DE1172742B - Process for the central acquisition of signal pulses occurring on telecommunication lines in random sequence, in particular of counting pulses in telephone systems - Google Patents

Description

FEDERAL REPUBLIC OF GERMANY

GERMAN

PATENT OFFICE

EDITORIAL

Boarding school Class: H 04 w

Number:
File number:
Registration date:
Display day:

German class: 21 a3- 75/60

S 81886 VIII a / 21 a3
4th October 1962
June 25, 1964

To record on telecommunication lines in random order, but with a certain time Minimum interval between signal pulses, in particular from counting pulses in telephone systems, it is already known to continuously scan the individual signal lines in a cyclical sequence. This scanning is generally done via coincidence gates, ζ. B, diode gates that in turn be steered up. If a scanning pulse meets a signal pulse, it becomes that resulting control pulse fed to a registration device. To ensure that all accruing Signal pulses are reliably detected, it is necessary that all signal lines during the Duration of the shortest signal pulse can be sampled once.

The registration of the individual signal pulses can be done in such a way that the count of the Sampling device fed to an automatically readable recording medium instead of a signal pulse or that in a memory arrangement with the individual signal lines individually assigned Memory sections the line-specific information contained in these depending on the position of the scanning device are stored and stored back again, depending the information stored in each case is changed or unchanged by the presence of a control signal is restored. The information change can consist, for example, in each case a "1" is added.

Furthermore, according to another method, it is already known to add bistable storage elements to the signal lines, ζ, B. ferromagnetic toroidal cores, for the temporary storage of one of the incoming Assign signal pulses individually, so that the signal pulse to be registered is easier Way is extended beyond the actual pulse duration, and the storage elements in place of the To scan signal lines, the resetting of the memory elements takes place simultaneously with the Sampling in the pause between two signal pulses. As a result of the intermediate storage of the individual signal pulses can control the scanning speed and thus the processing speed of the recording device be chosen lower than in the first-mentioned method.

Since, in general, the incoming signal pulses on the signal lines to be interrogated periodically are formed by closing or opening contacts and these tend to bruise very easily, each bruise signifying a glitch, be-method for the central detection of on
Telecommunication lines in random order
resulting signal pulses, in particular from
Counting pulses in telephone systems

Applicant:

Siemens & Halske Aktiengesellschaft,

Berlin and Munich,

Munich 2, Wittelsbacherplatz 2

Named as inventor:

Dipl.-Ing. Ullrich Tanke, Graefelfing

With both known methods there is the risk that one and the same signal pulse will be evaluated several times is, namely every time the interrogation pulse falls into a bounce gap.

To avoid such multiple counts, it is also known to provide a further binary memory element in the form of a marker bit for each signal line, in which the last query result is recorded until the subsequent query result arrives, and a registration only when a signal state changes, e.g. »1 «, In the others, eg» 0 «, which can be determined in a simple way by comparing two successive query results.
The invention also relates to a method for detecting signal pulses arriving on several signal lines in random sequence, but with a certain minimum time interval, in particular counting pulses in telephone systems, by periodically scanning the signal lines in a time sequence that is shorter than the duration of the shortest Bounce-free signal pulse, using the signal lines individually assigned, for example combined in the form of a matrix, bistable memory elements, e.g.

ferromagnetic toroidal cores. The invention is characterized in that the scanning The registration pulses occurring on the signal lines are fed to the individual memory elements, that the memory elements are coupled in time with the sampling pulses for the signal lines and with respect to the respective temporally closest sampling pulse by at least by

405 627/106

The greatest possible disturbance time caused by bruises on the die the criterion for the final recording At the beginning or at the end of a signal pulse phase control pulses forming a signal pulse can

shifted reset pulses in a time of the central recording device directly

borrowed sequence are deferred, which is greater than or allocated via downstream intermediate storage

the longest undistorted signal pulse, with only leading in 5 and the recording in known

the pause between two signal pulses back-type trigger.

can be made, and that only with the return When using downstream intermediate position The control pulses triggered by the storage elements can store the scanning pulse and the return pulse the central recording device actuating pulse to all signal lines or all memory directly or indirectly via intermediate memory train- io elements are fed simultaneously. At Leads will. On the other hand, orders without a buffer store would have to deviate from that without a buffer store for the reset pulses of the individual storage elements The method that works with signal pulses will be supplied as cyclical one after the other, since the zen-Merkbit Acting individual storage element not central recording device only individually synchronized may be controlled with each line scan 15. However, the impulse and the query result contained therein is not fully utilized at a time following the reset pulses. Comparison used, but the withdrawal. Normally, the reset would be followed due to phase pulses for the individual memory elements on the compared to the sampling pulses shifted separate reset pulses and only available for a complete cycle Pause between two signal pulses. This distributes the period evenly. As a result of the one possible, bounce impulses without the effort of keeping a minimum distance from the in a full one Comparator and the associated logic continuous cycle falling sampling pulses stands every element but only a limited period of time between two for resetting or new storage of the query result, scanning pulses are reliably available for each scan. According to a further exit. As a result of the formation of the invention by bruises, however, it is more advantageous caused disturbance time-dependent phase shift only one group of signal lines between them the scanning of the signal lines and the timely and the individual groups in such a way delayed the associated memory element is set to sample one after the other that the times in any case only once during the downtime - for the delayed resetting of the associated keyed and saved or only once back - 30 groups of memory elements set on without gaps. This remains in the case of storage as a result of different rows.

a bounce pulse at the beginning of a signal pulse. A further embodiment of the invention results the subsequent reset pulse without any effect resulting from the use of band ring cores as since this is in each case with the actual memory signal pulse individually assigned to the signal conductors overlaps, and likewise can with 35 elements. While the previously usual one in a bounce gap at the end of a signal pulse ferrite ring cores a minimum excitation a minimum falling reset pulse do not have to be offered a new injection time in order to complete take place, because the following scanning pulse can constantly be tilted, you can with band ring security in the pause between two nuclei the excitation up to a certain limit other signal pulses falls. 40 also choose lower when you get this excitement On the other hand, the phase shift allows it to act for a longer period of time. The exploitation between the sampling and reset pulses these properties have the advantage that one can learn simultaneous use of the individual memory signal pulses of sufficient length fewer wind elements as a buffer memory, so that the actual gene is required for the input winding. Recording of each signal pulse is essential 45 Furthermore, it is advisable that the individual rings can be carried out more slowly than the scanning of the signal cores in the known manner with pre-excitation for the presence of a signal pulse. drive, as this results in greater tolerances for the

A registration pulse can in fact only be stored while the signal pulses are flowing for the duration of a signal pulse, F i g. Fig. 1 shows an exemplary embodiment for worst-case scenario at the end of the signal pulse. 50 Execution of the process on which the invention is based. The signal pulses arrive via the signal with at least the entire pause time between two lines Si to Sn to the signal pulses available in the form of a matrix M , so that the time-summarized memory elements, which stood between two successive reset scanning device AW cyclically pulse back one after the other accordingly large can be chosen. 55 are placed. Since the recording of a signal pulse only by monitoring the pending signal pulses are derived from the scanning device when the memory elements are reset and via the control pulse, the maximum output e can be fed to the matrix M. The control impulses that occur at the required processing speed of the line-up arrive at a correspondingly lower level via the drawing device. 60, the control line α to the recording means by which the invention is based encryption AE and from this, depending on the driving Stel it is therefore possible, when using only development of the scanning processes. The use of individual assemblies in an individual memory element, which equates the circuit design and acts as a memory bit and as a buffer memory, has been omitted because these multiple counts caused in bruises are certainly sufficiently well known and that the invention is based on to be switched off and the processing speed Method not to lower the recording device compared to the scanning a certain formation of the assemblies delivery speed. that is.

F i g. 2 shows the pulse diagram on which the method is based. The upper curve shows the signal voltage occurring, for example, on the signal line S η. This is followed by the scanning pulses acting on a signal line via the control line e and the reset pulses given to the storage elements by the scanning device AW, as well as the memory state curve and the control pulses fed to the recording device via the control line α. According to the invention, the memory element can only be brought into memory position “1” when a signal pulse coincides with a sampling pulse. The pulse repetition time t _{Pe of} the sampling pulses must be shorter than the shortest duration t _{Jmin of} the interference-free signal pulse. If this time condition is not met, there is a risk that, with two successive scanning pulses, the first will fall into a bounce gap at the beginning and the other will fall into a bounce gap at the end of a signal pulse and the signal pulse will not be detected.

The pulse repetition time t _{Pa of} the reset pulses, on the other hand, can be much longer, since only the condition must be met that it is _{shorter than the minimum pause time} t Pmin between two successive signal pulses, otherwise there is a risk that a storage element brought into the storage position before the arrival of the subsequent registration pulse is not switched back to the starting position and thus only one is registered instead of two signal pulses.

To avoid double counting as a result of the bruises at the beginning of a signal pulse, it is It is also necessary that the scanning precedes the reset pulse so far that at one through a first bounce pulse triggered storage of a register pulse the subsequent reset pulse certainly does not coincide with a subsequent bounce gap. In this case the storage element, before the actual signal pulse has ended, reset and the following Sampling pulse stored again, so that the same in the subsequent pulse pause Signal pulse is registered one more time.

Likewise, the scanning pulse must lag behind the subsequent reset pulse so far that when the memory element is reset in the first bounce gap at the end of a signal pulse, the subsequent scanning pulse does not coincide with a subsequent bounce pulse. In this case, a registration pulse would be stored again before the start of the actual pause and a further registration would be initiated by the subsequent deletion. Each reset pulse must therefore lead or disadvantage the two sampling pulses which are closest in time by a time period t _t or t ₂ which is greater than the greatest possible disturbance time t _prmax caused by bruises.

As a result of this minimum distance between the reset pulses and the scanning pulses, which must be observed, the reset pulses required for the cyclical reset of all storage elements cannot be distributed evenly over the period t _Pa available for a complete operating cycle. Each sampling pulse falling in the reset cycle rather means an interruption, so that in the case of simultaneous sampling of all signal lines, the time determined by the pulse train time t _Fa cannot be fully used for the reset. However, this can be achieved if the sampling pulse is fed to individual groups of signal lines offset in time with respect to one another.

F i g. 3 shows the pulse diagrams of the sampling and reset pulses for three successively activated groups Gl, G2 and G3 of a total of «groups. The group Gl is sampled, for example, at the time t ₀ via the control line e , then the reset pulses for the first group Gl are used later by the fixed disturbance time I _1. The time t _ab available for resetting the memory elements belonging to this group results from the pulse _{repetition time t Pe} for the sampling pulses and the disturbance times t _t and t ₂ at the beginning and at the end of a signal pulse, so that t _ab = t _Fe - (t _± + t _t ). In order to enable the resetting of the storage elements belonging to the following group Gl to follow immediately, this group must already be scanned at the time t _x = t _Pe - (i _x + i ₂ ) = t _ab , that is, the following time for the Sampling of the individual groups of signal lines is equal to the time t _ab that is available for resetting the memory elements of each group. The number of groups required results from the ratio of the pulse repetition time t _Pa for the reset pulses and the group reset time i _from .

Claims (5)

Patent claims: 1. Method for the detection of signal pulses arriving on several signal lines in random sequence, but with a certain minimum time interval, in particular counting pulses in telephone systems, by periodically scanning the signal lines in a time sequence that is shorter than the duration of the shortest bounce-free signal pulse, under Use of the signal lines individually assigned, for example, summarized in the form of a matrix bistable memory elements, z. B. ferromagnetic ring cores, characterized in that the registration pulses occurring during the scanning of the signal lines (Sl to Sri) are fed to the individual memory elements, that the memory elements are coupled with the scanning pulses for the signal lines and at least by at least relative to the scanning pulse that is closest in time the greatest possible disturbance time (t _t or i ₂ = t _prmax ) at the beginning or at the end of a signal pulse caused by bruising, reset pulses are delayed in a time sequence that is greater than the longest undistorted signal pulse, with only in the pause between two Signal pulses can be reset, and that only the control pulses triggered with the resetting of the storage elements are fed to the central recording device (AE) directly or indirectly via intermediate storage. 2. The method according to claim 1, characterized in that all signal lines (51 to Sn) are scanned simultaneously. 3. The method according to claim 1, characterized in that in each case only one group of Signal lines at the same time and the individual groups offset in time one after the other be sampled that the times for the Subsequent resetting of the associated groups (Gl ... Gx) are stringed together without any gaps. 4. The method according to any one of claims 1 to 3, characterized by the use of Band ring cores as storage elements individually assigned to the signal lines. 5. The method according to claim 4, characterized in that the band ring cores with pre-excitation operate. 1 sheet of drawings 409 627/106 6.64 © Bundesdruckerei Berlin