in this way must be both a
transmitter and a receiver, and the delay times
used to control reception, demodulation and
detection (for the receiver) and modulation and
transmission (for the transmitter) must be
compatible with the bit time
64.6 Pollution and EMC Conformity
- the EMC problems caused by wire links, in the
form of differential pairs (parallel or twisted)
dedicated to CAN applications. - The problems of ensuring EMC mainly arise when
high-speed CAN is used - basic diagram of the standard CAN application in
simplified form, two CAN nodes with their line
drivers, together with line termination resistors
(120O) at each end of an unscreened line, not
forgetting the essential ground return, without
which nothing would work between the two nodes.
7basic diagram of the standard CAN application
84.6.1 The standards
- The standards are complex (IEC 806-6) but, the
requirements are - for transmission the system must not radiate, to
prevent interference with radio or mobile
telephone reception, etc. - for susceptibility the system must resist
disturbance by the proximity of electrical fields
of the order of 100-200 V m-1 in a frequency band
ranging from continuous current to several GHz.
9The theory
- The choice of a differential pair which is
symmetrical and also twisted provides important
structural advantages - - It provides freedom from numerous constraints
concerning the presence of parasitic or transient
signals occurring on adjacent power supply line
as these affect both wires of the bus
simultaneously, they will not degrade the data
signal. - - The active signal (the transmitted data)
develops on either side of the constant mean
level with equal and opposite values at all
times. - the immunity of this transport medium is
essentially determined by what is known as its
'common-mode' performance.
10The reality
- in fact, signals have rise times which are never
exactly equal and complementary to the signal
developed on the other wire. Consequently, there
is a difference signal which may produce
electromagnetic pollution which must be allowed
for and/or cancelled if possible.
11Description of the equivalent diagram
- Each node contributes to the radiation with two
emission sources. The diagram shows - Vemel, which is the voltage source equivalent to
the pollution contributions due to the additional
components of this node. This source is included
in the ground connection of the node in question - Vceml, which represents the voltage equivalent to
the emission which is produced by poor
compensation between the voltages developed on
the CAN_L and CAN_H wires, with respect to the
reference potential (ground) of the data - Vdist, which represents the voltage equivalent to
that due to the effect of the power of the
electromagnetic interference captured by the
CAN_L and CAN_H wires.
124.6.2 Measurements and results of measurements
- In Common mode
- The common mode signal rejection is by far the
most important criterion. - 82C250
- This circuit, normally supplied at 5 V, has a
common mode rejection voltage range from -7 to
12 V, corresponding to the equivalent injection
of a sinusoidal signal of 19 peak to peak, i.e. a
sinusoidal voltage of approximately 7 V effective
on the bus!
13Symmetry of the output signals
- The output stages of the transmission interface
must be as symmetrical and simultaneous as
possible, in order to reduce to a minimum the
voltages present during the phases of switching
from recessive to dominant and vice versa. - In practice, this perfect complementarity is
almost achieved because the output stages consist
of complementary transistors whose performance is
adjusted by means of carefully applied physical
differences (optimal choice of conductivity and
mobility of the type of free minority and
majority carriers of the semiconductors used).
14Rise times of signals
- excessively high speed of the edges of the
signals used can be resolved by optimizing the
slew rate as a function of the loop and the
chosen speeds. - provision is made for the continuous adjustment
(calculation below) of the signal slope (by
increasing the rise time), while maintaining the
speed, the precision of the sampling point, the
length and the type of cable used.
15(No Transcript)
16Example of calculation of the slew rate (dV/dt)
- the value of the resistor Rs to be connected
externally between this pin and the ground. The
relation between these parameters is then - Sr is the slew rate to be obtained.
- As the manufacturer states that a slew rate of 14
µs V-1 is obtained for a resistance Rs of 47kO, - After calculation, the value of this constant k
is 3.8 x 10-3 µs kO-1.
174.6.3 Numerous consequences and conclusions if
problems arise
- In general, all these recommendations enable us
to use parallel pairs with a ground return
forming part of the connecting cable. - For specific problems, in order to reduce even
further the effect of undesired signals in common
mode, it may be necessary to use twisted pairs
plus a separate ground wire or to twist all three
together.
18Making the output signals symmetrical
- connecting a balancing network for the CAN_H and
CAN_L lines at the output of the driver stages. - For this purpose, each of the adaptation
resistors at the line ends (both 120 O) can be
divided into two series resistors of 60 O, and
their midpoints are connected to the AC ground by
means of a capacitor
19Smoothing inductance
- two (very small) surge inductances, coupled
together as required, can be used to eliminate
the flow of current in the same direction on both
outputs of the bus.
20a single diagram the main solutions used
conventionally to resolve most cases
214.6.4 Results
- For immunity
- The graph shows how this network withstands an
injection of parasitic signals on the bus at
levels up to 20 V effective
22For radiation
- The three graphs in Figure 4.54 show the effect
of the different components on the non-radiating
quality of the network - without slope control, without coil
- with slope control
- - without coil
- - with coil (Figure 4.55).
- One last comment on high speeds, the slopes used
to reduce dV/dt, dI/dt, etc. of the signal and to
reduce the EMC levels.
23(No Transcript)
24(No Transcript)
254.6.5 Conclusion
- what is expected of a normally constructed CAN
line driver, in terms of managing line short
circuits, bus line faults, partial networks, low
and very low power consumption, local and remote
wake-up, recognition of wake-up modes, management
of diagnostics, good EMC performance (in terms of
immunity and susceptibility) and ESD resistance,
compatibility with all the 3.3 V, 5 V and other
microcontrollers in the market and resistance to
heat stress.