Symmetrical signal transmission
Symmetrical signal transmission is a method to be able to transmit signals as tolerant as possible to interference, even over longer transmission paths. The transmission takes place with a pair of similar signal conductors instead of just a single signal conductor. The actual signal is transmitted on one line and a reference signal known to the receiver on the other. The effect of coupling on the transmission path on the useful signal is almost the same on both conductors if they have the same source impedance , the same line impedance and the same load impedance . When the difference between the two conductor potentials is formed, the disturbance is then (almost) canceled.
In addition to the term symmetrical signal transmission (derived from the fact that interference has a symmetrical effect on both signal conductors), one also speaks synonymously of differential signal transmission , derived from the fact that the useful signal is obtained on the receiver side by forming the difference between the two signal conductors.
The term symmetrical signal transmission is often associated with differential transmission and voltage-symmetrical supply with respect to a reference potential ( ground ), but a distinction must be made between the three aspects of impedance properties, voltage curve and reference potential. With symmetrical transmission , engl. balanced line , strictly speaking, only impedance properties are meant. Depending on the application, the reference potential is transmitted as a third conductor in addition to the line pair for symmetrical signal transmission.
principle
If an electrical signal is to be transmitted in a wired manner, a closed circuit is generally required. There are therefore always two conductors between the signal source and the signal sink. This also applies to "single-wire technology" - here the second conductor is represented by earth .
If an electrical signal is transmitted on a line, all inductive and capacitive environmental influences have a disruptive effect on this signal. Depending on the nature of the cable (microphone cable, telephone line, LAN cable, ...) and signal properties, the interference can have a significant impact on the signal quality and can be stronger than the useful signal itself after a few centimeters or kilometers. Attempts in the 19th century to transmit telephone calls asymmetrically (e.g. with only one line and with earth as the return conductor) were therefore unsuccessful.
An alternating electric field causes z. B. by motors, transformers or AC lines, generates a frequency-proportional rectified induction voltage in each wire of a line. Capacitive influences add a ( displacement ) current proportional to the frequency to each wire. According to the Faraday cage principle , the cable can be shielded from the alternating electrical field by a conductive surface (e.g. a metal foil or a braided shield). A (low-frequency) alternating magnetic field also penetrates metallic conductors, it can practically not be shielded and thus induces in any line, however shielded. There is therefore no means of transmitting a signal over a conductor undisturbed. In addition, a distant recipient is u. U. at a completely different ground potential, which, even if the signal arrived there undisturbed, would generate an interference signal.
One solution to this problem is to not only transmit a signal, but also to transmit it as a reference signal
- a signal of opposite polarity that is identical to the useful signal (differential transmission) or
- a zero signal (pseudo-differential transmission).
If the signal is transmitted together with the reference signal via a suitable cable, all inductive and capacitive interference has an identical effect on both wires and the interference signal can be precisely eliminated by forming the difference between the two signals on the receiver side.
So-called symmetrical cables are used as suitable cables . With these, the two wires of a pair of conductors have identical electrical properties. In addition, the cores are usually twisted against each other ( stranding ) and usually also surrounded by an electrically shielding sheath (shield). If there are many pairs of wires in one cable (telephone technology), different twisting of the different pairs proves to be the most suitable means of balancing mutual signal coupling and thus minimizing it. Since the interference is induced voltages and impressed currents, identical coupling and termination impedances of both wires are required for an exactly symmetrical coupling of the interference.
use
Symmetrical signal routing was used worldwide in the analog telephone network. Today it is used almost exclusively in professional sound engineering ; Moving coil microphones with their symmetrical connection are popular in live sound engineering because they are simple and robust - the moving coil is also part of the transducer and transformer . Even in digital technology, symmetrical cable routing is used today. How to use USB and modern Ethernet symmetrical signal transmission.
Asymmetrical signal routing, on the other hand, is still preferred in areas of application where cost reasons and mechanical robustness are in the foreground. Typical examples are
- electrical measurement technology
- Sound transmission in non-professional (HiFi) and partially semi-professional environments ( cinch , DIN , jack plugs (also used for symmetrical signal transmission), S / PDIF , BNC ),
- analog video transmission ( SCART ),
- digital transmission ( EIA-232 (RS-232), IEEE 1284 (Centronics), HDBaseT , 10BASE2 - and 10BASE5 - Ethernet )
- High frequency technology ( antenna , LNB , BK network )
Application and theory
In the following considerations, real existing line impedances, which lead to attenuation and distortion of the phase and frequency response of both wires, are not taken into account. It is assumed that disturbances have the same effect on both cores, which can be assumed in particular when the cores are twisted.
Non-symmetrical transmission
In non-symmetrical transmission, the two wires have different electrical properties (source, line and terminating impedances), for example with a real signal wire and the cable shield. A radiated disturbance arrives at the signal receiver differently on both wires, resulting in a disturbed signal.
Symmetrical transmission, pseudo differential signals
With symmetrical transmission, the two wires have identical electrical properties, for example in the case of two real signal wires of the same type. A radiated interference arrives at the signal receiver equally on both wires; the interference can be eliminated by forming the difference.
Pseudodifferential transmission means that the actual, time-variable useful signal is only transmitted on one wire, the other wire (reference signal wire) carries a zero signal that cannot be changed over time, the reference potential.
In mathematical terms, the pseudo differential symmetrical signal transmission can be described as follows:
The signal to be transmitted may bear the time function at the source .
The reference signal is static zero: .
During the transmission, a temporally variable disturbance s (t) may act identically on both wires.
The interference can be completely eliminated at the recipient's location by forming the difference:
Symmetrical transmission, differential signals
The balanced transmission of differential signals is different from the pseudo-differential signals in that as a reference signal the negated useful signal (of opposite polarity) will be transmitted: .
At the recipient's location, the interference can be completely eliminated by forming the difference:
The differential signal transmission is somewhat more complex on the transmitter side, but leads to a doubled amplitude deviation on the receiver side - as a positive side effect - i.e. an increase in the signal-to-noise ratio by 6 dB .
Realization details
If the source and terminating impedances of both wires are not identical, jack plugs , BNC or Cinch connectors can also be used for transmission. The main advantage of symmetrical transmission, namely the feeding of the receiver from the signal and reference wire and the associated significant improvement in the signal transmission quality compared to conventional routing, is retained.
When transmitting signals within devices, it may be necessary to treat an asymmetrical output signal from one module in the next module as a pseudo differential signal. This is necessary if no common ground point can be implemented within the device , to which all analog signals within the housing are related. Due to the changing electrical currents in the ground lines between the assemblies , voltages are generated based on Ohm's law , which add up to the transmitted asymmetrical signal, because the interference is fed into only one line, the reference potential of the asymmetrical signal. One example are computers with an active loudspeaker box built into the drive bay . The active loudspeakers receive the audio signals through a three- / four-wire line directly from the sound card on the mainboard, but receive the ground for the voltage supply of the audio amplifier built into the loudspeaker through the cable harness, which also supplies the drives with operating voltage. High pulse currents to the processor and drives lead to different potentials in the two ground connections. Amplifiers with input circuitry for symmetrical signals are designed for the potential difference in the ground connections, while amplifiers for asymmetrical input signals can neither eliminate the interference signals if the ground line of the audio line is not connected to the amplifier, nor if they short-circuit the ground potentials. The latter connects the ground lines to form a conductor loop, which induces further interference currents (see conductor loop in the magnetic field ) and through the short circuit leads to excessive potential equalization current surges in the ground wire of the audio line.
Components
A subtraction of time-dependent signals can in principle be achieved with various means. For example, an isolating transformer only transmits the difference between the signal voltages applied to the primary winding. With the means of modern microelectronics, a subtraction amplifier can also be used to form the difference if isolation is not required (building technology, sound technology, computing technology in the house).
In sound engineering, depending on the application, one weighs up between the use of a transformer that allows passive switching, but shows a non-ideal frequency response , and active switching with a differential amplifier.
- Transformers : Transformers or Transformers are used where insulation is required. You can find them as so-called LAN magnetics in Ethernet .
- Subtraction amplifiers : Operational amplifiers are preferred when linear frequency response is required (analog technology, sound technology). Although better results are achieved than with a balanced, pseudo differential transmission with protective insulation, the effort is high. The subtraction amplifier is reserved for special applications that, for example, have protective earth. The principle of the subtraction amplifier is simple: the reference signal is negated by the operational amplifier and then added to the signal. In its simplest form, a single, negating operational amplifier and two resistors are required to add the signals. Highly integrated electrometer subtractors achieve higher precision, greater common mode rejection and offer high input resistances so that they can be adapted to any line impedance .
Difference principle in digital technology
In high-frequency technology (USB, Ethernet, television transmission), there is another problem in addition to the problem of differential transmission. Each cable has a defined wave impedance that is in the order of magnitude between 50 and 300 Ω . If the source or terminating impedance deviates from the wave impedance, signal reflections occur at this point (known e.g. from ghost images on television).
For example, for extremely high data rates or transmission frequencies, impedances in the range of the wave impedance are required. However, given the usual signal voltages, these lead to high power losses in the signal-driving gates.
Techniques have thus become established that manage with the lowest possible signal swing (equal to the lowest power loss). These are necessarily differential transmission methods.
The line is terminated with the wave impedance; that is, the differentially transmitted wires are connected to a resistor whose value corresponds to twice the wave impedance (see theory).
Because of the low-resistance terminating resistor , the transmission gate must be able to deliver a high output current. Such gates are called line drivers or buffers . Schmitt trigger gates are mostly used as receivers in order to increase interference immunity.
However, since the asymmetrical structure is relatively sensitive to external interference (on the ground line ), the symmetrical structure is preferred. The second line is fed with the complementary signal and uses a differential amplifier with a comparator as a receiver. An external disturbance has the same effect on both lines and causes a common-mode modulation , which is filtered by the difference formation in the comparator.
The complementary signals must not have any time shift, which is why in practice components ( ICs ) with complementary outputs are used on the one hand and the lengths of both lines are dimensioned exactly the same on the other.
Complementary outputs are available a priori in circuits using ECL technology , which is why no special modules have to be used. ECL modules are therefore particularly suitable for symmetrical data transmission. At the receiver, comparators with an ECL-compatible output signal are used, which are referred to as line receivers .
In the modern digital technology ( USB , Ethernet , RS485 , LVDS ) ECL components were differentially working CMOS - logic gates with bipolar output stage ( BiCMOS fail).
literature
- Michael Ebner: Manual of PA technology. 1st edition, Elektor-Verlag, Aachen, 2002, ISBN 3-89576-114-1
- Hubert Henle: The recording studio manual. 5th edition, GC Carstensen Verlag, Munich, 2001, ISBN 3-910098-19-3
- Thomas Görne: Microphones in theory and practice. 8th edition, Elektor-Verlag, Aachen, 2007, ISBN 978-3-89576-189-8
- Thomas Görne: Sound engineering. 1st edition, Carl Hanser Verlag, Leipzig, 2006, ISBN 3-446-40198-9
- Siegfried Wirsum: Nf tricks for the audio freak. 1st edition, Franzis Verlag GmbH, Munich, 1990, ISBN 3-7723-3321-4
- Helmut Röder, Heinz Ruckriegel, Heinz Häberle: Electronics 3rd part, communications electronics. 5th edition, Verlag Europa-Lehrmittel, Wuppertal, 1980, ISBN 3-8085-3225-4
- Michael Dickreiter, Volker Dittel, Wolfgang Hoeg, Martin Wöhr, Handbuch der Tonstudiotechnik , 7th completely revised and expanded edition, published by the ARD.ZDF medienakademie, Nuremberg, 2 volumes, publisher: KG Saur, Munich, 2008, ISBN 3- 598-11765-5 or ISBN 978-3-598-11765-7