DE19548172A1 - Low-distortion amplitude limiting for electric signals - Google Patents

Abstract

The input signal is applied to a store (1) and also to peak value (3) and zero-crossing (2) determining stages. The output peak value is compared (4) with a preset control value (max. Pegel) to derive a correcting control signal to be fed to a multiplier (5). The latter modifies the signal level in the store. For each halfwave of the input signal, the peak value is determined. The max. level can be continuously preset.The correction signal is only applied to those input signals exceeding the required threshold. Pref. the correction signal acts on the signal within a necessary delay, which is provided by the store.

Description

When transmitting electrical signals with varying Amplitude over real transmission media must be the maximum permissible amplitude of the transmission channel is observed will. If this is not observed, u. a. the following Problems occur:

• - signal distortion,
• - Interference with other channels not involved, e.g. B. by Adjacent channel interference in radio transmissions,
• - Destruction of transmission components, e.g. B. by Overloading active components such as transistors or Tubes.

Two groups of methods are known for avoiding these effects:
Compression and limitation (clip) procedures.

The compression process is based on the Envelope of the signal the gain of the signal continuously changed so that the signal at least does not exceed a maximum level in the long term. The Time constants with which the gain changes are always a compromise between striving to maintain both the dynamics of the signal largely also not to exceed a maximum level. Man achieves low distortion of the Signals, but must briefly exceed a Accept maximum amplitude.

With the limitation methods, the signal becomes linear amplified as long as the instantaneous value of the signal remains below a specified value. This value exceeded, the signal level is at the maximum Value held d. H. limited. The occurring Significant distortions can result from subsequent Filtering can be reduced. By relocating the Operation in the high frequency range is the number of Distortion products that are not filtering either can be eliminated, further reduced (so-called HF clipper). It is achieved with this method that the signal never exceeds the maximum amplitude, but must accept relatively high distortions.

The present invention is one of the limitations proceed as a reliable limitation to the maximum level is made. The Distortions that occur in the known methods avoided.  

The previous methods change the signal in the area maximum amplitudes. This immediately results Distortion products with a high amplitude. The particular of the present invention is that the Gain only at or near the zero crossing Zero crossing is changed and over a half wave is always constant. It already results from this Low level distortion products. (The term Half wave means any in this application Waveform between two zero crossings of the signal.)

Fig. 1 shows the basic process flow as a block diagram. The following process steps are carried out for each half-wave:

• 1. The signal is delayed in a memory ( 1 ) by a constant time. When the signal enters the memory, the beginning and end of a half-wave is determined by a zero crossing detector ( 2 ). At the same time, the peak value of each half-wave is determined by a peak value detector ( 3 ). Every time a half-wave has completely run into the memory, the associated peak value is also fixed.
• 2. In a comparator ( 4 ) the peak value is compared with the maximum permissible value and a new gain factor is calculated.
• 2.1 Is the peak value lower or equal to the maximum permissible value, this half-wave passes through unchanged the memory (gain = 1).
• 2.2 If the peak value is higher than the maximum permissible value, this half-wave is limited as follows: A separate amplification factor v is calculated for this half-wave. v = u _max / u _crest (u _max = maximum permissible level, and _crest = peak value of the half-wave) Still in the memory is the full half-wave (and only this very half-wave) multiplied v by an amplifier / multiplier (5) with the calculated factor . The peak value of this half-wave is then exactly at the maximum permissible value.
• 3. The values running out of the memory are not changed any further. The process is over.

The memory or the possibility of storing the signal sections with their own gain factors is a special feature of this process.

The maximum allowable delay time of the memory is not limited by the procedure. The minimum necessary Delay time is the reciprocal of the lowest too transmitting frequency. This is necessary so that  the lowest frequency a full half wave in the Memory fits.

The implementation of the method can be done by interconnecting components or assemblies. Fig. 1 take place. The storage is a specialty. For the realization you will e.g. B. Use random access memory, which is a suitable external circuit or programming to a memory similar to a shift register with a constant cycle time.

Especially because of the special memory implementation of individual or all steps as software recommend in a suitable digital calculator. Until the memory are all functions as elementary Operations on a digital computer with analog input and Exit available. The memory can be used as software Ring buffers can be implemented in the working memory, whereby Read and write pointers at a constant distance be forwarded. This results in the constant Delay time for continuous signals. Since the Memory in the working memory can be used at the same time Gain change optionally added to each signal value be gripped.

In summary, it can be said that the invention a reliable amplitude limitation of a signal least distortion allowed.

Claims (2)

1. A method for limiting a signal to a maximum level with the step of influencing an input signal by a correction signal,
characterized in that

• the peak value of the level is determined for each half-wave of the input signal,
• - the maximum level can be specified continuously,
• the correction signal is determined from the peak value of the half-wave and the maximum level,
• - the correction signal acts on a complete half-wave,
• - Signals below the maximum level are not affected.

2. The method according to claim 1, characterized in that the correction signal is already within the necessary Delay memory acts on the signal.