DE3423818A1 - Digital generator for audible signals which are to be fed into a PCM time division multiplex telecommunications system, in particular a telephone system - Google Patents

Description

T * iefonbau and Normalzeit GmbH, - SOOQ 'Frankfurt "aiä Hain, Mainzer Landstrasse 128 -

Digital generator for in a PCM time division multiplex telecommunication, In particular, acoustic signals to be fed into the telephone system

The invention "relates to a digital generator for in a PCM time division multiplex telecommunications, in particular telephone system Acoustic signals to be fed in, as described in more detail in the preamble of claim 1.

From DE-AS 19 45 155 it is known in one arrangement for feeding audio signals into time division multiplex PCM switching systems to store the amplitude curve of audible tone frequencies in the form of digitized samples. For each individual frequency, a memory area must be provided in which so many digitized Amplitude samples, expediently directly as PCK words are formed, are stored so that a or several complete sine waves can be displayed. By repeatedly reading out such a memory area can from the resulting PCM word sequences after a digital-to-analog conversion, the corresponding one Frequency of an audio signal can be obtained.

A method based on this principle and an arrangement for the distribution of audio signals and call signals in a time division multiplex communication system is specified in DE-PS 23 14 272. It is proposed that the audio signal frequencies and the audible character rhythms separately

(X) are stored, and each time a "special TT link is achieved that the desired ^{audible signal C 1} O is created. Since this method is used at subscriber stations, only one audible tone frequency (fundamental tone) above ^(v> 5 the incoming line is received and only then fed into the receiver of the subscriber's phone in a specific time slot linked to a specific rhythm via a digital-to-analog converter.

A digital audible tone generator, with which frequencies and rhythms are freely linked to different audible tones can be linked, is given in DE-OS 28 55 151-Here mainly programmable memories are used, in which not only the various digitized Audible tone amplitudes are stored as PCM words, but also independent information about the various Audible tones are stored. The rhythms occurring with audible tones are used as pause information in the same stored programmable read-only memory, in which also the PCM word clusters for the representation of a certain Frequency are stored. To get an assignment of the various Additional memories and adding networks are used to enable audible tones for certain time channels intended. In order to achieve a freely selectable combination of frequencies for audio signals are in the DE-OS 28 55 151 specified digital audible tone generator at least 6 different programmable memories required.

Additional memories, adder networks and comparators represent a considerable expense, and the number of available ones Frequencies and rhythms cannot easily be increased.

It is the object of the present invention to provide a digital generator for in a PCM time division multiplex telecommunications, in particular Telephone system to be fed acoustic signals to propose, in which the assignment of each The audio type can be freely selected for a specific time channel,

-D-

and also speech announcements of any length r— can be realized with the same principle that is used for the generation of audible sound. This task CsI 5 is achieved by a combination of features as specified in PQ patent claim 1.

In the case of the digital generator according to the invention, it is achieved in an advantageous manner that only a few are different Read-only memories to be programmed must be available for a freely selectable combination of rhythms and frequencies at the different audible tones and that also different spoken announcements can be generated. It is possible to assign audio tones and their assignment independently of one another to freely change individual time channels as well as voice announcements. Further developments of the invention result from the subclaims.

An exemplary embodiment of the invention is explained in more detail below with reference to a drawing. One with that System clock ST controlled clock counter TZ, which synchronizes in a known manner by a sync pulse SYIT supplies the control addresses for the timing memory ZL-SP with its higher-value outputs. With the A control store ST-SP is used for lower-value outputs controlled, which has a memory area for each individual time slot with a plurality of individually addressable memory lines, in which individual information items are stored, the Form clock-accurate control signals within each time slot when cyclically reading out the memory. Such Control signals could also be formed by an extensive logic logic, but then there is Disadvantage that difficult adjustments are required when changes are made.

In the time slot memory ZL-SP, a memory line is provided for each individual time slot, in which information about the type of acoustic signal to be sent can be stored if such a signal is sent out at this time wea? den should. If the timing memory ZL-SP

CQ "for example, has a capacity of 64 χ 8 bits, so ^v ~ can be 64 time slots on the" Assign optionally up to 256 different acoustic co signals. When reading out a ^ bit combination from the time slot memory ZL-SP, the CO 5 signal type is controlled according to predetermined memory areas in the tone memory TON-SP or speech memory SPR-SP, where the PCM word groups are stored that belong to the desired tone frequency or the desired voice announcement.

Fine addressing for reading out individual PCM words from the tone memory TOIT-SP or the speech memory SPR-SP is connected by an address counter ADZ made with an address buffer ADP, with a control only takes place if there is a time slot in which an acoustic signal is to be sent. this happens in that control signals for driving the address buffer ADP and the address counter ADZ only then the control memory ST-SP are output when there is a situation in which audible tones or voice announcements are sent should be. The address buffer ADP and the address counter ADZ work together so that the clock counter TZ specified addressing of a time slot is then switched through to the address buffer ADP if a time slot is present, in which audible tones or voice announcements are to be sent. In doing so, the one in the address buffer ADP is sent Information, i.e. the address with which the tone memory TON-SP or the speech memory SPR-Sp was controlled, to the output of the address buffer ADP and is loaded into the address counter ADZ. Immediately afterwards the counter ADZ is incremented by one by a control clock coming from the control memory ST-SP, so that an address appears at its output with which the next PCM word can be read out. This cycle will continued within each time slot until together with the last in a manner not shown PCM word of a word group an additional information appears, which the address counter ADZ back to the first Resets the address of the memory area concerned.

Kir the generation of pauses between the "listening sessions" or A clock memory TA-SP is pre-OO between the announcements see, which is continuously addressed by a sampling counter AZ

^ _n J will. The sampling counter AZ receives its control clock via * · * 5 a clock divider TT at which, depending on the OO

minimum or maximum pause length can be set via a routing R1, in which sequence, for example 5 or 10 or 20 ms the sampling counter AZ is activated. The input of the clock divider TT can, for example, with the most significant bit of the clock counter TZ, so that a very precise time clock for the sampling counter AZ is available. The outputs of the sampling counter AZ are directly connected to the addressing inputs of the clock memory TA-SP connected, so that at the output of this memory at each sampling time the previously programmed bit pattern a memory line appear the clock memory TA-SP is organized in such a way that a bit column for each type of rhythm is provided, for example the bits to be read out during a pause with KuIl and the bits in between Bits can be described with one. The..this way out

ι extracted airs ,

the clock memory TA-SP «reach a clock multiplexer TA-MUX whose control inputs are connected to the outputs of the sound memory TOK-SP. According to the information that is read out together with a PCM word, and reaches the clock multiplexer TA-MUX, one of several bits on the output of the clock multiplexer TA-MUX switched and arrives at a combination logic VL2. The linkage logic VL2 is at certain times within a time channel with one cycle from the control memory ST-SP controlled so that an ITuIl generator ITG is controlled. When the Kullgenerator KG is controlled is, instead of a PCM word from the tone memory TON-SP or the speech memory SPR-SP a special, Word representing the value zero to the output buffer AP. The digital generator thus gives in each case Time slot as long as Kull words, as from the clock memory TA-SP pause bits of the respective audio key can be read out.

With the help of the read out from the sound memory TOIT-SP

OQ so-called * ~ selection bits which do not belong to the actual PCM word and which are ^{routed to the clock multiplexer TA-I 1 TUX}

co, so it is possible simply by corresponding

^ Programming of the tone memory TON-SP, rhythms and OO 5 frequencies can be freely combined, so that audible tones can be created in any desired form.

For the formation of advisory tones, which consist of different frequencies sent one after the other, one is Combination logic VL1 provided. This combination logic VL1 is activated when from the time slot memory ZL-SP a bit combination is read out, which identifies an advisory tone time slot. The TA-SP clock memory has special features Bits for advisory tones are provided, with which the addressing of the tone memory TON-SP being affected. In the combination logic VL1 several bits of the clock memory TA-SP with at least one Bit of the time slot memory ZL-SP linked, so that several addressing bits of the tone memory TON-SP each can change, although this is only shown single-core in the drawing. The clock memory TA-SP causes another memory area in the tone memory TON-SP after the time lapse specified for a tone frequency is controlled, from which PCM word groups of a different frequency are read. According to the number of bits which are provided for this purpose in the clock memory TA-SP and to vary via the combination logic VL1 the addressing of the tone memory TON-SP can be used, the number of which can be transmitted one after the other

JO different tone frequencies can be of different sizes.

A speech memory SPE-SP, which is independent of the tone memory TON-SP, is provided for voice announcements, which addresses only then is when a signal is output from the timing memory ZL-SP or from the control memory ST-SP that indicates a time slot in which an announcement is to be sent. By means of a multi-core routing R2 that is shown in the drawing only single-core, it can be achieved that the rough addressing of the speech memory SPR-SP either via the time slot memory ZL-SP or via

the address counter ADZ takes place. So the control of memory areas, as in the OO drawing shown, from time slot memory ZL-SP

^ _x J pending, this results in shorter announcements in different * - * 5 time slots. But are the be-OO

The relevant bits of the gross address of the voice memory SPR-SP are connected to the more significant bits of the address counter ADZ, in this way several areas of the voice mailbox SPR-SP can be combined by changing them directly in time are controlled one after the other, so that announcement texts of greater length can be formed.

The output from the digital generator, i.e. from the sound memory TOIT-SP, the speech memory SPR-SP or the PCM words originating from zero generator HG arrive in a Output buffer AP if there is a time slot in which an acoustic signal is to be sent. To this end a pulse is then emitted from the control memory ST-SP, with which the output buffer AP for reception of a PCM word is turned on. Via a downstream A serial bit sequence is generated from each PCM word in a known manner from the parallel-serial converter PS. The serial PCM word is either output directly for further processing at a multiplex output AM or with Using a delay device VZ via an output multiplexer A-MUX in an existing busbar system fed in on time. The output multiplexer A-MUX is controlled in such a way that it only works in the time slots in which the digital generator has to deliver a PCM word, picks it up and forwards it to the receiving bus E-SS. In all other time slots the output multiplexer A-MUX switches the sending busbar S-SS directly to the receiving busbar E-SS through. The digital generator can thus be easily adapted to existing PCM systems.

P 3893

Gö / we

18 = 6.84

Claims (1)

Telefonbau and! Tormalzeit GmbH, "" gOÖÜ Frankfurt am Main, Mainzer Landstrasse 128 - 146 Digital generator for in a PCM time division multiplex telecommunication, in particular, acoustic signals to be fed into the telephone system Claims 1. Digital generator for a PCM time division multiplex telecommunications, in particular telephone system to be fed acoustic signals, in which these signals are digital as Groups of PCM words are stored, with frequencies and rhythms separated for each signal are stored and logically linked as required, characterized, that in a time slot memory (ZL-SP) for each time slot a line is provided in which the designation of the acoustic signal to be sent in this time slot is stored is that with the data read out from the time slot memory (ZL-SP) memory areas in one Tone memory (TOl-T-SP) or a voice memory (SPR-SP) are controlled, in which the for single frequencies or - Announcement texts stored in digitally saved PCM word groups are that a clock memory (TA-SP) is provided, which is continuously addressed by a sampling counter (AZ) and delivers rhythm control words that have a Clock multiplexer (TA-MUX) linked with control information which are read out together with the PCM words stored in the tone memory (TOK-SP), with which the acoustic signal in frequency and rhythm 23818 set _{isti that} time slot memory Tone memory (TON-SP), voice memory (SPE-SP) and clock memory (TA-SP) are designed as programmable read-only memories, with which tones, voice announcements and pauses can be freely assigned to each type of signal, and that for voice announcements by optional addressing (R2 ) of the voice mailbox (SPR-SP) either several shorter or a few longer announcements can be implemented.
10 2. Digital generator according to claim 1, characterized in that that for the individual addressing of the lines in the tone memory (TON-SP) and in the speech memory (SPR-SP) an address counter (ADZ) in connection with an address buffer (ADP) provided is, with one line for each of the time slots in the address buffer (ADP) that can be assigned an acoustic signal is provided in each time a PCM word is read from the tone memory (TOIT-SP) or the speech memory (SPR-SP) the control address is temporarily stored for it, which will be used for the next control for read out the same time slot, loaded into the address counter (ADZ) and increased by 1, whereby the control address for the next PCM word arises. 3. Digital generator according to claim 1, characterized in that that a control memory (ST-SP) is provided which has a memory area with several for each time slot Has lines per time slot that are continuously addressed individually by a clock counter (TZ) that this Control memory (ST-SP) is designed as a programmable read-only memory and according to its Programming within each time slot occupied by acoustic signals, clock-accurate control clocks for memory, Counter and logic logic (VL) supplies. 4-. Digital generator according to Claim 1, characterized in that
that an output buffer (AP) for PCM words with a subsequent switched parallel-serial converter (P / S) is provided, over which the PCM words serially with one for a PCM time division multiplex system required speed - are output, 5. Digital generator according to claim 4-, characterized, that the parallel-serial converter (P / S) has a delay device (VZ) is connected downstream, and that an output multiplexer (A-MUX) connected to it is an existing one Busbar system (S-SS, E-SS) is switched to the digital generator whenever a Time slot occupied by an acoustic signal is present, one PCM word then being output in a timely manner will. 6. Digital generator according to claim 1, characterized in that that in the case of audible tone sequences (warning tone) consisting of several frequencies, special areas in the tone memory (TON-SP) can be controlled one after the other by taking from the time slot memory (ZL-SP) and from the clock memory (TA-SP) bits read out with one another via a link logic (VLI) be linked, 7 · Digital generator according to claim 1, characterized in, that with bits read out from the clock memory (TA-SP) a combination logic (VL2) is controlled which delivers an output signal to the during the pauses occurring with acoustic signals Control of a zero generator (NG), which is then a PCH word with the meaning "zero" generated. P 3893 Gö / we 6/18/84