FIELD OF THE INVENTION An aspect of the invention relates to a frequency-division circuit. Other aspects of the invention relate to a method of frequency division and a signal processing arrangement that comprises a frequency-division circuit. The signal processing arrangement may be, for example, a receiver for processing a radio frequency signal that conveys information in the form of audio, video, or data or any combination of those. Yet another aspect of the invention relates to an information-rendering apparatus that comprises such a signal processing arrangement. The information-rendering apparatus may be, for example, a video display set.

BACKGROUND OF THE INVENTION United States patent 4,866,741 describes a 3/2 frequency divider. The frequency divider comprises two D flip-flops. Each D flip-flop has a D input, a clock input, a Q output and a Q-not output. One flip-flop receives an input signal at its clock input, the other flip-flop receives the inverse of the input signal at its clock input. A logic circuit, which is in the form of an OR gate, is responsive to the Q-not output of each flip- flop. The logic circuit provides a low-state signal to the D input of each flip-flop when the Q-not output of at least one flip- flop provides a low-state signal. The logic circuit provides a high-state signal to the D input of each flip-flop when the Q-not output of each flip- flop provides a high-state signal. A further logic circuit, which is in the form of an AND circuit, is responsive to the Q output of each flip-flop and provides an output signal. The output signal has a frequency that is 2/3 times the frequency of the input signal.

SUMMARY OF THE INVENTION It is an object of the invention to allow a reduction of power consumption at a given frequency and/or to allow a higher frequency at the same power consumption. The invention is defined by the independent claims. The dependent claims define advantageous embodiments. According to an aspect of the invention, a frequency-division circuit comprises the following characteristics. The frequency-division circuit comprises a pair of multi-state circuits, each of which can be switched throughout a cycle of states. One multi-state circuit switches to a next state in response to a rising edge in an input signal. The other multi-state circuit switches to a next state in response to a falling edge in the input signal. Each multi- state circuit has at least one state in which the multi-state circuit inhibits the other multi-state circuit so as to prevent the other multi-state circuit from switching to the next state. The invention takes the following aspects into consideration. A frequency divider typically comprises a multi-state circuit composed of, for example, one or more flip- flop circuits. The multi-state circuit receives an input signal with successive rising edges and successive falling edges. The multi-state circuit may switch from one state to another state in response to a rising edge or a falling edge. An output circuit derives an output signal from the multi-state circuit. For example, the output signal may have a low level for one or more states, whereas the output signal has a high level for one or more different states. Accordingly, the output signal has a frequency that is a submultiple of the frequency of the input signal. In the prior-art 3/2 frequency divider identified hereinbefore, each flip-flop constitutes a two-state circuit. At the occurrence of a rising edge, one flip-flop takes a decision, as it were, what its next state will be. The other flip-flop takes such a decision at the occurrence of a falling edge. Each decision on the next state is based on a feedback signal that is present at the D input of the flip-flop concerned. The feedback signal depends on the signal at the Q output of each flip-flop. Consequently, for each decision, the state of each flip-flop must be evaluated so as to derive therefrom a desired new state. This evaluation and decision process, which takes a certain time, imposes a high-frequency limit beyond which frequencies are no longer correctly divided. In principle, it is possible to speed up the evaluation and decision process by increasing the speed of the logic circuitry that is involved in this process, which is the OR gate in the 3/2 frequency divider. However, such a speed increase will generally be at the expense of higher power consumption. Consequently, the prior-art 3/2 frequency divider will need to consume relatively much power in order to divide relatively high frequencies. In accordance with the above-described aspect of the invention, each multi- state circuit switches throughout a cycle of states and has at least one state in which the multi-state circuit inhibits the other multi-state circuit so as to prevent the other multi-state circuit from switching to the next state. The principle of operation is thus that a multi state circuit is from time to time inhibited to switch to the next state. The next state is by definition defined by the cycle of states. Accordingly, there is no evaluation and decision process in order to determine what the next state will be. The inhibiting operation in accordance with the invention can be significantly faster than the prior art evaluation and decision process for a given power consumption. Consequently, the invention allows a higher frequency limit for a given power consumption. Conversely, the invention allows a reduction in power consumption for given frequency. These and other aspects of the invention will be described in greater detail hereinafter with reference to drawings.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram that illustrates a video display set. FIG. 2 is a block diagram that illustrates a frequency conversion circuit that forms part of the video display set. FIG. 3 is a circuit diagram that illustrates a pair of two-state circuits for the frequency conversion circuit. FIGS. 4A and 4B are state diagrams that illustrate the operation of the pair of two-state circuits. FIG. 5 is a time diagram that illustrates the operation of the pair of two-state circuits. FIG. 6 is a circuit diagram that illustrates a pair of three-state circuits for the frequency conversion circuit. FIGS. 7A and 7B are state diagrams that illustrate the operation of the pair of three-state circuits. FIG. 8 is a time diagram that illustrates the operation of the pair of three-state circuits. FIG. 9 is a circuit diagram that illustrates a pair of four-state circuits for the frequency conversion circuit. FIGS. 1OA and 1OB are state diagram that illustrate the operation of the pair of four-state circuits. FIG. 11 is a time diagram that illustrates the operation of the pair of four-state circuits. FIG. 12 is a circuit diagram that illustrates an alternative pair of four-state circuits for the frequency conversion circuit. FIG. 13 is a circuit diagram that illustrates a quadrature generation circuit for the frequency conversion circuit. FIG. 14 illustrates a combination of the pair of two-state circuits and the quadrature generation circuit. FIG. 15 illustrates an output circuit for the pair of multi-state circuits. FIG. 16 illustrates a pair of two-state circuits with embedded switches. FIG. 17 illustrates a switchable latch circuit.

DETAILED DESCRIPTION FIG. 1 illustrates a video display set VDS that comprises a receiver REC and a display device DPL. The receiver REC receives a radio frequency signal RF and provides in response thereto a video signal VID. The display device DPL displays the video signal VID. The receiver REC comprises a front-end circuit FEC, a frequency conversion circuit FCV, a back-end circuit BEC, and a control circuit CTRL. The front-end circuit FEC filters and amplifies the radiofrequency signal RF so as to obtain a frequency-conversion input signal FI. The frequency conversion circuit FCV carries out a frequency conversion so as to obtain an intermediate frequency signal IF. The intermediate frequency signal IF is a frequency-shifted version of the frequency-conversion input signal FI and thus a frequency- shifted version of the radiofrequency signal RF. The back-end circuit BEC selects a certain portion of the frequency spectrum of the intermediate frequency signal IF. The back-end circuit BEC processes that portion of the frequency spectrum so as to obtain the video signal VID to be displayed on the display device DPL.

FIG. 2 illustrates the frequency conversion circuit FCV. The frequency conversion circuit FCV comprises an oscillator circuit OSC, a pair of multi-state circuits MSCA, MSCB, a quadrature generation circuit QGC, and a mixer circuit MIX. Multi-state circuit MSCA can be switched through a cycle of N different states SA(I), .., SA(N), N being an integer. Multi-state circuit MSCB can also be switched through a cycle of N different states SB(I), .., SB(N). In FIG. 2, a circle represents a state. Multi-state circuit MSCA and multi-state circuit MSCB provide enable signals EA and EB, respectively. Enable signal EA has a binary value that depends on the state in which the multi-state circuit MSCA is. The binary value is indicated in the circle that represents the state concerned. Accordingly, FIG. 2 illustrates that enable signal EA is '0' (zero) for all states except state SA(N). In state SA(N), the enable signal is ' 1 ' (one). The same applies to multi-state circuit MSCB: enable signal EB is '1 ' in state SB(N) and '0' in all the other states. FIG. 2 further illustrates that multi-state circuit MSCA receives enable signal EB that multi-state circuit MSCB provides. Multi-state circuit MSCB receives the enable signal EA that multi-state circuit MSCA provides. The frequency conversion circuit FCV operates as follows. The oscillator circuit OSC produces an oscillator signal OS that comprises rising edges Er and falling edges Ef. Multi-state circuit MSCA is responsive to the rising edges Er, whereas multi-state circuit MSCB is responsive to the falling edges Ef. Multi-state circuit MSCA switches from a current state to the next state when a rising edge Er occurs if and only if enable signal EB is '1'. Likewise, multi-state circuit MSCB switches from a current state to the next state when a falling edge Ef occurs if and only if enable signal EA is ' 1 ' . Each multi-state circuit is thus prevented from switching to the next state when the enable signal that the other multi-state circuit provides, is O'. Accordingly, the pair of multi-state circuits MSCA, MSCB provides a pair of frequency-divided signals DOA, DOB, which have certain phase relationship Δφ with respect to each other. Each frequency-divided signals has a frequency that is 2N- 1 times lower than the frequency of the oscillator signal OS. For example, let it be assumed that each multi-state circuit has a cycle of 4 states. In that case, each frequency-divided signal will have frequency that is 1/7 of the oscillator signal OS frequency (N=4; 2N- 1=7). The quadrature generation circuit QGC derives a set of 4 quadrature mixer signals MI+, MQ+, MI-, MQ- from the pair of frequency-divided signals. The set of 4 quadrature mixer signals MI+, MQ+, MI-, MQ- have the following phase relationship: 0°, 90°, 180°, and 270°. Quadrature mixer signals MI+, MQ+, MI-, and MQ- can be associated with a 0°, 90°, 180°, and 270° phase condition, respectively. In effect, the set of 4 quadrature mixer signals MI+, MQ+, MI-, MQ- constitute a vectorial signal that can either have a positive or a negative frequency. The mixer circuit MIX mixes the frequency conversion input signal FI with the set of 4 quadrature mixer signals MI+, MQ+, MI-, MQ-. Accordingly, the intermediate frequency signal IF is obtained. The mixer circuit MIX preferably has a double quadrature architecture as described in, for example, the article "a 48-860MHz digital cable tuner IC with integrated RF and IF selectivity" by Jan van Sinderen et al., ISSCC2003, session 25, RF infotainment, paper 25.3. Such a mixer circuit allows a satisfactory suppression of unwanted signals at so-called image frequencies. FIGS. 3, 6, and 9 illustrates various pairs of multi-state circuits that provide frequency division ratios of 3, 5, and 7, respectively. As explained hereinbefore, a pair of two-state circuits provides a frequency division ratio of 3, which is illustrated in FIG. 3. A pair of three-state circuits provides a frequency division ratio of 5, which is illustrated in FIG. 6. A pair of four-state circuits provides a frequency division ratio of 7, which is illustrated in FIG. 9. FIG. 3 illustrates a pair of multi-state circuits MSCA3, MSCB3 that provides a frequency division ratio of 3. The numeral 3 is used in the reference signs to indicate the frequency division ratio concerned. For example, numeral 3 has been added to the reference signs DOA and DOB illustrated in FIG. 2 so as to indicate that the frequency-divided signals have frequency that is 3 times lower than that of the oscillator signal OS. Multi-state circuit MSCA3 comprises a flip-flop circuit DF31, and an and-gate AN31. Multi-state circuit MSCB3 comprises a flip-flop circuit DF32, an and-gate AN32, and an inverter INV3. Each flip-flop circuit DF is of the "d"-type and has a data input D, a clock input CIk, an output Q and an inverted output Q-. Each flip-flop circuit DF is responsive to a rising edge at the clock input. Inverter INV3 converts a falling edge in the oscillator signal OS into a rising edge. And-gates AN31 , AN32 can be considered to operate as switches under the control of enable signals EB3, EA3, respectively. Flip-flop circuit DF31 receives the oscillator signal OS when enable signal EB3 is ' 1 '. Accordingly, flip-flop circuit DF31 will be responsive to a rising edge in the oscillator signal OS. In contradistinction, flip-flop circuit DF31 will not be responsive to the oscillator signal OS when enable signal EB3 is '0'. Likewise, flip-flop circuit DF32 receives the oscillator signal OS when enable signal EA3 is T. Accordingly, flip-flop circuit DF32 will be responsive to a falling edge in the oscillator signal OS because of inverter INV3. Flip-flop circuit DF32 will not be responsive to the oscillator signal OS when enable signal EA3 is '0'.

FIGS. 4A and 4B are state diagrams that illustrate the operation of multi-state circuits MSCA3 and MSCB3, respectively. FIG. 4A illustrates that multi-state circuit MSCA3 can be switched through a cycle of two states: state SAI3 and state SA23. In state SAl3, the output Q of flip-flop circuit DF31 provides a T and enable signal EA3 is O'. In state SA23, the output Q of flip-flop circuit DF32 provides a 'O'and enable signal EA3 is '1 '. Let it be assumed that enable signal EB3 is ' 1 '. In that case, multi-state circuit MSCA3 switches from state SAl 3 to state SA23 in response to a rising edge Er(i) in the oscillator signal. Multi- state circuit MSCA3 switches from state SA23 back to state SAl 3 in response to a subsequent rising edge Er(i+1). FIG. 4B illustrates that multi-state circuit MSCB3 operates in similar fashion. Multi-state circuit MSCB3 can be switched through a cycle of two states: state SBI3 and state SB23. In state SBl 3, the output Q of flip-flop circuit DF32 provides a '1' and enable signal EB3 is O'. In state SB23, the output Q of flip-flop circuit DF32 provides a 'O'and enable signal EB3 is '1 '. A difference is that multi-state circuit MSCB3 is responsive to falling edges in the oscillator signal, whereas multi-state circuit MSCA3 is responsive to rising edges. Let it be assumed that enable signal EA3 is '1'. In that case, multi-state circuit MSCB3 switches from state SBl3 to state SB23 in response to a falling edge Ef(i). Multi-state circuit MSCB3 switches from state SB23 back to state SBl3 in response to a subsequent falling edge Ef(i+1).

FIG. 5 is a timing diagram that illustrates the operation of the pair of multi- state circuits MSCA3 and MSCB3. FIG. 5 illustrates from top to bottom: the oscillator signal OS, the state of multi-state circuit MSCA3, the state of multi-state circuit MSCB3, enable signal E A3, enable signal EB3, frequency-divided signal DO A3, and frequency-divided signal DOB3. The oscillator signal OS comprises successive rising edges ErI, .., Er9, and successive falling edges EfI, .., Ef9. At rising edge ErI, multi-state circuit MSCA3 switches to state SAl3. In that state, enable signal EA3 is O'. Consequently, multi-state circuit MSCB3 cannot change state when falling edge EfI occurs and remains in state SB23. At rising edge Er2, multi-state circuit MSC A3 switches from state SAl3 to state SA23. In that state, enable signal EA3 is ' 1 '. Consequently multi-state circuit MSCB3 is allowed to switch from state SB23 to state SBl3 when falling edge Ef2 occurs. In state SBl 3, enable signal EB3 is '0'. Consequently, multi-state circuit MSCA3 cannot change state when rising edge Er3 occurs and remains in state SA23. FIG. 5 illustrates that this process continues. When multi- state circuit MSCA3 is in state SAl 3, multi-state circuit MSCB3 cannot change state because enable signal EA3 is '0' in this states. Conversely, when multi-state circuit MSCB3 is in state SBl 3, multi-state circuit MSCA3 cannot change state because enable signal EB3 is '0' in this state.

FIG. 6 illustrates a pair of multi-state circuits MSCA5, MSCB5 that provides a frequency division ratio of 5. The numeral 5 is used in the reference signs to indicate the frequency division ratio concerned. For example, numeral 5 has been added to the reference signs DOA and DOB illustrated in FIG. 2 so as to indicate that the frequency-divided signals have frequency that is 5 times lower than that of the oscillator signal OS. Multi-state circuit MSCA5 comprises two flip-flop circuits DF51 and DF53, a nand-gate NA51, and an and-gate AN51. Multi-state circuit MSCB5 comprises two flip- flop circuits DF52 and DF54, a nand-gate NA52, an and-gate AN52, and an inverter INV5. The flip-flop circuits DF are similar to those in FIG. 3. Inverter INV5 converts a falling edge in the oscillator signal OS into a rising edge. And-gates AN51, AN52 can be considered to operate as switches under the control of enable signals EB5, EA5, respectively. Flip-flop circuits DF51and DF53 receive the oscillator signal OS when enable signal EB5 is '1 '. Accordingly, flip-flop circuits DF51 and DF53 will be responsive to a rising edge in the oscillator signal OS. In contradistinction, flip- flop circuits DF51 and DF53 will not be responsive to the oscillator signal OS when enable signal EB5 is O'. Likewise, flip-flop circuits DF52 and DF54 receive the oscillator signal OS when enable signal EA5 is T. Accordingly, flip-flop circuits DF52 and DF54 will be responsive to a falling edge in the oscillator signal OS because of inverter INV5. Flip-flop circuits DF52 and DF54 will not be responsive to the oscillator signal OS when enable signal EB5 is O'.

FIGS. 7A and 7B are state diagrams that illustrate the operation of multi-state circuits MSCA5 and MSCB5, respectively. FIG. 7A illustrates that multi-state circuit MSC A5 can be switched through a cycle of three states: state SAl5, state SA25 and state SA35. In state SAl5, the respective outputs Q of flip-flop circuits DF51 and DF53 provide a ' 1 '. Enable signal EA5 is O'. In state SA25, the output Q of flip-flop circuit DF51 provides a '0' whereas the output Q of flip-flop circuit DF53 provides a ' 1 '. The enable signal EA5 is still '0'. In state SA35, the output Q of flip-flop circuit DF51 provides a '1 ' whereas the output Q of flip-flop circuit DF53 provides a O'. The enable signal EA5 is now ' 1'. Let it be assumed that enable signal EB5 is '1 '. In that case, multi-state circuit MSCA5 switches from state SAl 5 to state SA25 in response to a rising edge Er(i) in the oscillator signal. Multi- state circuit MSCA5 switches from state SA25 to state SA35 in response to a subsequent rising edge Er(i+1). Multi-state circuit MSCA5 switches from state SA35 back to state SAl5 in response to a further subsequent rising edge Er(i+2). FIG. 7B illustrates that multi-state circuit MSCB5 operates in similar fashion. Multi-state circuit MSCB5 can be switched through a cycle of three states: state SB15, state SB25 and state SB35. In state SBl5, the respective outputs Q of flip-flop circuits DF52 and DF54 provide a T. Enable signal EB5 is O'. In state SB25, the output Q of flip-flop circuit DF52 provides a '0' whereas the output Q of flip-flop circuit DF54 provides a T. The enable signal EB5 is still O'. In state SB35, the output Q of flip-flop circuit DF52 provides a '1 ' whereas the output Q of flip-flop circuit DF54 provides a '0'. The enable signal EB5 is now T. A difference is that multi-state circuit MSCB7 is responsive to falling edges in the oscillator signal, whereas multi-state circuit MSCA7 is responsive to rising edges. Let it be assumed that enable signal EA5 is '1'. In that case, multi-state circuit MSCB5 switches from state SBl 5 to state SB25 in response to a falling edge Ef(i). Multi-state circuit MSCB5 switches from state SB25 to state SB35 in response to a subsequent falling edge Ef(i+1). Multi-state circuit MSCB5 switches from state SB35 back to state SBl 5 in response to a further subsequent falling edge Ef(i+2).

FIG. 8 is a timing diagram that illustrates the operation of the pair of multi- state circuits MSCA5 and MSCB5. FIG. 8 illustrates from top to bottom: the oscillator signal OS, the state of multi-state circuit MSCA5, the state of multi-state circuit MSCB5, enable signal EA5, enable signal EB5, frequency-divided signal D0A5, and frequency-divided signal DOB5. The oscillator signal OS comprises successive rising edges ErI, .., Er9, and successive falling edges EfI, .., Ef9. At rising edge ErI, multi-state circuit MSCA5 switches to state SAI5. In that state, enable signal EA5 is O'. Consequently, multi-state circuit MSCB5 cannot change state when falling edge EfI occurs and remains in state SB35. At rising edge Er2, multi-state circuit MSCA5 switches from state SAl5 to state SA25. In that state, enable signal EA5 is still O'. Consequently, multi-state circuit MSCB5 can still not change state when falling edge Ef2 occurs and still remains in state SB35. At rising edge Er3, multi-state circuit MSCA5 switches from state SA25 to state SA35. In that state, enable signal EA5 is ' 1 '. Consequently, multi-state circuit MSCB5 is now allowed to switch from state SB35 to state SBl5 when falling edge Ef3 occurs. In state SBl5, enable signal EB5 is '0'. Consequently, multi-state circuit MSCA5 cannot change state when rising edge Er4 occurs and remains in state SA35. At falling edge Ef4, multi-state circuit MSCB5 switches from state SBl5 to state SB25. In that state, enable signal EB5 is still '0'. Consequently, multi-state circuit MSCA5 can still not change state when rising edge Er5 occurs and still remains in state SB35. At falling edge Ef5, multi-state circuit MSCB5 switches from state SB25 to state SB35. In that state, enable signal EB5 is T. Consequently, multi-state circuit MSCA5 is now allowed to switch from state SA35 to state SAl5 when rising edge Er6 occurs. FIG. 8 illustrates that this process continues. When multi-state circuit MSCA5 is in state SAl5 or SA25, multi-state circuit MSCB5 cannot change state because enable signal EA5 is '0' in these states. Conversely, when multi-state circuit MSCB5 is in state SBl5 or SB25, multi-state circuit MSCA5 cannot change state because enable signal EB5 is '0' in these states.

FIG. 9 illustrates a pair of multi-state circuits MSCA7, MSCB7 that provides a frequency division ratio of 7. The numeral 7 is used in the reference signs to indicate the frequency division ratio concerned. For example, numeral 7 has been added to the reference signs DOA and DOB illustrated in FIG. 2 so as to indicate that the frequency-divided signals have frequency that is 7 times lower than that of the oscillator signal OS. Multi-state circuit MSC A7 comprises two flip-flop circuits DF71 and DF73, two and-gates AN71 and AN73. Multi-state circuit MSCB7 comprises two flip-flop circuits DF72 and DF74, two and-gates AN72 and AN74 and an inverter INV7. The flip-flop circuits DF are similar to those in FIG. 3. Inverter INV7 converts a falling edge in the oscillator signal OS into a rising edge. And-gates AN71 , AN72 can be considered to operate as switches under the control of enable signals EB7, EA7, respectively. Flip-flop circuits DF71and DF73 receive the oscillator signal OS when enable signal EB7 is '1 '. Accordingly, flip-flop circuits DF71 and DF73 will be responsive to a rising edge in the oscillator signal OS. In contradistinction, flip-flop circuits DF71 and DF73 will not be responsive to the oscillator signal OS when enable signal EB7 is O'. Likewise, flip-flop circuits DF72 and DF74 receive the oscillator signal OS when enable signal EA7 is '1 '. Accordingly, flip-flop circuits DF72 and DF74 will be responsive to a falling edge in the oscillator signal OS because of inverter INV7. Flip-flop circuits DF72 and DF74 will not be responsive to the oscillator signal OS when enable signal EB7 is O'.

FIGS. 1OA and 1OB are state diagrams that illustrate the operation of multi- state circuits MSCA7 and MSCB7, respectively. FIG. 1OA illustrates that multi-state circuit MSCA7 can be switched through a cycle of four states: state SAl7, state SA27, state SA37 and state SA47. In state SAl7, the output Q of flip-flop circuit DF71 provides a '1 ' whereas the output Q of flip-flop circuit DF73 provides a '0'. Enable signal EA7 is O'. In state SA27, the respective outputs Q of flip-flop circuits DF71 and DF73 provide a T. Enable signal EA7 is still O'. In state SA37, the output Q of flip-flop circuit DF71 provides a '0' whereas the output Q of flip-flop circuit DF73 provides a '1 '. The enable signal EA7 is still O'. In state SA47, the respective outputs Q of flip-flop circuits DF71 and DF73 provide a O'. The enable signal EA7 is now T. Let it be assumed that enable signal EB7 is ' 1 '. In that case, multi-state circuit MSCA7 switches from state SAl7 to SA27, then to SA37, and then to SA47 in response to a series of rising edges Er(i), Er(i+1), Er(i+2), and Er(i+3), respectively, in the oscillator signal. FIG. 1OB illustrates that multi-state circuit MSCB7 operates in similar fashion. Multi-state circuit MSCB7 can be switched through a cycle of four states: state SBl7, state SB27, state SB37 and state SB47. In state SB 17, the output Q of flip-flop circuit DF71 provides a '1' whereas the output Q of flip-flop circuit DF73 provides a O'. Enable signal EB7 is O'. In state SB27, the respective outputs Q of flip-flop circuits DF71 and DF73 provide a '1'. Enable signal EB7 is still O'. In state SB37, the output Q of flip-flop circuit DF71 provides a '0' whereas the output Q of flip-flop circuit DF73 provides a T. The enable signal EB7 is still O'. In state SB47, the respective outputs Q of flip-flop circuits DF71 and DF73 provide a 'O'.The enable signal EB7 is now ' 1'. A difference is that multi-state circuit MSCB7 is responsive to falling edges in the oscillator signal, whereas multi-state circuit MSCA7 is responsive to rising edges. Let it be assumed that enable signal EA7 is '1'. In that case, multi-state circuit MSCA7 switches from state SBl7 to SB27, then to SB37, and then to SB47 in response to a series of falling edges Ef(i), Ef(i+1), Ef(i+2), and Ef(i+3), respectively, in the oscillator signal.

FIG. 11 is a timing diagram that illustrates the operation of the pair of multi- state circuits MSCA7 and MSCB7. FIG. 11 illustrates from top to bottom: the oscillator signal OS, the state of multi-state circuit MSCA7, the state of multi-state circuit MSCB7, enable signal EA7, enable signal EB7, frequency-divided signal D0A7, and frequency- divided signal DOB7. The oscillator signal OS comprises successive rising edges ErI, ..,Er9, and successive falling edges EfI, ..,Ef9. At rising edge ErI, multi-state circuit MSCA7 switches to state SAl7. In that state, enable signal EA7 is O'. Consequently, multi-state circuit MSCB7 cannot change state when falling edge EfI occurs and remains in state SB47. At rising edge Er2, multi-state circuit MSCA7 switches from state SAl7 to state SA27. In that state, enable signal EA7 is still O'. Consequently, multi-state circuit MSCB7 can still not change state when falling edge Ef2 occurs and still remains in state SB47. At rising edge Er3, multi-state circuit MSCA7 switches from state SA27 to state SA37. In that state, enable signal EA7 is still '0'. Consequently, multi-state circuit MSCB7 can still not change state when falling edge ED occurs and still remains in state SB47. At rising edge Er4, multi-state circuit MSCA7 switches from state SA37 to state SA47. In that state, enable signal EA7 is T. Consequently, multi-state circuit MSCB7 is now allowed to switch from state SB47 to state SBl7 when falling edge Ef4 occurs. In state SBl7, enable signal EB7 is O'. Consequently, multi-state circuit MSCA7 cannot change state when rising edge Er5 occurs and remains in state SA47. At falling edge Er5, multi-state circuit MSCB7 switches from state SBl7 to state SB27. In that state, enable signal EB7 is still '0'. Consequently, multi-state circuit MSCA7 can still not change state when rising edge Er6 occurs and still remains in state SB47. At falling edge Ef6, multi-state circuit MSCB7 switches from state SB27 to state SB37. In that state, enable signal EB7 is still O'. Consequently, multi-state circuit MSCA7 can still not change state when falling edge Ef6 occurs and still remains in state SB47. At falling edge Ef7, multi-state circuit MSCB7 switches from state SB37 to state SB47. In that state, enable signal EB7 is T. Consequently, multi-state circuit MSCA7 is now allowed to switch from state SA37 to state SAl7 when rising edge Er8 occurs. FIG. 8 illustrates that this process continues. When multi-state circuit MSCA7 is in state SAl7, SA27, or SA37 multi-state circuit MSCB7 cannot change state because enable signal EA7 is '0' in these states. Conversely, when multi-state circuit MSCB7 is in state SBl7, SB27, or SB37 multi-state circuit MSCA7 cannot change state because enable signal EB7 is '0' in these states.

FIG. 12 illustrates an alternative pair of multi-state circuits MSCA7a, MSCB7b that provides a frequency division ratio of 7. The alternative pair of multi-state circuits comprises the same elements as the pair of multi-state circuits MSCA7, MSCB7 illustrated in FIG. 9. Corresponding elements have similar reference signs, the letter "a" being added to the reference sign of each element in FIG. 12. There is a difference in the manner in which elements are coupled. In the alternative pair of multi-state circuits illustrated in FIG. 12, the clock input CIk of flip-flop circuit DF53a is coupled to the output Q of flip- flop circuit DF51a, whereas in the pair of multi-state circuits MSCA, MSCB illustrated in FIG. 9 the clock input CIk of flip-flop circuit DF53 receives the oscillator signal OS via and-gate AN51. In the alternative pair of multi- state circuits illustrated in FIG. 12, the data input D of each flip- flop circuit is coupled to the inverted output Q- of the flip- flop circuit. In contradistinction, in the pair of multi-state circuits MSC A7, MSCB7 illustrated in FIG. 9, the data input D of flip- flop circuits DF71 and DF72 are coupled to the inverted output Q- of flip-flop circuit DF73 and DF74, respectively. The data input D of flip-flop circuits DF73 and DF74 coupled to the output Q of flip-flop circuit DF71 and DF72, respectively. The alternative pair of multi-state circuits illustrated in FIG. 12 is functionally equivalent to the pair of multi-state circuits MSCA, MSCB illustrated in FIG. 9. The embodiment illustrated in FIG. 9 can be regarded as a synchronous implementation, whereas the embodiment illustrated in FIG. 12 can be regarded as an asynchronous implementation.

FIG. 13 illustrates the quadrature generation circuit QGC, which forms part of the frequency conversion circuit FCV illustrated in FIG. 2. The quadrature generation circuit QGC comprises two latch circuits DLl and DL2 of the "d"-type. Each latch circuit DL has a data input d, and output q, an inverted output q-, and a clock input represented by a triangle. The data input d of latch circuit DLl is coupled to the inverted output q- of latch circuit DL2. The data input d of latch circuit DL2 is coupled to the output q of latch circuit DLl. Latch circuit DLl receives the frequency-divided signal DOA at its clock input. Latch circuit DLl receives the other frequency-divided signal DOB at its clock input. It should be noted that in the embodiments described hereinbefore with respect to Figs. 3-11, the frequency-divided signals DOA and DOB are 180° phase shifted with respect to each other. The respective outputs q of latch circuits DLl and DL2 provide quadrature mixer signals MI+ and MQ+, respectively. The respective inverted outputs q- of latch circuits DLl and DL2 provide quadrature mixer signals MQ- and MI-, respectively. The quadrature mixer signals have a frequency that is half the frequency of the frequency-divided signals DOA and DOB that the multi-state circuits provide. That is, the quadrature generation circuit QGC provides a frequency division by two.

FIG. 14 illustrates a combination of the pair of multi-state circuits MSCA, MSCB illustrated in FIG. 3 and the quadrature generation circuit QGC illustrated in FIG. 13. The combination comprises three switches SWl, SW2, and SW3. Each switch has two positions: the position as shown in FIG. 14 and an opposite position. Let it be assumed that each switch is in the position as illustrated in FIG. 14. In that case, multi-state circuit MSCA provides a frequency division by two: the frequency of frequency-divided signal DOA2/3 is half the frequency of the oscillator signal OS. The frequency of the quadrature mixer signals MI+, MQ+, MI-, MQ- will be one quarter of the frequency of the oscillator signal OS. Let it now be assumed that each switch is in the opposite position. In that case, the multi-state circuit MSCA provides a frequency division by three: the frequency of frequency-divided signal DOA2/3 is one-third the frequency of the oscillator signal OS. The frequency of the quadrature mixer signals MI+, MQ+, MI-, MQ- will be one sixth of the frequency of the oscillator signal OS. Consequently, the combination illustrated in FIG. 14 can provide different frequency division ratios with relatively few additional elements.

FIG. 15 illustrates an output circuit that can be combined with the pair of multi-state circuits MSCA, MSCB illustrated in FIG. 3. The output circuit comprises an exor- gate EXOR. One input of the exor-gate EXOR receives frequency-divided signal D0A3, another input receives frequency-divided signal DO A3. An output of the exor-gate EXOR provides a signal whose frequency is two-third of the frequency of the oscillator signal OS. That is, the combination of the output circuit illustrated in FIG. 15 and the pair of multi-state circuits MSCA, MSCB illustrated in FIG. 3, provide a frequency division by 1.5.

FIG. 16 illustrates a pair of multi-state circuits MSCA3a, MSCB3a that is functionally equivalent to the pair of multi-state circuits MSCA3, MSCB3 illustrated in FIG. 3. Multi-state circuit MSCA comprises a switched flip-flop circuit DFS31. Multi-state circuit MSC comprises a switched flip-flop circuit DFS32 and an inverter INV30. Flip-flop circuit DFS31 comprises two latch circuits of the "d" type DL31 and DL33 and a switch SWl 1. Flip-flop circuit DFS32 is similar: it comprises two latch circuits of the "d" type DL32 and DL34 and a switch SW 12. Each latch circuit DL has a data input d, and output q, an inverted output q-, and a clock input, which is represented as a triangle. Latch circuits DL33 and DL34 are locked when the respective clock inputs receive a signal level that is '0', whereas latch circuits DL31 and DL32 are locked when the respective clock inputs receive a signal level that is '1 '. This difference is symbolized by a circle in front of the respective clock inputs of latch circuits DL31 and DL32. The switches SWl 1 and SWl 2 are functionally equivalent to and-gates AN31 and AN32, respectively, illustrated in FIG. 3. Each switch can be switched between an open position and a closed position. Latch circuits DL33 and DL34 are responsive to the oscillator signal OS when switches SWl 1 and SW12, respectively, are in a closed position. Latch circuit DL33 is unlocked when the oscillator signal OS has a level that is T. Latch circuit DL34 is unlocked when the oscillator signal OS has a level that is '0'. In contradistinction, latch circuits DL33 and DL34 are always locked when switches SWl 1 and SW12, respectively, are in the open position. Latch circuits DL33 and DL34 are no longer responsive to the oscillator signal OS. The inverted output q- of latch circuit DL34 controls switch SWl 1. Switch SWl 1 is in the open position when the inverted output q- provides a '0' whereas it is in the closed position when the inverted output q- provides a ' 1 '. The inverted output q- of latch circuit DL33 controls switch SW 12 in a similar fashion. The embodiment illustrated in FIG. 16 allows even lower power consumption at comparable frequencies than the embodiment illustrated in FIG. 3. This advantage is due to the fact that the enabling or disabling of a state change, has been embedded within the flip- flop circuits that constitute the multi-state circuits. This embedding reduces delay.

FIG. 17 illustrates a transistor implementation of a switched latch circuit DLS. The switched latch circuit DLS is functionally equivalent to the combination of switch SWl 1 and latch circuit DL33 illustrated in FIG. 16. The switched latch circuit DLS has a differential data input dd, a differential output qd, a differential clock input clkd, and a switch control input swc. The switched latch circuit DLS comprises various transistors Ql, .., Q7, a current source IT and two resistors Rl, R2. In one state, transistor Q6 is conducting whereas transistor Q7 is non-conducting. In another state, transistor Q7 is conducting whereas transistor Q6 is non-conducting. The current source IT provides a tail current that is divided among transistors Ql , Q2, and Q3. The tail current will substantially not flow through transistor Q3 when the voltage at the base of this transistor is sufficiently low. The tail current will either flow through transistor Ql or transistor Q2, or both, depending on the signal received at the differential clock input clkd. In this case, the switched latch circuit DLS is allowed to change state. However, the switched latch circuit DLS is prevented from changing state when the voltage at the base of transistor Q3 is sufficiently high so that the tail current substantially flows through this transistor. In this case, the tail current will substantially not flow through transistor Ql ; transistor Ql is dry as it were. The result is that the differential pair formed by transistors Q4 and Q5 is switched off. This prevents a signal that is present on the differential data input dd from having any effect on the switched latch circuit DLS. The switched latch circuit DLS is locked in its current state. The pair of multi-state circuits MSC A3a, MSCB3b illustrated in FIG. 16 can be implemented on the basis of the switched latch circuit DLS illustrated in FIG. 17. The combination of switch SWl 1 and latch circuit DL33 can be implemented in the form of the switched latch circuit DLS illustrated in FIG. 17. An inverting circuit should be added so that the voltage at the switch control input swc is high when the inverted output q- of latch circuit DL34 is '0' and vice versa. In a similar manner, the combination of switch SWl 2 and latch circuit DL34 can be implemented in the form of the switched latch circuit DLS illustrated in FIG. 17. Latch circuits DL31 and DL32 in FIG. 16 can also be implemented in the form of the switched latch circuit DLS illustrated in FIG. 17. However, transistor Q3 may be left out. Further circuits may be used to modify the DC component of a signal. Such circuits are commonly referred to as level-shift circuits. For example, one or more level-shift circuits may be used to ensure that the signal applied at the differential clock input clkd illustrated in FIG. 17 has an appropriate DC level with respect to the signal applied at the switch control input swc. The signal applied at the differential clock input clkd will be derived from the oscillator signal OS illustrated in FIG. 16. The signal applied at a switch control input will be derived from the inverted output q of latch circuit DL33 or DL34 illustrated in FIG. 16. It is advantageous that signals and circuits are in a differential form. In this manner, it is relatively easy to invert a signal. A differential signal has two components. It is sufficient to replace one component by the other component and vice versa, in order to obtain an inverted version of the differential signal. This is just a matter of wiring; no circuitry is required. For example, inverter INV31 illustrated in FIG. 16 need not be a circuit if the oscillator signal OS is in the differential form. The same applies to the inverters shown in the other embodiments.

The detailed description hereinbefore with reference to the drawings illustrates the following characteristics. A frequency-division circuit comprises a pair of multi-state circuits (MSCA, MSCB). Each multi-state circuit (MSCA, MSCB) can be switched throughout a cycle of states (SA(I), .., SA(N); SB(I), .., SB(N)). One multi-state circuit (MSCA) switches to a next state in response to a rising edge (Er) in an input signal (oscillator signal OS). The other multi-state circuit (MSCB) switches to a next state in response to a falling edge (Ef) in the input signal (OS). Each multi-state circuit (MSCA, MSCB) has at least one state (SA(I), SB(I)) in which the multi-state circuit inhibits the other multi-state circuit (MSCB, MSCA) so as to prevent the other multi-state circuit (MSCB, MSCA) from switching to the next state.

The detailed description hereinbefore further illustrates the following optional characteristics. The frequency-division circuit comprises an output circuit (the quadrature generation circuit QGC illustrated in Fig. 13 or the exor-gate illustrated in Fig. 15). The output circuit receives an output signal (frequency-divided signals DOA, DOB) from each multi-state circuit (MSCA, MSCB). The output circuit derives a processed output signal (quadrature mixer signals MI+, MQ+, MI-, MQ- , or signal Fos÷3/2) from respective output signals of the multi-state circuits. An advantage of these characteristics is that it is possible to generate a great variety of output signals. The output circuit (in the form of the quadrature generation circuit QGC illustrated in Fig. 13) comprises a first and a second latch circuit (DLl, DL2). Each latch circuit (DLl, DL2) has a data input (d), a clock input (>), an output (q), and an inverted output (q-). The clock input (>) of the first latch circuit (DLl) receives an output signal (DOA) of one multi-state circuit (MSCA). The clock input (>) of the second latch circuit (DL2) receives an output signal (DOB) of the other multi-state circuit (MSCB). The data input (d) of the first latch circuit (DLl) is coupled to the inverted output (q-) of the second latch circuit (DL2). The data input (d) of the second latch circuit (DL2) is coupled to the output (d) of the first latch circuit (DLl). An advantage of these characteristics is that it is possible to generate a set of quadrature signals. Such signals allow good image suppression when applied to a quadrature mixer circuit. The output circuit (in the form of exor-gate EXOR illustrated in Fig. 15) comprises a logic-gate circuit (the exor-gate EXOR). An advantage of this characteristic is . that it is possible to obtain a non-integer division ratio. Each multi-state circuit comprises a switchable latch circuit (DLS illustrated in Fig. 17) having a switch control input (the input swc). The switchable latch circuit (DLS) does not change state when an inhibit signal is present at the switch control input (In Fig. 17, the inhibit signal corresponds to a signal that causes the tail current to substantially flow through transistor Q3). An advantage of these characteristics is that it is possible to achieve even lower power consumption or operation at even higher frequencies, or both. Each multi-state circuit comprises differential pairs (Q1-Q2, Q4-Q5 illustrated in Fig. 17) that receive and provide differential signals (differential input signals dd, clkd and a differential output signal at qd). An advantage of these characteristics is that inversion of a signal requires few circuitries only or even none at all.

The aforementioned characteristics can be implemented in numerous different manners. In order to illustrate this, some alternatives are briefly indicated. The cycle of states throughout which one multi-state state circuit is switched, may be different from the cycle of states throughout which the other multi-state circuit is switched. Perfect symmetry is not required. Accordingly, it is possible to generate phase relationships between respective output signals of the multi-state circuit that are different from the examples given in the description. The multi-state circuit may comprise any type of flip-flop circuit. For example, JK flip-flop circuits can be used to implement a multi-state circuit. There are numerous different output circuits that can be combined with the pair of multi-state circuits. For example, an output circuit may comprise gates other than an exor-gate. Functions may be implemented by means of items of hardware or software, or both. In this respect, the drawings are diagrammatic, each representing only one possible embodiment of the invention. Thus, although a drawing shows different functions as different blocks, this by no means excludes that a single item of hardware or software carries out several functions. Nor does it exclude that an assembly of items of hardware or software or both carry out a function. The remarks made herein before demonstrate that the detailed description with reference to the drawings, illustrate rather than limit the invention. There are numerous alternatives, which fall within the scope of the appended claims. Any reference sign in a claim should not be construed as limiting the claim. The word "comprising" does not exclude the presence of other elements or steps than those listed in a claim. The word "a" or "an" preceding an element or step does not exclude the presence of a plurality of such elements or steps.