DE2846207A1 - Input signal level control for digital receiver - uses closed loop feedback for setting level limit - Google Patents

Abstract

The proposed control optimises the level of the input analog signal to a digital receiver having an analog first stage. The control of the input level prevents saturation of the front end components and thus give linear response. One channel of a multichannel digital output is used as the level setting feedback in conjunction with a level reference signal. The normal application is to receivers with multichannel output because here no extra control channel has to be added. The system makes use of fast Fourier transform circuits, and where these do not exist in the receiver they have to be added. No ensor for actual input level is required.

Description

Arrangement for level control for digital receiver systems

The invention relates to an arrangement for level control according to the preamble of the claim.

In digital receiver systems, digital technology is used for selection, Frequency conversion, amplification and demodulation are used. It's already a filter bank in the form of a digital transversal filter for filter multiplexing with a Pass-through characteristic made up of equidistant bandpass curves with constant bandwidth exists, in the case of the M - 1 delay elements connected in series with the same delay times are provided at their mutual connection points and at the entrance of the first and at the output of the last delay element there are multiplier units, multiply the signals fed to them by M given weight coefficients and at their outputs there are further multiplier units for N filter channels a multiplication by N M predetermined trigonometric coefficients from a Carry out read-only memory, and in which a slimming device is provided, which links the calculation results of the other multiplier units, known, where the number N of filter channels equals 2t, q s is an integer, is selected in the case of performing the function of the further multiplying units the. Function of the read-only memory of the trigonometric coefficients and the function the summing unit, the series connection of an adder, which its input signals in consecutive order in bundles of thickness F = N and summarizes them superimposed on each other to form N sums of g signals (where ff is an integer), is provided with a Bourier computer, the transformation quantity of which is chosen to be equal to N. and that according to the algorithm of the 'Fast Fourier Transform' '(Fast Fourier Transform, -FS) works (DE-OS 22 62 652).

Such a receiver requires an A / D converter at its input. Since the dynamic range of converters possible today is smaller than the signal dynamics, which, as the ratio of the smallest to the largest signal, can be over 120 dB a level control before the converter is necessary. The level control ensures that the Converter is always fully controlled without being overdriven.

This level control is different from the conventional control Recipients. With the latter, the output level of a channel is kept constant. This is often not necessary in digital systems. Instead, this is the input level kept constant.

Since every A / D converter acts like a hard limiter as soon as its range of numbers is exceeded, strong non-linear values arise even with a low level of overdrive Distortions that are very annoying, especially in the form of intermodulation very reduce the usable dynamic range of the converter.

So it is important to prevent such overmodulation. That Signal at the input is usually a mixture of many oscillations within the bandwidth of the A / D converter, since the selection and thus the selection of one or several individual channels only in the digital filters behind the converter he follows.

The amplitude distribution of such a composite signal is approximate Gaussian, so that there is no absolute security against overloading of the A / D converter is given, even if the signal RMS value C is many times smaller sets as the control limit G.

Since, with decreasing modulation / G, the dynamics used by the A / D converter with respect to the quantization noise decreases, with respect to the distortion products as a result Limitation, but increasing, is opt for a certain optimal modulation (# / G) expect a maximum of effective dynamics.

This optimal degree of modulation depends on the signal statistics. if a single oscillation within the input bandwidth B strongly dominates, then the amplitude distribution is no longer of the Gaussian type and (o / G) 1 / tz2 applies.

opt It is therefore better not to set the level control to a fixed fir / G ratio, but to lay it out adaptively.

The task is therefore to find an adaptive optimal level control to accomplish.

The problem is solved according to the invention by that specified in the claim Measures resolved.

The invention is illustrated below with reference to a drawing of an exemplary embodiment explained in more detail.

If the A / D converter is slightly overdriven by signal peaks an almost white interference spectrum is created. It is superimposed on the white quantization noise of the converter. Both noise sources together can be at a free point in the spectrum to be observed. The band gap Bo - B is suitable for this. Bo is the one through that Sampling rate given bandwidth of the A / D converter. You can get them a little bigger than that Make pre-filter bandwidth B and thus the signal-free gap for the noise measurement create.

For the level control, the noise level in the measuring channel is used as Criterion used. He must in the steady state of the control around a certain factor (approx. 3 dB) above the known and fixed value of the quantization noise lie.

The control loop is shown in detail in the figure. The input signal is fed in analog form to an attenuator 1, from there a low-pass filter or Bandpass filter 2 fed and then fed to a sample and hold circuit 3. From this will it is given to an analog-to-digital converter 4 and then fed to the digital system 5, which in the present case is a multi-channel system, which therefore contains many useful channels 6. A control channel 7 leads from the digital system 5 to a level formation element 8, in which the channel signal is amplitude demodulated and smoothed. The control loop also consists of the subtraction element 9, to which the setpoint value of the level is fed will. The signal processed in this way is used as a disturbance variable in digital or analog form fed to the attenuator 1 of the input part.

In multi-channel systems, the control channel when using the F? T (fast Fourier transformation) is present from the start. The additional effort for the regulation is low. For reasons of cost, single-channel systems will only rarely be implemented digitally. Multi-channel systems without the use of the FFT must have an extra channel.

The advantage of the method described is that it is automatic Adaptation of the modulation to the signal statistics and furthermore that no There must be a sensor in hardware for the input level. The realization of the Sensor on the digital side does not need to mean additional hardware expenditure, if the digital system is implemented by software or firmware.

The method according to the invention is based on the embodiment described not restricted, but is also suitable for other digital systems.

Claims (4)

Claims: 1. Arrangement for level control for digital receiver systems with an analog input part, d a d u r c h e -k e n n n z e i c h n e t that a control loop is arranged between the digital system and the analog input section is. 2. Arrangement according to claim 1, d a d u r c h g e -k e n n z e i c h n e t that the control loop consists of a level forming element for averaging the amplitude and a subtracter for introducing the setpoint. 3. Arrangement according to claim 1, d a d u r c h g e -k e n nz e i c h n e t that the signal for the control loop is a channel of the digital system is removable. 4. Arrangement according to claim 1, d a d u r c h g e -k e n n z e i c h n e t that the control signal in digital or analog form has a damping element, ed can be supplied in the analog input section.