FR2534087A1 - Method and device for frequency demodulation. - Google Patents

Abstract

The present invention relates to a frequency demodulator of the type used for analog or digital transmissions. The method of frequency demodulation of frequency-modulated signals consists essentially in performing a discrimination of the duration of presence of each modulating frequency f1 or f2, f1 and f2 being the relevant modulating frequencies. The demodulating device of the invention comprises for example a monostable 2 placed in series with a flip-flop 3, the monostable 2 receiving the calibrated modulated signal E, the flip-flop 3 supplying a "one" or a "zero" upon the appearance of the first front of the second period of each frequency. Application to transmissions over cables or over optical fibres.

Description

 The present invention relates to a frequency demodulator of the type used for anlogic or digital transmission on cables, optical fibers, etc.

 Frequency demodulators are already known; they generally consist of a phase loop permitting analog frequency demodulation. For example, demodulators have been described in the S.C.

Gupta at the Proceedings of the IEEE vol. 63 No. 2 in February 1965.

 However such systems are slow to "hang on" to the correct frequency. They are applicable to the frequency demodulation of FSK (frequency shift keying) coded binary signals of the order of a few tens of kbit / s but do not apply to higher digital bit rates. The response time of such demodulators is often slow and the setting complicated. Indeed the setting is different according to the frequencies used.

In addition, the major disadvantage of these systems is that the maximum ratio, between the frequencies f 1 and f 2 of the signal FSK and the bit rate, not to be exceeded when one wishes to transmit digital data, is of the order of a few tens
The present invention aims to overcome all these disadvantages of analog demodulators.

 The present invention aims to achieve a demodulation even when the frequencies f1 and f2 are very close to each other.

The invention essentially consists in discriminating the duration of presence of each modulation frequency f1 or f2, f1 and f2 being the frequencies of modulation considered.
The device of the invention comprises a monostable place in series with a flip-flop, for example of type D, said monostable receiving the calibrated modulated signal and having a time constant t, T being chosen such that i 41 or fi and f2 are the frequencies of
modulation, said D flip-flop providing a "one" or "zero" of the onset of the first rising (or falling) edge of the second period of each frequency.

According to one variant, the device of the invention comprises placed in series two monostables and a flip-flop, for example of the JK type, the inputs of said flip-flop, receiving, on the one hand, the signal to be demodulated, on the other hand, the signal coming from of the second monostable, the time constants 1 of the first and T2 of the second monostable being chosen to be close to 11 A 1 1 + 1 t and 172 ovt and T2 (1 -
- T1 T2 t Tz fl the output signal of said flip-flop being the desired signal.

 Other advantages and features will become apparent upon reading the following description illustrated by drawings.

 Figure la is a general diagram of the demodulator circuit according to the invention.

 FIG. 1b represents a timing diagram of the signals according to the version of FIG.

 Figure 2a is a modified version of Figure 1.

 Figure 2b shows a timing diagram of the signals according to the version of Figure 2a.

 The incoming signal to be demodulated is for example a digital signal formed by a succession of pulses of frequencies f1 or f2. This signal may as well be an analog FSK module signal. (Frequency - shift keying).

 Referring to FIG. 1a, the demodulator circuit according to the invention consists essentially of a comparator 1 effecting a shaping of the incoming signal and outputting a digital signal E formed of a series of "zeros" and "ones".

The signal E is applied, on the one hand, the input of a second time constant mo # 2, on the other hand, at the input
D of a D-type flip-flop 3. The time constant Z is appropriately chosen as a function of the frequencies f1 and f2 to be demodulated. It is therefore interesting to choose a time constant # such that: 1 = T1 # RS T2 = 1
fi f2.

 OR f1 and f2 are the two aforementioned frequencies and f2 is less than f1. The signal Sm coming from the mon-ostable is applied to the input of an inverter 4 and then applied to the input CK of a D-type flip-flop whose input D receives the signal E coming from the comparator 1. The flip-flop D is triggered by the descending fren of the monostable 2 and outputs a signal S which is a series of "zeros" and "ones", the "zeros" corresponding to the duration of the frequency signal fl while that the "ones" correspond to the duration of the signal of frequency f2. The "hooking" "takes place in about a period of the minimum frequency, namely a period of f1.

 Figure la shows a variant of Figure 1. The modulated input signal is pre-filtered by means of a bandpass filter 6 to eliminate spurious signals.

At the output of the filter 6 the signal is applied to the input of a comparator 7 which outputs a digital signal forms of slots of variable widths. The signal E 'coming from the comparator 7 is applied, on the one hand, to the input of a monostable 8, on the other hand, to the input CK of a flip-flop 10, for example of the JK type. If fl and f2 (with f2 <fl) are the frequencies of the incoming signal to demodulate the monostable 8 has a time constant Z1 = 1 1 + 1) which
2 (f1 f2 makes it possible to remain in the state 1 when the signal E 'of frequency f1 is applied and to make it go to state 0 on the appearance of the frequency signal f2. The rising edge of the monostable 8 is then triggered by the second edge m.on the signal E 'and the signal Q1 returns to 1 at the same time as E. The signal Q1 inverse of Q1 is applied to the input of a second monostable 9. This second monostable 9 has a time constant 72 such that # 2 # '' 1 (1 -1
(f2 f1)
This makes it possible to trigger this second monostable as soon as a falling edge of Q1 appears and to maintain it in the zero state for a duration greater than that of the slot of the signal Q1. The rising edge of the signal E 'corresponding to the frequency f2 is thus framed by the state 1 as regards the signal Q2 obtained on the output Q of the monostable 9. This output Q2 is connected to the input of a flip-flop 10, for example of the JK type whose other input CK receives the signal E 'from the comparator 7. Ai-nsi that represents on the timing diagram of Figure 2b the rising edge of the signal E' corresponding to the frequency f2 switches the flip-flop 10 to the state 1. This signal S <remains in state 1 for the duration of the frequency 2 and returns to the state O of the second rising edge of the incoming signal of frequency f1. It is again noted that the attachment takes place in a period of the frequency fl which is the highest.

 According to the invention, the desired frequency is detected from the end of the first period to the appearance of each frequency.

The process of the invention is therefore very fast and very safe at the same time
Indeed if fl and 2 sought frequencies are very close, for example 30 MHz and 31 MHz, the method of the invention has the same efficiency.

Claims (3)

 1 - Frequency modulated signal frequency demodulation method, characterized by the fact that  a discrimination of the duration of presence of each modulation frequency f1 or f2, f1 and f2 being the frequencies considered.  2 - Device for implementing the process of the re  claim 1 characterized by the fact that it includes  Seen monostable placed in series with a flip-flop for example of type D, said monostable receiving calibrated modulated signal, and having a time-constant Z, being chosen such that 1 & G 1 or fl and f 2 are  f1 f2 modulation frequencies, said D-type flip-flop providing a "one" or a "zero" as soon as the first rising (or falling) edge of the second period of each frequency occurs.  3 - Device according to claim 1 characterized by  the fact that it comprises essentially placed in series two monostable and a rocker for example JK type, the  said flip-flop receiving, on the one hand, the signal to  modulate calibrated, on the other hand, the signal from the second mo  nostable, the time constants 1 of the first monostable  and 2 of the second being chosen adjacent to # 1 # (1 + 1) and - T-2 1 ≈signal of  2 f1 f2 2 f2- T1 signal  Release. said flip-flop being the desired signal.