Patent Application
20060251428
This invention relates to a method for converting a received optical pulse train into an electric output pulse train, wherein the received optical pulse train has pulse packets and packet pauses and wherein the received optical pulse train is converted into a light-dependent current pulse train that is transformed into a voltage pulse train, whose amplitude is controlled and whose pulses are transformed into digital output pulses of the output pulse train.
The invention also relates to an assembly for implementing the method to convert a received optical pulse train into an electric output pulse train.
It is known in prior art that an infrared receiver is comprised of a bias voltage generator for the photodiode, a transimpedance amplifier for transforming the photocurrent into a voltage, and a comparator for forming a digital signal for further processing.
Such solutions use DC control loops to compensate for offsets occurring in circuits for processing differential signals as a result of asymmetries or to compensate for any existing DC light.
The control loop that is very slow when using a DC-coupled amplifier generates pole sites in the full transmission function that generate a higher-order polynomial (>2).
One principle of every infrared transmission, for example, according to the IrDa standard, is to transform only the light pulses sent. However, further pulses may also result from the higher-order system, particularly after long data transmissions, which are characterized by long pulse packets, from the asymmetrical signal character due to the method of data transmission or to the transmission protocol, and/or from high photocurrent.
The asymmetrical signal character is caused by the introduction of energy into the system by light pulses only. However, the return to the operating point is carried out as an intrinsic process with a system-specific time constant.
The control assembly described is, therefore, loaded with a rectified voltage that corresponds to the integral of the input signal as a function of the signal intensity and the duty cycle. This charge is cleared in the form of an amplifier overshooting and results in one or more additional pulses, the so-called ‘satellites, after receiving a pulse packet and when there are no additional light pulses at the receiver's input.
A further drawback of the prior art is that after a high light intensity transmission the amplifier is unable to receive a signal of low intensity for a relatively long time. The result is the so-called latency time during which the system returns to its basic state.
The invention therefore has as its objective to provide a method and a corresponding circuit assembly for converting a received optical pulse train into an electric output pulse train that is designed to improve transmission quality and to reduce the latency time.
According to the invention, this objective is achieved by a method of the type mentioned above for converting a received optical pulse train into an electric output pulse train by transforming the current pulse train into a first voltage pulse train in a controlled manner and by then transforming that said first voltage pulse train into a second voltage pulse train through controllably limiting its amplitude. As a function of the amplitude of said second voltage pulse train, a third voltage pulse train is then generated below a first amplitude size that is free of a static offset of said second voltage pulse train. Above the aforementioned first amplitude size, the third voltage pulse train is free of said static offset only to a reduced degree, and above a second amplitude size larger than the aforementioned first amplitude size, said third voltage pulse train is generated free of a dynamic offset. Below this second amplitude size, said third voltage pulse train is free of said dynamic offset of said second voltage pulse train only to a reduced degree when a packet pause occurs, the amplitude of said third voltage pulse train is set to zero, and the output pulse train is generated from the third voltage pulse train.
The received pulse train, which reaches the receiver diode in the form of light pulses, consists of pulse packets and packet pauses. A pulse packet may include one or more single pulses, each consisting of a certain pulse duration and a certain pulse pause. As a result, if a pulse packet contains only one pulse, the pulse duration is identical to the pulse packet duration and the pulse pause is identical to the packet pause.
The optical input pulses received are converted into a light-dependent current pulse train and subsequently transformed into the first voltage pulse train by a transimpedance amplifier. This transformation involves the control of the amplification of the transimpedance amplifier. In a subsequent step, the first voltage pulse train produced is thus transformed into the second voltage pulse train, likewise involving amplification control. In both transformation processes, the purpose of the amplification control is to produce the second voltage pulse train with identical voltage amplitude for all occurring signal intensities.
The transformation of the second voltage pulse train into the third voltage pulse train is performed using different procedures and depends on the signal amplitude of the second voltage pulse train.
The third voltage pulse train is generated free of a static offset below a first amplitude size, that is, the undesired offset produced by transforming the pulse trains up to the second voltage pulse train is separated from the useful signal and thus no longer has any impact on the subsequent process steps, for example, on a downstream comparator, whose signal transformation is falsified by any offset at its input.
Above a second amplitude size, the third voltage pulse train is produced free of a dynamic offset, that is, the static offset produced by transforming the pulse trains up to the second voltage pulse train is not separated from the useful signal. A high-pass circuit performs this step, for example. In the transmission of a signal of asymmetrical signal character, an undesired integration of the mean values of the voltage pulse trains occurs at the high-pass output. To avoid this effect, herein referred to as dynamic offset, integration is eliminated for the voltage pulse trains above the second amplitude size.
The effect of the static offset separation decreases as the amplitude size of the second voltage pulse train increases in the area above the first amplitude size. Likewise, the effect of the dynamic offset reduction decreases below the second amplitude size as the amplitude of the second voltage pulse train decreases. This decrease may show a linear or a non-linear functional course.
The third voltage pulse train generated in these ways is set to zero when a packet pause occurs during signal transmission; that is, in order to improve the accuracy of subsequent steps, any error occurring in the signal amplitude is eliminated.
In the final step the third voltage pulse train is transformed into the output pulse train. In this process, for example, the signal conditions required for downstream steps are produced.
In one embodiment of the invention all electric pulse trains are differential signal pulse trains.
In the method according to the invention, the conversion of a received optical pulse train into an electric output pulse train can be carried out using both rectangular signals and differential signals.
In an additional embodiment of the invention for asymmetrical pulse trains of the second voltage pulse train a dynamic offset separation is carried out to produce the third voltage pulse train, and the second voltage pulse train is subjected to high-pass filtering using a capacitor. In each pulse pause, a discharge of the capacitor by an amount that is a function of the size of the amplitude of the second voltage pulse train is performed.
The transformation of the second voltage pulse train into the third voltage pulse train is performed using an assembly of a high-pass character. However, when transmitting signals of asymmetrical signal character, there will be an undesired up-integrating of the voltage pulse trains at a high-pass output. To avoid these dynamic offsets the process of integration is counteracted as per the invention. To this end, at least a partial discharge of the capacitor is effected during the pulse pauses. The amplitude of the second voltage pulse train determines the intensity of the discharge in direct proportionality.
In a special embodiment of the invention each pulse of the output pulse train starts a time comparison with a predefined time period, so that after the lapse of the predefined time period a “lock” signal is generated marking the end of a pulse packet.
The method according to the invention ensures continuous detection of a running data transmission and the end of the data transmission. To implement this detection, each pulse of the output pulse train starts a time comparison with a predefined time period. If, for example, the last pulse of an output pulse train packet has started the time comparison and no additional pulses arrive before the conclusion of the time comparison, then upon the end of the time comparison, a signal is generated that indicates the end of the current data transmission. If a further pulse arrives during the time comparison, the comparison process is reset and the time comparison starts again from the beginning.
In one embodiment of the invention the time period is shorter than a latency time and longer than the longest pulse pause.
For example, after a high light intensity data transmission, an amplifier of a converting unit is unable to properly process a signal of low signal intensity for a length of time that depends on the overall configuration. This period is the latency time, the definition of which determines the upper limit of the time range for the time period. When dimensioning the upper limit of the time period, the maximum pause time within the STR protocol of 115 kBit/s must also be considered.
The lower limit for the time period is defined by the pulse width of the slowest transmission mode to be supported. If a data transmission is conducted, for example, in the SIR protocol at 9.6 kBit/s, the maximum pulse width is 22 μs. Accordingly, the time period must be longer than these 22 μs in order to prevent a lock signal from being generated during a long pulse, i.e., before its termination.
In one embodiment of the invention a first “AGCSH” control voltage is generated as a function of the amplitude of the second voltage pulse train to control the amplitude of the first voltage pulse train that is generated in the transformation of the current pulse train into a voltage pulse train.
According to the invention, at least the transformation of the current pulse train into the first voltage pulse train is carried out so that it can be controlled. The “AGCSH” control voltage is generated as a graduated, i.e., analog voltage to avoid abrupt changes in the pulse position or pulse width and to prevent the formation of satellite pulses.
In one embodiment of the invention a second “zero” control voltage is generated as a function of the first “AGCSH” control voltage for setting the amplitude of the third voltage pulse train to zero when a packet pause occurs if the amplitude of the second voltage pulse train exceeds a first amplitude size and the lock signal is active.
Controlled by the “AGCSH” control voltage generated, the second “zero” control voltage is formed. The purpose of this control voltage is to set the amplitude of the third voltage pulse train to zero when a packet pause occurs, but the “zero” control signal is formed only if the amplitude of the second voltage pulse train exceeds a first amplitude size and the lock signal indicates the end of data transmission. Setting the amplitude of the third voltage pulse train to zero reduces the high-pass time constant and it also prevents any differential signal activation of a downstream assembly for generating the digital output pulse train when the differential signal technique is used.
In a special embodiment of the invention a third “limiter” control voltage is generated as a function of the first “AGCSH” control voltage and thus the amplitude of the second voltage pulse train is limited.
Aside from the controlled transformation of the current pulse train into the first voltage pulse train, the subsequent transformation of the first voltage pulse train into the second voltage pulse train is also carried out in a controlled manner. Ideally, this control is effected by the “limiter” control signal in such a way that the second voltage pulse train has an identical amplitude for all signal intensities. The “limiter” control voltage is generated as a function of the “AGCSH” control voltage, so that, for example, after reaching the minimum amplifying value of the “AGCSH” control voltage, control is continued for the further reduction of the amplitude by changing the “limiter” control voltage in order to avoid the interference of non-linear large signal effects occurring with large signals.
In another embodiment of the invention a fourth “short” control voltage is generated as a function of the first “AGCSH” control voltage and used to generate a third voltage pulse train that is free of the static or the dynamic offset.
The fourth “short” control voltage is also generated as a function of the size of the signal amplitude. This control voltage controls the generation of the third voltage pulse train that is generated free of any static offset when small signal amplitudes occur, that is, the constant component of the second voltage pulse train is eliminated.
The static offset separation is switched off using “short” for large signal amplitudes of the second voltage pulse train. Thus, for example, a reduction of the dominant pole sites from two to one is achieved, resulting in an improvement in the return behavior of the third voltage pulse train into the operating point following an oscillation.
In a further embodiment of the invention a fifth “level” control voltage is generated as a function of the first “AGCSH” control voltage and used to control the operating threshold and the hysteresis of a comparator during the transformation of the third voltage pulse train into the output pulse train.
Due to the optical transmission principle, the oscillation of an amplifier in the conversion unit occurs only in a positive direction through a “light on” process, wherein the return of the signal level is effected exclusively by means of the intrinsic process of the amplifier. Therefore, the comparator threshold position must be asymmetrical and positive relative to the differential zero point. To avoid any erroneous detection, “level” is used to adjust the operating threshold and the hysteresis of a comparator configuration generating the output pulse train
According to the invention the purpose of an assembly of the above type for converting a received optical pulse train into an electric output pulse train is achieved by connecting the photodiode with each of its electrodes to the inputs of an transimpedance amplifier via a capacitor, connecting the output of the transimpedance amplifier to the input of a buffer amplifier, and the output of the buffer amplifier to the input of a comparator via a high-pass circuit. At the same time, the capacitor of the high-pass circuit is bridged by a controllable element, and the input of an amplifier control assembly controlling the amplification of the transimpedance amplifier is connected to the output of the buffer amplifier, and the output of the amplifier control configuration is connected to a control signal input of the transimpedance amplifier.
A photodiode receiving the optical input pulses is connected to an assembly for bias voltage generation. This assembly is used to achieve a negative bias voltage of the photodiode.
The anode and the cathode of the photodiode are each also connected via capacitors to the inputs of a transimpedance amplifier. The capacitors perform the separation of the DC light or DC current component of the input signal prior to the current/voltage conversion in the transimpedance amplifier, thus providing the first voltage pulse train.
The output of the transimpedance amplifier is connected to the input of a buffer amplifier that transforms the first voltage pulse train into the second with a predetermined amplification V and provides it at its output.
The output of the buffer amplifier is connected to the input of a downstream high-pass circuit. According to the invention, this high-pass circuit has the special feature of bridging the capacitor of the high-pass circuit with a controllable element. For example, this controllable element may be a transistor that allows the capacitor of the high-pass to become effective in the signal path for the separation of the direct voltage or neutralizes its function. Due to the characteristic curve of the transistor and a corresponding activation, further states are possible between these two extreme states. The high-pass circuit converts the second voltage pulse train into the third voltage pulse train.
The output of the high-pass circuit is connected to the input of a comparator that converts the third voltage pulse train into the output pulse train and provides it at the output pulse train output.
For the controlled conversion of the pulses, an amplifying control assembly is connected at its input to the output of the buffer amplifier and at the output to the AGC input of the transimpedance amplifier. The amplifying control assembly controls the amplitude of the first voltage pulse train for the purpose of keeping the amplitude constant for varying received light intensities.
In another embodiment of the invention connections between the transimpedance amplifier and buffer amplifier, the buffer amplifier and high-pass circuit, and high-pass circuit and comparator assemblies consist of two signal paths for transmitting differential signals.
The assembly according to the invention can be adapted for processing rectangular pulses or, by appropriately designing the assemblies, for processing differential signals.
In a further embodiment of the invention the high-pass circuit has two inputs, an input IN connected to a first contact of a first capacitor and an input IP connected to a first contact of a second capacitor. At the same time, a second contact of the first capacitor is connected to an output ON, and a second contact of the second capacitor is connected to an output OP of the high-pass circuit. In this same embodiment, a first voltage divider is provided at the input, and a second voltage divider is provided at the output, whose center taps are each connected to each other. The first and the second capacitors are each bridged by a source-drain structure of a MOS-FET as the respective controllable element, the gate contacts of the two transistors are connected to an input “short,” and the outputs ON and OP are each connected to a contact of a source-drain structure of a third MOS-FET whose gate contact is connected to the “zero” input.
For processing signals by means of the differential signal technique, the high-pass circuit according to the invention consists of two symmetrical high-pass circuits, each of which has one capacitor connected to an input and the output of the assembly and a resistor. The two resistors are connected and positioned between the outputs of the high-pass circuit and have a first center tap. The circuit further consists of a voltage divider at the input, consisting of two resistors. This divider is positioned between the two inputs of the circuit and has a second center tap. The first and the second center taps are connected to each other. According to the invention, a voltage means of the upstream buffer amplifier is introduced at the second center tap, so that the voltage is superimposed on the signal at the output of the high-pass circuit via the voltage divider at the output by means of the connection of the center taps, without the assembly interfering with the effect of the offset separation for the differential signal. This transfer of the buffer operation point to the input of the comparator constitutes an efficient way of generating the operation point for the comparator input.
The two high-pass capacitors are each bridged with a source-drain structure of a MOS-FET, which may be, for example, a p-channel transistor, whose gates are connected to the “short” control signal, for example, for bridging the high-pass capacitor in the event of high input signal intensities, which leads to a pole reduction of the entire system.
Another MOS-FET with a source-drain structure is positioned between both outputs ON and OP of the high-pass circuit. Its gate contact is connected to the “zero” control signal. This control signal allows, for example, the reduction of the output signal to zero for a predetermined time upon detection of the end of a transmission. Thus, the downstream comparator cannot detect erroneous pulses. At the same time, this reduces the time constant of the high pass, thus allowing a fast recharge of the static offset of the upstream buffer amplifier output to the high-pass capacitors.
In a further embodiment of the invention an output of the comparator is connected to the input of an interlocking logic that signals a pulse packet end, and a first output (“time”) of the interlocking logic is connected to a second input (“time”) of the amplifying control assembly.
An output of the comparator providing the output pulse train is connected to the “comp” input of an interlocking logic. The purpose of this interlocking logic is to detect the end of a current data transmission and to generate a “time” output signal indicating the end. For this purpose, the assembly includes a timer that carries out a time comparison with a predefined time period. The timer is restarted after each detected and provided output signal such that the “time” output signal is provided only after the time comparison following upon an at least preliminarily final output pulse. The “time” output of the interlocking logic is connected to a downstream input of the amplifying control assembly. This signal makes it possible to selectively influence the amplification control, for example, to accelerate the upward-regulation of amplification in the packet pauses for the purpose of reducing the system's latency time.
In a further embodiment of the invention a second output (“lock”) of the interlocking logic is connected to a first input (“lock”) of an end control assembly, a second input (“AGCSH”) of the end control assembly is connected to the output of the amplifying control assembly, and a first output (“zero”) of the end control assembly is connected to the “zero” input of the high-pass circuit.
A further output signal of the interlocking logic is the “lock” output signal. This output signal is generated by combining the output pulse train of the comparator with the “time” signal and is provided at the lock output. This signal also serves the detection of the end of a data transmission. The “lock” output of the interlocking logic is connected to the “lock” input of the end control assembly according to the invention. That assembly's “AGCSH” input is connected to the “AGCSH” output of the amplifying control assembly. By means of this arrangement and also by using combinatorics, the “zero” output signal for the high-pass circuit is generated and is provided at the “zero” output of the assembly. The “AGCSH” signal supplied by the amplifying control assembly is compared with a set point using a comparator. Upon detection of the transmission end, the “zero” signal is supplied to the high-pass circuit for the end processing described above. At the same time, the current amplitude value of the “AGCSH” voltage is compared with the set point value, and once the latter is reached, the logical status of the “zero” signal is changed, and thus the running end processing is terminated.
In a further embodiment of the invention a second output (“intense”) of the end control assembly is connected to a third input (“intense”) of the amplifying control assembly.
A further output signal generated by the end control assembly is the “intense” signal. This signal supplies information on the status of the comparison of the “AGCSH” voltage with the set point performed in the assembly. Connecting this signal output with the “intense” input of the amplifying control assembly in conjunction with the “time” signal allows for the start of the end processing in the amplifying control assembly following the detection of the end of the data transmission and for its selective completion after the set point of the comparator in the end control assembly has been reached. Thus, it becomes possible to reduce the assembly's latency time.
The invention is described in further detail with reference to an exemplary embodiment in the accompanying drawings:
The conversion assembly according to the invention is shown in
The method according to the invention will be discussed with reference to a specific embodiment of the assembly for converting the received optical pulses into electric output pulses for processing differential signals as shown in
The signal path starts at the bias voltage generator 2, which generates the negative bias voltage of a photodiode receiving optical pulses, for example, in accordance with IrDA protocol.
Subsequently, high-pass capacitors 5 separate the continuous light and DC current components at the input of the transimpedance amplifier 3. The latter operates as a current-voltage converter and converts the light-dependent current pulse train generated by the photodiode 1 into the first voltage pulse train. Transimpedance conductances for converting the photocurrent signal of the photodiode are provided, for example, in the form of depletion transistors whose conductance can be changed at the gate by means pf voltage control.
Following in the signal path, the buffer amplifier 6 with a predetermined amplification ‘V’ is coupled with the transimpedance amplifier 3 by direct voltage. Thus, in addition to the pulsed signal voltage of the second voltage pulse train, an offset direct voltage is also present at the output of the buffer amplifier 6.
To minimize the static offset voltage and its temperature course, both the transimpedance amplifier 3 and the buffer amplifier 6 are embodied as bipolar differential amplifier stages.
The high-pass circuit 7 according to the invention is positioned between the buffer amplifier 6 carrying the offset voltage and the comparator 4. This high-pass circuit is incorporated into the total circuit assembly in such a way that for small signal amplitudes of the second voltage pulse train that are below a first amplitude size, an offset separation is performed using the symmetrical high-pass assembly implemented in the high-pass circuit 7. The offset separation is also necessary to permit offset values of the transimpedance amplifier 3 in connection with the buffer amplifier 6 that are within the range of the operating threshold of the comparator 4.
Thus, for small input signals, the dominant poles of both high passes, i.e., those of the continuous light separation high pass defined by the capacitors 5 and those of the capacitors arranged in the high-pass circuit 7 itself, are tolerated. This is possible because the pauses of every IrDA protocol are sufficient for returning the differential output voltage of the high-pass circuit 7 to the setting zero point.
In contrast to this, with large input signal amplitudes of the second voltage pulse train that exceed a second amplitude size, the output differential voltage of the buffer amplifier 6 being significantly above the offset voltage, the symmetrical high pass incorporated into the high-pass circuit 7 is made ineffective by the fourth control voltage “short.” In this way, the dominant poles are reduced from originally two to one. This reduction is advantageous because any approximation to the setting zero point of the differential output voltage of the high-pass circuit 7 has an asymptotical effect.
At the end of each pulse packet with large input signal amplitudes, a further intervention into the high-pass circuit 7 by means of the second “zero” control voltage provides the capability of recharging the high-pass capacitors discharged by the fourth “short” control voltage back to the level of the offset output voltage of the buffer amplifier 6 within a short time. For this purpose, the differential output voltage of the high-pass circuit 7 is bridged to zero for a certain time following the end of a pulse packet. Two effects are achieved in this manner. First, an extreme reduction of the high-pass time constant is achieved by which the offset of the buffer amplifier 6 is transferred directly to the high-pass capacitors of the high-pass circuit 7, and, second, any differential signal activation of the comparator 4 is avoided.
In the area between the first and the second amplitude sizes it is likewise desirable to prevent an activation of the comparator 4 after the end of a pulse packet by means of the second “zero” control voltage. In the case of very small input signals that are below the first amplitude size, the second “zero” control voltage is not active, since interferences can be coupled in by the parasitic elements of the MOS-FET that bridges the differential output voltage of the high-pass circuit 7 to zero.
For efficiency reasons, the output-side common-mode operating points of both outputs of the buffer amplifier 6 are therefore averaged and transmitted to the comparator 4 as the input-side operating point.
A comparator 4 connected to the output of the high-pass circuit 7 detects the voltage pulse train and converts it into the digital output pulse train. Due to the manner of activation of the amplifying assembly as well as of processing signals in accordance with the differential principle where an oscillation of the amplifier is effected by the “light on” status exclusively in a positive direction and the return into the operating point is effected only by the intrinsic process of the amplifier, the comparator operating threshold of the comparator 4 must be located asymmetrically positively relative to the differential zero point of the voltage. The comparator 4 has a “level” control channel that is used according to the invention to control the operating point and the hysteresis of the comparator 4 by means of the “level” control voltage in order to avoid erroneous detections depending on the signal intensity.
The IrDA standard provides for infrared signals to be transmitted in a very high dynamics range. The equivalent electric currents generated by the receiving photodiode are within the range of about 50 nA to above 10 mA.
This corresponds to a dynamics range of more than 5 powers of ten.
For the conversion assembly to be able to transmit all occurring signal currents converted into voltage pulse trains in accordance with the IrDA and according to the protocols defined for the respective speed range, and to convert them into a digital voltage pulse train of specified characteristics regarding pulse duration and pulse position, amplifying control is provided and, to avoid overdrive effects—if the available control range of the transimpedance amplifier 3 cannot prevent them—, limitation of the signal.
An amplifying control assembly 8 that derives its input information from the signal path carries out the amplitude control of the voltage pulse trains. For this purpose, in the assembly according to the invention as shown in
The resulting control loop is intended to influence the transmission parameters of the transimpedance amplifier 3 and/or of the buffer amplifier 6, ideally in such a way that at the output of the buffer amplifier 6 an identical signal is created for all signal intensities of the voltage pulse trains.
As this ideal case is virtually impossible to achieve, due to the characteristic curve properties of the available components and also due to the limited setting range and the technical offsets of the subcircuits due to the very small operating voltages ranging between 1.8 V and 2.4 V depending on the specification or requirement, this control can only be effective up to an upper photocurrent limit. To avoid interference of nonlinear large signal effects occurring with large signals and based on the knowledge of the protocol definitions of the IrDA standard, the signal path is further influenced by the controls according to the invention as described below.
To avoid any negative effects during the signal transmission, such as abrupt changes of the pulse position and/or of the pulse width of the pulse trains or possibly even erroneous pulses, so-called satellites, which may be caused both by changes of the input signal intensity and by changes of the properties of the signal path by the control process, all control signals if possible, generated in the amplifying control assembly 8 must be produced by graduated, i.e., analog signals.
From the first “AGCSH” control signal obtained from the amplifying control assembly 8 that contains information on the signal intensity and is used to regulate amplification, which, in the present case, means to regulate the transimpedance transconductance in the transimpedance amplifier 3, the analog control signals for the signal limitation “limiter” for the buffer amplifier 6, “short” for the corresponding offset reduction in the high-pass circuit 7, and “level” for influencing the operating threshold and/or the hysteresis of the comparator 4 are obtained in accordance with the invention.
The third “limiter” control voltage is generated in such a way that after reaching the control limit for regulating the transimpedance transconductance of the transimpedance amplifier 3, the limitation of the output signal of the buffer amplifier 6 starts.
Furthermore, according to the invention, there are two additional subassemblies in the circuit assembly: interlocking logic 9 and end control assembly 10. Depending on the “AGCSH” intensity information obtained from the amplifying control assembly 8 during signal transmission and by means of a suitable signal management, these components 9 and 10 generate the digital “zero” and “intense” signals. The “zero” signal controls the recharge of the offset to the capacitors of the high-pass circuit 7, and “intense” controls changes in the time response of the above-described “AGCSH” output signal and thus also of the “limiter, “short”, and “level” signals that are dependent on “AGCSH.” As a result, this assembly according to the invention makes it possible to minimize the time interval after which the assembly is again able to receive signals at the sensitivity limit, the so-called latency time, following level control of any intensity.
A circuit assembly (not described here) for adjusting the output pulse train provided by the comparator 4 to the pulse requirements of the various IrDA protocols may be connected to the output of comparator 4.
The subcomponents and their functional characteristics are described below.
The cathode of the photodiode 1 is positioned between the source contact of the voltage source 11 and the drain contact of the current source 12. Using a voltage source 11 at the cathode of the photodiode 1 is advantageous as high operating voltage suppression with regard to the external supply voltage “VDDext” is achieved.
The anode of the photodiode 1 is connected to the connection between the source contact of the current source 12 and, according to the invention, to a grounded shunt connection of a linear resistor 13 and a diode 14. This shunt connection 13, 14 allows a low photocurrent to flow largely unhindered into the input of the transimpedance amplifier 3. This can only be the case if the current divider consisting of resistor 13 of the bias voltage generator 2 and the current in the convergence nodes of the transimpedance amplifier 3 falls in favor of the input current of the transimpedance amplifier 3 via the capacitor 5.
In the case of large photocurrents, wherein the transimpedance amplifier 3 is not able to take in any more due to its level control range, diode 14 ensures that the maximum input voltage of the transimpedance amplifier 3 does not exceed the amount of the diode flux voltage for a large signal. The part solution according to the invention within the bias circuit ensures that, for low photocurrents, in observance of the rule R(13)>>Zin (3), the current divider always falls in favor of the input current.
However, if in the case of a larger photocurrent the flux voltage value of diode 14 is reached, this technique ensures that a negative bias voltage of the photo diode 1 is, nevertheless, maintained. Independently of the size of the photocurrent, the linear component 13 is necessary to allow a maximum negative bias voltage of the photodiode 1 at the sensitivity limit. By redistributing the current division in the case of increasing photocurrent, an increasing portion of the latter is directly diverted to ground and can thus not contribute to an overdrive of the transimpedance amplifier 3.
In this manner the transformation transconductance is made to follow the current-voltage conversion as a function of intensity. Thus a minimum first “AGCSH” control voltage results in a maximum transconductance of the current-voltage conversion. The sensitivity limit is thus defined. The minimum value of the transformation transconductance obtainable by using n-depletion transistors 16 is reached with the control limit of the first “AGCSH” control voltage.
To improve the bandwidth for acceptable operating currents the principle of the feed-forward technique is used in the form of the capacitors 20.
The core piece of the circuit assembly shown in
According to the invention, this high pass is implemented by a voltage divider 28 to average the operating point of the buffer amplifier 6 and of a connection to a tap point between the high-pass resistors 26, by means of which the identical direct voltage signals to the two “OP” and “ON” outputs of the assembly are superimposed on the useful signal in such a way that there is no detrimental influence on the effect of the offset separation.
Furthermore, according to the invention, the contacts of the two capacitors 27 are each connected to the source and the drain contact, respectively, of a p-channel transistor 29 whose gates are controlled by the fourth “short” control voltage also generated in the amplifying control assembly 8. For large input signal amplitudes of the second voltage pulse train, this control signal bridges the capacitors 27 and thus serves the above-described pole reduction of the entire system.
Additionally, the two output nodes “OP” and “ON” of the high-pass circuit 7 in
Due to the associated drastic reduction of the time constant of the symmetrical high-pass 26, 27, whose resistors 26 are bridged by the resistor 30, a fast recharge of the static offset present at the output of the buffer amplifier 6 to the high-pass capacitors 27 occurs at the same time.
The defined time mentioned is determined adaptively by means of the subcircuits 8, 9, and 10 in
Comparator 4 is arranged downstream from the high-pass circuit 7. Its operation is shown in
Likewise according to prior art, a switch point shift with regard to the differential input signal at 31 is effected by load elements 33 and 34 by predetermined operating-point currents. The current source 36 leads to a positive shift of the switch point with regard to the differential input signal. At the same time, a current source 37 is used to generate hysteresis of the comparator by impressing directional information onto either the one or the other load element in the same manner by means of a current switch 38 as a function of the digital output signal. For this shift to be defined, the currents from 36 and 37 must be in a fixed relation to the current source 32 of the input differential amplifier 31, the relation being I(36)+I(36)<I(32).
The capacitors 39 are feed-forward capacitors and compensate for the input capacitor of the differential amplifier. The two bipolar transistors 40 limit the maximum output rise of the amplifier 31. Both measures are advantageous for improving the response time of the comparator.
To perform an adaptation of the operating thresholds to the input intensity of the third voltage pulse train according to the invention, currents 36 and 37 are increased as the input intensity increases while the proportionality to the current of the current source 32 must still be maintained. The fifth “level” control voltage as an analog voltage generated by the amplifying control assembly 8 is used as the intensity information.
To track the operating threshold and the hysteresis of the comparator 4 as a function of intensity according to the invention, the currents of the current sources 36 and 37 are generated from the branch currents in
The circuit assembly consisting of the elements 43 and 50 to 54 generates a branch current from the basic current “I” of current source 43 in transistor 55 as a function of the fifth “level” control voltage. As the fifth “level” control voltage rises, an increasing portion of the current flowing in current mirror 50 to current mirror 54 is passed on to the single-mirror transistor 55 that influences the outgoing mirror currents of transistors 47 and 49. If the “level” control voltage is zero at the sensitivity limit, no current will be conducted into current mirror 54 and comparator 4 will possesses the operating threshold and hysteresis defined by the currents flowing in transistors 46 and 48.
There is an upper switching threshold “UTHO” and a lower switching threshold “UTHU” resulting from the operating threshold and hysteresis. Reference voltage “VREF” in
Transistor 104 operates as a single mirror transistor of a current mirror for the current alternatively selectable via elements 42 and 45. Outgoing mirror transistor 105 is used to generate the current 32 for the comparator subcircuit 41.
The output signal of the DC-coupled buffer amplifier 6 in the signal path of
The input part of the circuit assembly comprises a CMOS-HF amplifier 59 with the two coupling capacitors 60 used as a differential former by means of elements 57 and 58. As the input elements exhibit high-pass behavior for the offset separation of the buffer amplifier 6, the operating point at the input of the amplifier 59 is formed by means of the elements 61, 62, and 63 and directed to the junction summing point via resistors 64.
The voltage source 65, which may in practice be obtained, for example, from the stabilized voltage “VDDa” by means of a voltage divider, raises the output operation point of the amplifier 59 by a small amount and thus ensures high-frequency suitability of the assembly up to the lower control limit by already putting the amplifier 59 into an active range. In the practical application of the invention, the value of the voltage source 65 is 50 mV, for example, in order to achieve a compromise between HF-suitability on the one hand and acceptable restriction of the level-controllability on the other hand.
The purpose of this input circuit of the amplifying control assembly 8 is to form a single-ended signal out of the input differential signal that is grounded and transmits only the positive half-wave of the input differential signal:
Voutp (59)=|k*(V(INP)−V(INN))|+V(65)
wherein k=R(58)/R(57) balances the damping factor of the limiter 1/k.
This is the only measure for the light intensity at the input or for the current supplied to the overall circuit of
Thus the following applies to the charge Q on capacitor 66 after each pulse:
Q(66)=0.
Transistor 67 opens precisely at the point when the emitter voltage of 67 falls below the “zero voltage limit” due to the basic relation to the bipolar flux voltage from transistor 62. Thus, the solution according to the invention ensures that despite offset separation the pulses are transmitted with the full amplitude to the subsequent sample-and-hold circuit.
The following rail-to-rail operation amplifier 68 is fully adjustable between “VSS” and “VDDa” at the input and the output. It implements a sample-and-hold-type function together with at least the current source 69 and the low-pass capacitor 70 of the low pass comprising resistor 71 and capacitor 70. Thus a first “AGCSH” control voltage is available at the output of the low pass 70, 71, and this voltage corresponds on average to the peak value of the pulses provided by the operation amplifier 59.
The negators 107 and 108 adjust the polarity of the corresponding control signals.
In the application according to the invention of the circuit in
When the first “AGCSH” control voltage rises, it causes an amplitude reduction of the output signal at the buffer amplifier 6 in the control loop as seen in
The “AGCSH” signal of the assembly is the central control signal for the amplitude control and is supplied at the “AGCSH” output of the assembly. While the third “limiter” control voltage exerts regulating as well as controlling influences on the behavior of the assembly according to the invention, all other output signals of the amplifying control assembly 8 have a control signal character. According to the invention, all other control voltages are derived from the first “AGCSH” control voltage.
The fifth “level” control voltage used for the adaptive control of the comparator switching thresholds results from the further smoothing of the “AGCSH” voltage by means of a low pass consisting of a resistor 80 and a capacitor 81 and of a voltage amplification by a factor>1 in the operation amplifier 84 implemented by the resistors 82 and the capacitor 83. The amplification must be >1 so that after a data transmission, when the “AGCSH” control voltage is already decreasing, the comparator thresholds can be kept at the value fixed for the transmission, thus increasing the interference resistance by means of capacitor 70. Here, deliberate use is made of the upper control limit of the operation amplifier 84. In the practical implementation of the invention, a voltage amplification value of 2 is appropriate. Accordingly, with “AGCSH” control voltage values above half the operating voltage “VDDa,” the feedback of the comparator thresholds starts only when the “AGCSH” control voltage has already decreased to 50%.
At the same time, the dominant pole of the operation amplifier 84 according to prior art and the capacitor 83 contained in the application are used to further smooth the fifth “level” control voltage. This is necessary to keep the comparator thresholds and thus the pulse position at a stable level with regard to the sampling by the comparator 4 once the signal intensity has been determined. This means a reduction of the jitter with respect to the falling flank of the output signal “OUTN” of comparator 4 in
According to the invention, the third “limiter” control voltage is generated to reduce the amplification of buffer amplifier 6 as a function of the amplitude in the case of high input intensity. That is, a peak value limitation and a reduction of the static amplification of buffer amplifier 6 are carried out. At the same time, this reduces offset amplification for high signal intensities and thus advantageously influences the behavior of the circuit. The input signal of the generation circuit for the third “limiter” control voltage consisting of a voltage source 85 and the transistors 86 and 87 is taken from a low pass of the second order consisting of the elements 71, 80, 70, and 81 and is supplied to the interconnected gates of the p-channel transistor 87 and the n-channel transistor 86. A direct voltage source 85 is connected to the source contact of the transistor 86.
This voltage source 85 determines the level of the output differential signal of the buffer amplifier 6 above which a limitation takes place. The value of the voltage source 85 depends on the operating point of the outputs “OP” and “ON” of the buffer amplifier 6 and on the threshold voltage of the p-channel transistors 25 that are used for limitation. It should be noted for dimensioning the voltage source 85 that the p-channel threshold voltage increases by contacting its bulk contacts to “VDDs” due to the substrate control factor of the transistors 25. In the practical implementation of the invention, the voltage of the voltage source 85 can be generated in various ways.
As a constant voltage from the stabilized supply voltage “VDDa” using an ohmic voltage divider or adaptively using the circuit assembly in
In an implementation according to
The statically unloadable output signal “CM” shown in
Conveniently, the two resistors 86 and 87 are dimensioned in such a way that cross-fading and thus the activation of the limiter circuit starts at the center of the level control range defined by “VDDa.” This threshold is necessary to ensure reliable blocking of the limiter transistors 25 in the case of low photocurrents and thus full amplification of the buffer amplifier 6.
According to the invention, the fourth “short” control voltage serves as pole reduction by bridging the capacitors 27 shown in
A p-channel switch transistor 94 is paralleled with the load element 93, allowing feedback of the high-pass circuit 7 into the offset separation mode following the end of a data transmission by quickly blocking the transistors 29 that serve for bridging the high-pass circuit 7 in the case of a fast-rising fourth “short” control voltage. This process is triggered by the signal management according to the invention using the “time” signal that is generated in the interlocking logic 9 as seen in
Furthermore, according to the invention, the “time” signal, gated with the “intense” signal in the NAND gate 95, is used to switch on the current source 78 shown in
The circuit shown in
The digital differentiation 97 seen in
This assembly causes a differentiation of the positive comparator signal “comp” obtained, for example, from the output “OUTP” of comparator 4 and provides for equal treatment of pulses of different widths generated at the comparator output.
The time comparison with the predetermined time period starts anew at the beginning of each light pulse. The time period is dimensioned, using the current of the current source 98 or the capacitor 99 shown in
Time period>22 μs and time period<tlatency time−tzero.
In this regard, “tlatency time” is a product characteristic to be defined and which the IrDA standard divides into various classes. According to the classification, the target minimum value for the latency time should be <50 μs in the best class for an effective protocol transmission. The maximum time for discharging the “AGCSH” capacitor 70 from any given maximum value, i.e., also the one defined by “VDDa,” down to a lower threshold value indicated by a high-low flank of the “intense” signal of the end control assembly 10 shown in
The time period thus defined causes each pulse in a transfer of 9.6 kBit/s, including pulses with a pulse duration of 22 μs, to be detected as single pulses and the signal management of the end processing according to the invention to become effective after each of these pulses. With pulses that long, there is the great danger that, in the case of a high energy content of the pulses, a static offset of the entire amplifier assembly located in the signal path causes an overshoot at the input of the comparator 4 and thus probably a detrimental erroneous pulse. In the faster mode, the end processing management is in each case activated only during longer transmission pauses. This is particularly advantageous for modes such as FIR (4 Mbit/s) and VFIR (16 Mbit/s), so that as many control voltages as possible are essentially constant during the data transmission and thus the jitter of the comparator's output pulses is kept to a minimum. For any changes in the amplifier properties that are affected by these control voltages also result in a change in the time position of the voltage pulse trains at the amplifier output.
The “lock” signal derived from the inverted output of the static master-slave flip-flop 100 shown in
The time difference between the high-low flanks of the two “time” and “lock” output signals of the assembly shown in
The assembly of the end control assembly 10 shown in
The “intense” signal ensures that the fast discharge of the “AGCSH” capacitor 70 triggered by the timer during the end processing is performed only down to a minimum value defined by the “VREF” reference voltage as seen in
During the fast discharge process of capacitor 70, i.e., when the end detection has been released by the “lock” signal, a low level is produced at the “zero” signal. This low level opens the p-channel transistor 30 in the high-pass circuit 7 shown in
The described control voltages are generated in a process where, following the detection of a packet end, the “lock” signal is interpreted as a pulse at the flip-flop 101 which then supplies a low level at the “zero” output. The completion of the fast discharge phase of capacitor 70 and the clamping of the comparator input is indicated as a low level at the “intense” output. The latter resets the flip-flop 101 and regenerates a high level at the “zero” output.
The “intense” signal is formed using the comparator 102 by comparing the current value of the “AGCSH” control voltage at the “AGCSH” contact of the comparator 4 to the “VREF” reference voltage provided by the voltage divider 103 from the stabilized “VDDa” operating voltage.
FIGS. 20 to 22 show typical signal paths of the signals that are important for the solution according to the invention. While
The effect of the signal management, according to the invention, during the end processing is shown in
| Number | Date | Country | Kind |
|---|---|---|---|
| DE 103 49 130.9 | Oct 2003 | DE | national |
| 10 2004 009 037.8 | Feb 2004 | DE | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | PCT/DE04/01383 | Jul 2004 | US |
| Child | 11405023 | Apr 2006 | US |
