252018/12/ Manchester coding - differential - RZ - NRZ

CODING
MANCHESTER Summary
of the Site

Before: Phase Locked Loop Set: Synchronous codes

Last updated: October 27, 2018, 1:27 PM

Why so many different types of coding?

The problems that can arise from transmitting a digital signal over a transmission medium
are of several types:

Important average value

Statistical insufficiency of transitions.

Inadequate bandwidth.

Important average value

See here: the definition of the average value of a signal.

The signal from above has a zero average value.

The one at the bottom, a positive average value.

Electronics engineers know that a signal with a significant average value

creates amplification difficulties and propagates poorly on the lines beyond a certain length.

Statistical insufficiency of transitions.

We saw on the previous page the necessity of a sufficiently large occurrence of the transitions of the coded signal
in order to synchronize the remote decoder's clock with that of the transmitter's encoder.

Inadequate bandwidth.
Most metal wire lines have considerable bandwidth even over long distances.
Provided that these lines are continuous from the transmitter to the receiver.
Emission referred to as 'Base Band'. (see here: )
This is the case for connections between various devices that produce digital streams.
between them on a local network
or between them and the modulator/demodulator equipment for remote transmissions.

This is not the case for remote transmissions going through exchanges.
who multiplexes them :( ) or : ( ) into carrier signals of a large number of simultaneous exchanges
from multiple origins to multiple destinations.

[Link] 1/6
252018/12/ Manchester coding - differential - RZ - NRZ
The bandwidth of the signal to be transmitted must be shifted to normalized values.
It is the role of MODULATION. (Generalities see here: )

That is why I warn the reader not to confuse the coding methods we are examining here.
with the modulation methods that are explained elsewhere on this site.

Amplitude modulation (principles)

QAM Modulations

DSL Modulations

A solution for synchronization: a SMART CODING

The problem raised on the previous page,
Ensure that the signal has a sufficient number of transitions
so that the PLL can maintain the isochronism of the transmitter-receiver clocks.
go find an acceptable solution in the way we code the transmitted bits.

The basic idea is to code the information in such a way that the transmitter's clock
can be deduced from the signal received itself.
Self-supporting clock coding in summary.

This is the case of the MANCHESTER code.

Manchester encoding rules

The transmitting clock has a fixed period T (called 'Bit Time').

The amplitude of the Manchester encoded signal only takes two values [ +U and -U ].

The coded signal must necessarily show an upward or downward transition.

umilitude jacket weather of bit.

Upward unetransition: (L H) corresponds to the value "0" of the binary signal to

transmit.
Downward transition: (H L) corresponds to the value "1" of the binary signal at
transmit.

Under these conditions, the transmitted signal has a transition at each bit time.
(The binary signal to be transmitted is always either '1' or '0').
The clock will be very easily extractable to synchronize the PLL.

It will be observed that a sequence of '0's or '1's does not produce a constant signal,
but transitions exist at every bit time.
What allows the synchronization of the receiver (by PLL)

Manchester coding has a drawback....

I'll let you think about it for a moment...

[Link] 2/6
252018/12/ Manchester Coding - Differential - RZ - NRZ

Response here:

Differential Manchester Encoding

The "Differential Manchester" encoding avoids the pitfall of polarity.

Differential Manchester coding rules

We position ourselves at the very beginning of each bit time:

If the signal to be transmitted in the following bit time is '0'

the driving consists of a transition
(ascending or descending) at the beginning of the bit time.

Similarly, if the signal to be transmitted in the following bit time is '1'

the online signal must show no transitions at the beginning of the time of
bit.

The signal must change level in the middle of each bit time.
This prevents long periods without transition in sequences of several bits.
identical (0000... or .1111...) The goal is to be able to regenerate the clock at
the arrival.
code bearer of clock

You notice that with these codes, it is unnecessary to identify the ends of the wires upon arrival.
since the meaning of transitions is indifferent.
This prevents wiring errors.

It's well observed! Isn't it?

Questions
Question

From a reader (A.F) on 16/10/2017

For differential Manchester coding, during the first half-bit, the green line is in the high state.
but could just as well have been in a low state, right?
And in that case, the entire timeline would have been reversed, of course.
Did I make a mistake? ;)

Answer
First, this is a case that can occur: see the bit time of the second "1" (the central "1")
The online signal: first low stays low (no transition for the '1')
until the middle of the bit time when it goes to high level.

Now... will the entire timeline following be reversed?

Apparently yes, because this coding consists of inversions or non-inversions.
[Link] 3/6
252018/12/ Codages Manchester - différentiel - RZ - NRZ
of the signal present just before the moment of inversion.
The inversion of a signal will reverse all its successors.

But be careful! You cannot change the past!

The 'green line' could have been at the top. But it is not.
To be at the top, it would have required that the binary signal to be transmitted had been different!

Interest in MANCHESTER or DIFFERENTIAL MANCHESTER encodings

Remember what was said above: the desynchronization of the receiver with respect to the
The transmitter is due to long stationary passages of the signal.

With the two Manchester encodings in each bit time, there is at least one transition.
of the signal.
This code can even do without a phase-locked loop (PLL)!
And it is for this reason that he was employed for a long time,
even before the PLLs were introduced into digital signal receivers.

But a phase locking loop doesn't hurt, of course.

Do you think we have arrived at a satisfactory result?

Justify your answer and check it by clicking here:

Readers' Mail:

Conclusion
The previous response leads us to conclude that this coding method,
if it is profitable for short connections
(local industrial networks or connections) where the price of the lines does not intervene drastically,
it is revealed to be completely out of competition in the case of long-distance telecommunications lines
where bandwidth is so expensive.

He will find something else for our major fast telecommunications routes.

This will be the subject of the next chapter:

synchronous codes.
But stay a little longer here!

Coding violations

It is possible to violate the coding rules without affecting the transmitted signal.
What interest, you might ask?

Very interesting on the contrary!

If the receiver's decoding device is capable of detecting such a violation,
we can use each of these violations as a signal, for example,
to indicate the beginning or the end of a communication.

This signal is very precise. It can be detected within a small fraction of a bit.
It is used in networks to detect the beginnings of frames.

[Link] 4/6
252018/12/ Manchester encoding - differential - RZ - NRZ

The figure above shows two coding violations marked V.

Note that these are violations of the rule that states the signal must change.
in the middle of each bit time.

In the example of the figure, the violation detector will allow to signify
that the previous thread ends with: 0 1 0 1 1 0 0 1 (before the 2 violations)
and that the following begins with: 10.

RZ and NRZ codings

RZ stands for 'Return to Zero'
NRZ means 'No Return to Zero'

In NRZ encoding, there are only two line states: 0 or 1.

(0V or 5V) or (+12V or -12V)

In RZ coding, there is also a 'latency state'.

during which the line does not transmit any bit.
In the RZ diagram below, this state is a zero voltage in line.
I noted these latency times with brackets and little zeros at the bottom in red.
In the acronym RZ: Zero should be understood as this intermediate latent state 0.
Not as a significant binary state.

An NRZ signal consists of a voltage in line whose value

varies between a low level and a high level (0, 5 V, or -12, +12 V).

The problem that electronics engineers will understand very well

Does this signal have an average value?

[Link] 5/6
252018/12/ Manchester coding - differential - RZ - NRZ
This average value propagates poorly across the lines
The electronics of the receivers have difficulty processing this type of signal.
The signal thus carries useless electrical energy.

We prefer the coded RZ signal which is of the alternating type and thus does not have,
the reported disadvantages for RZ.

But on the other hand,

the frequency spectrum The RZ signal is higher than for NRZ.
The bandwidth of this signal is broadened.
In some applications, bandwidth is expensive...

Finally, let's recall the NRZI (No Return to Zero Inverted) coding studied here:

And now, let's hit the road for advanced coding!

Previously: Phase Lock Loop Cohort: Synchronous Codes

Summary

of the Site

[Link] 6/6