Patent Application
20190068146
The present invention relates in general to a low noise amplifier circuit. The present invention is particularly related to a semiconductor circuit comprising such an amplifier circuit, and to a sensor device comprising such an amplifier circuit.
A large number of amplifier circuits are known in the art. Amplifier circuits are used for many different applications with many different requirements, for example in terms of electrical requirements/constraints: input level, output level, power consumption, heat dissipation, power efficiency, signal bandwidth, signal distortion, noise level, group delay, but also in terms of component cost, PCB size, etc.
Typically several requirements may conflict each other, and thus typically a trade-off needs to be made. There is no single ideal circuit topology which is best for all applications, but each type or class of amplifier has its own advantages and disadvantages.
The present invention is related to a new class of amplifier circuits with a high level of integration (in a semiconductor device), which is suitable for amplifying or pre-amplifying relatively small signals (e.g. having an amplitude less than 1 mV or even less than 100 μV) and provides good noise characteristics by incorporating chopper modulation (to avoid increased noise levels due to offset and flicker noise).
It is well known in the art, for example from C. Enz and G. C. Temes; “Circuit techniques for reducing the effects of op-amp imperfections: autozeroing, correlated double sampling, and chopper stabilization,” Proc. IEEE, Vol. 84, No. 11, pp. 1584-1614, November 1996, further referred to herein as [Enz96], that chopping can be used to eliminating offset and flicker noise.
There is always room for improvements or alternatives.
It is an object of embodiments of the present invention to provide an amplifier circuit or amplifier circuit topology that can be sized for low-noise amplification of an analog signal.
It is an object of particular embodiments of the present invention to provide an integrated sensor device comprising such an amplifier circuit.
It is an advantage of circuits and sensor devices according to the present invention that they provide an inherent noise-filtering characteristic which make them suitable for directly sampling without significant deterioration of the overall noise performance because of noise-folding.
It is an advantage of at least some circuits and sensor devices according to embodiments of the present invention that they may be used for amplifying various signals, including, but not limited to, biomedical signals and signals from a transducer element (e.g. in sensor applications).
It is an advantage of at least some circuits and sensor devices according to embodiments of the present invention that they are particularly suitable for co-integration with sampling-based circuits that post-process an/the amplified signal, such as e.g. with switched-capacitor circuits and/or with analog-to-digital conversion circuits.
These objectives are accomplished by embodiments of the present invention.
In a first aspect, the present invention provides a semiconductor amplifier circuit, comprising: an input block adapted for receiving a voltage signal to be amplified, an integrator circuit arranged downstream of the input block and adapted for receiving the voltage signal or a processed version thereof, the integrator circuit comprising an integrating capacitor for storing an integrated signal and being adapted for providing a continuous-time signal representative for the combined integral of signals received by the integrator circuit wherein one of the signals is the voltage signal or a processed version thereof; the semiconductor amplifier circuit moreover comprising a first feedback path comprising a sample-and-hold block and a first feedback block, the sample-and-hold block being arranged downstream of the integrator circuit and being adapted for receiving the continuous-time signal and for sampling said signal using a sampling signal thereby providing a sampled signal, and for providing a zero-order-hold voltage signal proportional to said sampled signal; the first feedback block being arranged downstream of the sample-and-hold block, and being adapted for providing a first feedback signal proportional to the zero-order-hold signal (e.g. according to a predefined proportionality factor), the first feedback signal being provided to the integrator circuit or to a component or to a node upstream of the integrator circuit; and wherein the components of the circuit are chosen such that an amplitude ratio of the continuous-time signal and the voltage signal to be amplified is at least larger than 1 for a predefined frequency range (e.g. for frequencies of the voltage signal to be amplified from DC up to a frequency representative for a bandwidth of the semiconductor amplifier circuit); Only within the predefined frequency range (i.e. within the useful bandwidth of the circuit) the amplification should be larger than one; and wherein the first feedback block is adapted for providing the feedback signal such that, when the voltage signal to be amplifier is equal to zero, a charge stored on the integrating capacitor at the beginning of a sample period is linearly removed during the sampling period of the sampling signal, in such a way that the absolute value of the charge is smaller at the end of the sampling period than at the beginning of the sampling period. In embodiments of the present invention the absolute value of the charge may for example be reduced by more than 40%, or even by more than 60%, or even by more than 90%, at the end of the sample period compared to at the beginning of the sample period. In some embodiments of the present invention the charge on the integrator capacitor may be completely removed at the end of the (single) sampling period.
In embodiments of the present invention the semiconductor amplifier circuit may comprise at least one chopper adapted for providing a chopped input signal indicative of the voltage signal to be amplified prior to the integrator circuit wherein the sampling frequency of the sample-and-hold block is equal to the chopping frequency or to an integer times the chopping frequency of the at least one chopper.
It is an advantage of embodiments of the present invention that the voltage signal to be amplified can be separated (in frequency) from the offset and flicker noise of the at least one transconductor or amplifier (part of the integrator circuit). This is accomplished through the use of chopping. The use of chopping is not mandatory for tackling offset and flicker noise, since these can also be reduced by appropriate design of the transconductor or amplifier, e.g. using devices which are sufficiently large to reduce mismatch and flicker noise.
Many exemplary embodiments of the present invention have been detailed with the inclusion of chopping. These embodiments may also be operate at a zero chopping frequency, while retaining the above mentioned advantages, except that the offset and flicker noise of the at least one transconductor or amplifier is not separated from the voltage signal to be amplified. A zero chopping frequency implies that the chopper signal is constant (e.g. either +1 or −1). The choppers in these exemplary embodiments are then all in a fixed state, and a simplified embodiment may be obtained by replacing these choppers with their corresponding hard-wired interconnection.
In embodiments of the present invention, the predefined frequency range may for example be a lowpass range or a bandpass range with some specific bandwidth. The bandwidth of the amplifier may be as low as 1 kHz, or even lower, and as high as 1MHz, or even as high as 100 MHz, or even higher. The predefined frequency range may start from DC, or at a frequency smaller than 1% of the bandwidth, or at a frequency smaller than 10% of the bandwidth or at a frequency smaller than 50% of the bandwidth.
In embodiments of the present invention the choppers are a first chopper and a second chopper, and the semiconductor amplifier circuit comprises a first transconductance.
The first chopper is adapted for providing a chopped input voltage indicative of the voltage signal to be amplified, wherein the first transconductance is arranged downstream of the first chopper and adapted for receiving said chopped voltage signal and adapted for converting said chopped voltage signal into a chopped current signal; wherein the second chopper is arranged downstream of the first transconductance adapted for providing a demodulated current signal; and wherein the integrator circuit is a current integrator arranged downstream of the second chopper.
In embodiments of the present invention each of the first and second chopper may be adapted to be operated at a chopping frequency, and the sample-and-hold-block may be adapted to be operated at a sampling frequency, wherein the sampling frequency is equal to the chopping frequency or equal to an integer times the chopping frequency, the integer being larger than zero.
It is an advantage of embodiments of the present invention that low-noise signal amplification is obtained. It is an advantage that the circuit of the present invention is at the same time an amplifier and also has a highly desirable filter effect. It is an advantage that an output of the circuit can be directly sampled without significant amount of excess noise due to noise folding (allowing sampling-based circuits to be directly connected to its output).
Although, at first sight, several “sub-circuits” and techniques used in this integrated circuit are known per se in the art, the specific combination and arrangement of elements as specified in the claim provides an amplifier circuit with a unique behavior in terms of “low-noise amplification”, in particular related to broadband white noise. As far as is known to the inventors—the advantageous behavior of this circuit is unprecedented in the prior art, and it is the best possible behavior of its kind known thus far (as compared to any other amplifier circuit that includes chopping and has an overall first-order response).
It is an advantage that no dedicated analog filter is required for filtering, and/or that no additional current consumption is required for active filtering, and/or that no tuning of filter characteristics is required (due to large tolerances).
The inventors are of the strong opinion that the specific combination of features is not trivial, inter alia because it is very uncommon to use a discrete-time feedback-loop in an otherwise continuous-time circuit, and because it is counter-intuitive to reduce noise by feedback of a sampled signal because of a widespread belief that sampling always increases noise levels due to noise folding.
The main benefit of the circuit in terms of noise performance can be best appreciated from the spectral graphs illustrated in
It is an advantage of embodiments according to the present invention that a (relatively) large signal bandwidth (e.g. sensor bandwidth) and a (relatively) low latency can be obtained. The latency can be as small as only one or only two sampling periods Ts (where Ts=1/fs). The continuous-time output (Vct) has a latency smaller than one sampling period.
It is an advantage of embodiments of the present invention that the low-noise amplifier circuit works in continuous amplitude. This offers the advantage for applications that do not benefit from conversion to the digital domain, that no ADC and/or DAC or sigma-delta-modulator are needed, thus saving power and silicon area. It is a further advantage that no amplitude-quantization noise is introduced in the feedback-loop, resulting in a low overall noise behavior. Furthermore, the bandwidth of systems with an ADC, even with a sigma-delta ADC is typically severely limited by the ADC, for example because of an oversampling factor (typically at least a factor 8). As there is no DAC in the feedback path, the circuit of the present invention can exploit (almost) the full bandwidth.
It is an advantage of embodiments of the present invention that the circuit can optionally also provide a discrete-time voltage output signal.
It is a further advantage of using an integrator (in the forward path), comprising one or more active components (e.g. transistors or an OP-AMP or an OTA) that it provides a low-impedance output, which can be directly sampled e.g. for back-end processing, without requiring extra buffering.
It is an advantage that the overall noise-behavior of the entire circuit is mainly determined by the noise-behavior of the first transconductor, (also referred to herein as “low noise transconductance”) and is much less sensitive to the noise-behavior of other components such as the current integrator or the feedback-transconductor.
It is an advantage that this circuit can be used to amplify DC-signals.
It is an advantage that one or more of the continuous-time continuous-amplitude analog signal (Vct), and/or the sampled signal (Vdt), and/or the zero-order-hold signal (Vzoh) can be provided as output signal(s).
It is an advantage that the ratio of the integrated signal and the voltage signal to be amplified, also known as the “overall gain” of the circuit, can be chosen to be a specific predefined value, for example equal to 2.0 or 5.0 or 10.0 or 100.0 or 1000.0.
It is an advantage of embodiments of the present invention that no active or passive analog low-pass filter such as e.g. an RC-filter or a Butterworth filter or Elliptic filter or Chebychev filter is required, but that the circuit itself provides a low-pass filter characteristic, while reducing, e.g. minimizing excess noise due to noise-folding.
It is an advantage of semiconductor amplifier circuits according to the present invention that they can be sized (during design-time) such that they are capable of amplifying a signal (e.g. a voltage signal) with a specific gain factor, while adding less noise due to noise-folding than a prior art amplifier circuit implemented in the same technology and offering the same gain and having an overall first-order response with the same (3 dB)-bandwidth.
It is an advantage of semiconductor amplifier circuits according to embodiments of the present invention that they may be used as an instrumentation amplifier, e.g. for providing a high input impedance, a large common-mode rejection ratio, and a definite gain.
In an embodiment, the first feedback path is arranged in one of the following ways: i) wherein the first feedback block comprises or is a second transconductance adapted for providing a current feedback signal which is fed back between an output of the second chopper and an input of the current integrator; ii) wherein the feedback block comprises or consists of a scaler for providing a voltage feedback signal, and the voltage feedback signal is fed back upstream of the first chopper; iii) wherein the feedback block comprises a second transconductance for providing a current feedback signal proportional to the zero-order-hold signal, and the current feedback signal is fed back upstream of the first chopper; iv) wherein the feedback block comprises a third chopper and second transconductance connected in series, for providing a chopped current feedback signal, the third chopper operable at the chopping frequency, and wherein the chopped current feedback signal is fed back between an output of the first transconductance and an input of the second chopper; v) wherein the feedback block comprises a third chopper and a second transconductance connected in series, for providing a chopped current feedback signal, the third chopper operable at the chopping frequency, and wherein the chopped current feedback signal is fed back to an internal node of the first transconductance; vi) wherein the feedback block comprises a third chopper and a scaler connected in series for providing a chopped voltage feedback signal, the third chopper operable at the chopping frequency, and the chopped voltage feedback signal is fed back between an output of the first chopper and an input of the first transconductance.
It is an advantage that the semiconductor amplifier circuit can be used with a wide variety of different input blocks.
In an embodiment, the sampling frequency (of the sample-and-hold block in the first feedback path) is twice the chopping frequency.
It is an advantage of circuits of the present invention that the useful bandwidth (defined by the 3 dB cutoff frequency) of the signal to be amplified, is about 88% of the Nyquist frequency, while introducing only about 30% of in-band power noise due to aliasing (of white noise).
It is an advantage of particular embodiments of the present invention wherein the sampling frequency is chosen equal to twice the chopping frequency (thus fs=2·fc) that the ratio fc/fBW can be as low as 1.14, i.e. the chopping frequency fc only has to be 14% higher than the bandwidth fBW, which is particularly favorably in applications requiring a high bandwidth fBW but having practical constraints on the maximum possible chopper frequency.
In an embodiment, the sampling frequency (of the sample-and-hold block in the first feedback path) is equal to the chopping frequency.
In an embodiment wherein the sampling frequency (of the sample-and-hold block in the first feedback path) is twice the chopping frequency, the semiconductor amplifier circuit further comprises, if not already comprising, a third chopper arranged downstream of the current integrator and operable at the chopping frequency, and further comprising: a second feedback path for removing DC-offset and flicker noise, the second feedback path being adapted for providing a second feedback signal to the current integrator or to a component or to a node upstream of the current integrator, the second feedback path comprising: the third chopper, and a filter arranged downstream of the third chopper, the filter having a transfer function comprising at least one integration and a factor (1+Z−1).
In an embodiment wherein the sampling frequency (of the sample-and-hold block in the first feedback path) is an integer N times the chopping frequency, the semiconductor amplifier circuit further comprises a second feedback path for removing DC-offset, the second feedback path being adapted for providing a second feedback signal to the current integrator or to a component or to a node upstream of the current integrator, the second feedback path comprising: a second sampler operable at a second sample frequency, and a fourth chopper arranged downstream of the second sampler, the fourth chopper operable at the chopping frequency, and a filter arranged downstream of the fourth chopper, the filter having a transfer function comprising at least one integration and a factor (1+Z−1), and wherein the second sampling frequency is equal to twice the chopping frequency.
In both embodiments with a second feedback loop (as shown in
It is an advantage that the filter H3 comprises an integrator, which together with the choppers in the second feedback path provides a notch at the chopping frequency fc.
It is an advantage that the filter H3 comprises a factor (1+Z−1) to block signal-related frequencies (which are up-converted by the fourth chopper CH4 to the chopping frequency fc) while amplifying offset-related frequencies, thus making the second feedback loop more selective.
It is a further advantage of the circuit with offset-suppressing feedback loop that it reduces DC-offset, and flicker noise. Such a circuit is ideally suited for readout of sensor-signals having also a DC-offset, such as for example Hall-sensors, while reducing or eliminating the risk of clipping the amplified signal to one of the power rails (VDD or GND).
It is a major advantage of embodiments comprising a second feedback path, in particular if fs and fc are chosen such that fs=1·fc or fs=2·fc, that the transfer function in the pass-band can be improved (substantially flat to a higher frequency), and that the transfer function has a very steep fall near the end of the pass-band.
In a second aspect, the present invention also relates to a semiconductor amplifier circuit, comprising: an input block comprising a first chopper adapted for providing a chopped input voltage signal indicative of a voltage signal to be amplified; at least one amplifier having an inverting input port and a non-inverting input port and an output port; and wherein the output port of the first chopper is connected to the non-inverting input port of the at least one amplifier, and/or the output port of the first chopper is connected to the inverting input port of the at least one amplifier via at least one intermediate component adapted for converting the chopped input voltage signal into a chopped current signal; a second chopper having an input port and an output port, the input port being connected to the output port of the at least one amplifier; a third chopper having an input port and an output port, the output port being connected to the inverting input port of the at least one amplifier; at least one integration capacitor connected between the output port of the second chopper and the input port of the third chopper; a sample and hold circuit having an input port connected to the output port of the at least one amplifier or to the output port of the second chopper, directly and/or indirectly through one or more intermediate components, and having an output port for providing a zero order hold voltage signal; a proportional feedback network having an input port connected directly and/or indirectly through one or more intermediate components to the output port of the sample and hold circuit, and being adapted for providing a proportional feedback signal, and having an output port connected directly and/or indirectly through one or more intermediate components to the inverting input port of the at least one amplifier; wherein a voltage of the at least one integration capacitor provides a continuous-time voltage signal being an amplified version of the voltage signal to be amplified; and wherein the values of the proportional feedback network are chosen such that an amplitude ratio of the continuous-time signal and the voltage signal to be amplified is larger than 1.0 for a predefined frequency range; and wherein the value of the at least one integration capacitor is chosen such that, for a chopped input signal equal to zero, a charge stored on the integrating capacitor at the beginning of a sample period is linearly removed during the sampling period of the sampling signal, in such a way that the absolute value of the charge is smaller at the end of the sampling period than at the beginning of the sampling period; and wherein each chopper is adapted to be operated at a chopping frequency, and the sample-and-hold-block is adapted to be operated at a sampling frequency, wherein the sampling frequency (fs) is equal to the chopping frequency (fc) or to an integer multiple (N) larger than zero thereof. In embodiments of the present invention the non-inverting input port of the at least one amplifier may be connected to a biasing voltage. Embodiments according to the second aspect describe a slightly different topology than the circuits described in the first aspect, but offer the same advantages, and are illustrated for example in
In another aspect, the present invention also relates to a semiconductor amplifier circuit, comprising: an input block adapted for receiving a voltage signal to be amplified; at least one amplifier having an input port which comprises at least one input node and an output port which comprises at least one output node; at least one integration capacitor connected in a negative feedback configuration between the at least one output node and the at least one input node of the at least one amplifier. In such an embodiment the semiconductor amplifier circuit is configured such that, during operation, the voltage signal to be amplified directly or indirectly (i.e. through one or more intermediate components) arrives at the input port of the at least one amplifier. Embodiments of the present invention may comprise a connection for passing the voltage signal to be amplified to an amplifier input node which is also connected to the at least one integration capacitor. This connection may be an indirect connection comprising one or more intermediate components adapted for converting the voltage signal to be amplified into a current signal. Embodiments of the present invention comprises a sample and hold circuit having an input port connected to the output port of the at least one amplifier, directly and/or indirectly through one or more intermediate components, and having an output port for providing a zero order hold voltage signal; a proportional feedback network having an input port connected directly and/or indirectly through one or more intermediate components to the output port of the sample and hold circuit, and being adapted for providing a proportional feedback signal, and having an output port connected directly and/or indirectly (i.e. through one or more intermediate components) to the input port of the at least one amplifier; wherein a voltage of the at least one integration capacitor provides a continuous-time voltage signal being an amplified version of the voltage signal to be amplified; and wherein the values of the proportional feedback network are chosen such that an amplitude ratio of the continuous-time signal and the voltage signal to be amplified is larger than 1.0 for a predefined frequency range (e.g. for frequencies of the voltage signal to be amplified from DC up to a frequency representative for the bandwidth of the semiconductor amplifier circuit); and wherein the value of the at least one integration capacitor is chosen such that, for an input signal equal to zero, a charge stored on the integrating capacitor at the beginning of a sample period is linearly removed during the sampling period of the sampling signal, in such a way that the absolute value of the charge is smaller at the end of the sampling period than at the beginning of the sampling period.
Many variants are possible. For example, the input signal may be a single-ended or a differential input signal. In case of a differential input signal, several components should be present in pairs, or should have dual input ports and output ports. The proportional feedback network may comprise or may be for example a resistive network or a transconductance, etc. Two specific embodiments are described next.
In an embodiment of a semiconductor amplifier circuit according to the second aspect, the input block is adapted for providing a differential chopped input voltage signal indicative of a differential voltage signal to be amplified; the at least one amplifier comprises: a first and a second Operational Transconductance Amplifier (OTA), the first OTA having a non-inverting input port connected to a first output of the first chopper and being adapted for providing a first output signal, the second OTA2 having a non-inverting input port connected to a second output of the first chopper different from the first output and being adapted for providing a second output signal; the second chopper being adapted for receiving the first respectively second output signal from the first respectively second OTA, the second chopper having a first output connected to a first node N1 and a second output connected to a second node N2; and wherein the at least one integrating capacitor comprises a first and a second integrating capacitor, the first integrating capacitor being arranged between the first node N1 and a third node N3, and the second integrating capacitor being arranged between the second node N2 and a fourth node N4; the third chopper having a first input connected to the third node N3 and a second input connected to the fourth node N4 and having a first output connected to the inverting input of the first OTA and having a second output connected to the inverting input of the second OTA; and wherein the resistive feedback network comprises: a first feedback resistor arranged between the third node N3 and a fifth node N5, and a second feedback resistor arranged between the fourth node N4 and a sixth node N6, and a third resistor arranged between the third node N3 and the fourth node N4; and wherein the sample and hold circuit has a first input connected to the first node N1 and a second input connected to the second node N2 and a first output connected to the fifth node N5 and a second output connected to the sixth node N6.
This specific embodiment is illustrated in
In another embodiment of a semiconductor amplifier circuit according to the second aspect, the input block is adapted for receiving a differential voltage input signal to be amplified and for chopping said signal at a chopper frequency, a first output of the first chopper being connected to a base or gate of a first transistor or darlington-pair or the like, a second output of the first chopper being connected to a base or gate of a second transistor or darlington-pair or the like; and wherein the circuit further comprises a first and a second transistor or darlington-pair or the like, the first transistor or darlington-pair or the like having an emitter or a source connected through a first current source to ground, and having a collector or a drain connected through a second current source to a supply voltage, the second transistor or darlington-pair or the like having a source connected through a third current source to ground, and having a collector or drain connected through a fourth current source to a supply voltage; and wherein the at least one amplifier comprises: a first Operational Amplifier and a second Operational Amplifier, a non-inverting input of the first Operational Amplifier being connected to a biasing voltage and the non-inverting input of the second Operational Amplifier being connected to the biasing voltage; the sample-and-hold block having a first input connected to an output of the first Operational Amplifier, and having a second input connected to an output of the second Operational Amplifier, and having a first and second output connected via a resistor network to the emitter or source of the first transistor or darlington-pair or the like and to the emitter or source of the second transistor or darlington-pair or the like for providing a differential feedback signal; the second chopper having a first respectively second input connected to a first respectively second output of the first respectively second Operational Amplifier, and having a first respectively second output connected to a first respectively second output node, a potential difference between these nodes forming a differential output voltage as a continuous-time voltage signal being an amplified version of the differential input voltage; and wherein the at least one integrating capacitor comprises a first and a second integrating capacitor, the first integrating capacitor having one terminal connected to the first output node and another terminal connected to a first input of a third chopper, the second integrating capacitor having with one terminal connected to the second output node and another terminal connected to a second input of a third chopper; the third chopper having a first output connected to the collector or drain of the first transistor or darlington-pair or the like and connected to the inverting input of the first Operational Amplifier, and having a second output connected to the collector or drain of the second transistor or darlington-pair or the like and connected to the inverting input of the second Operational Amplifier.
This specific embodiment is illustrated in
In an embodiment according to the first or second aspect, the input block is further adapted for receiving the voltage signal to be amplified from one or two external pins or from a voltage source inside the integrated circuit; or wherein the input block further comprises an impedance (e.g. resistance) or transimpedance (e.g. Wheatstone-bridge) arranged upstream or downstream of the first chopper and is further adapted for receiving a continuous-time current signal from an external pin or from a current source inside the integrated circuit; or wherein the input block further comprises a transducer of the kind that converts a physical (e.g. non-electrical)signal or excitation into a voltage signal, the transducer being arranged upstream of the first chopper for providing the voltage signal to be amplified in response to the physical (e.g. non-electrical)signal or excitation; or wherein the input block further comprises a transducer of the kind that requires a biasing signal and provides a voltage signal indicative of a physical quantity, the transducer being arranged upstream of the first chopper for providing the voltage signal to be amplified in response to the physical quantity; or wherein the input block further comprises a transducer of the kind that requires a biasing signal and provides a voltage signal indicative of a physical quantity, the first chopper.
It is an advantage that the amplifier circuit according to embodiments of the present invention is very much suited for amplifying a signal from any of the above mentioned signal sources. In an embodiment according to the first or second aspect, at least all components downstream of the first chopper, including those of the first and/or second feedback path, are integrated on a single semiconductor die.
In a third aspect, the present invention provides a sensor device comprising: at least one sensor element or at least one transducer or at least one impedance or transimpedance for providing at least one voltage signal to be amplified; and a semiconductor amplifier circuit according to the first or second aspect, arranged for amplifying said at least one voltage signal.
The semiconductor amplifier circuit may be integrated on a first die, and the sensor element or transducer or impedance or transimpedance may be integrated on a second die. The first die and second die may be a single die. The first and second die may be two separate dies interconnected with each other but packaged in a single chip package.
In an embodiment, the sensor devices further comprises: an analog-to-digital convertor for converting the amplified voltage signal; and a digital processor for digitally processing the digitized signal.
The at least one transducer of the sensor may consist of a single Hall element or Hall sensor, or may comprise at least one Hall element or Hall sensor.
The sensor device may be comprise or may be a position sensor device or an electronic compass.
The circuit described above is ideally suited for amplifying voltage signals having an amplitude in the order of about 100 μV to about 10 mV, with low-noise. This makes the circuit ideally suited for amplifying a Hall voltage over a Hall plate, which low noise, in such applications as position sensors, in particular angular position sensors, or compass applications.
Particular and preferred aspects of the invention are set out in the accompanying independent and dependent claims. Features from the dependent claims may be combined with features of the independent claims and with features of other dependent claims as appropriate and not merely as explicitly set out in the claims.
These and other aspects of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.
In
In
In
In
In
In
In
In
In
The ratio of the sampling frequency fs and the chopping frequency fc in all embodiments shown in
The drawings are only schematic and are non-limiting. In the drawings, the size of some of the elements may be exaggerated and not drawn on scale for illustrative purposes. Any reference signs in the claims shall not be construed as limiting the scope. In the different drawings, the same reference signs refer to the same or analogous elements.
The present invention will be described with respect to particular embodiments and with reference to certain drawings but the invention is not limited thereto but only by the claims. The drawings described are only schematic and are non-limiting. In the drawings, the size of some of the elements may be exaggerated and not drawn on scale for illustrative purposes. The dimensions and the relative dimensions do not correspond to actual reductions to practice of the invention.
Furthermore, the terms first, second and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a sequence, either temporally, spatially, in ranking or in any other manner. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other sequences than described or illustrated herein.
Moreover, the terms top, under and the like in the description and the claims are used for descriptive purposes and not necessarily for describing relative positions. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other orientations than described or illustrated herein.
It is to be noticed that the term “comprising”, used in the claims, should not be interpreted as being restricted to the means listed thereafter; it does not exclude other elements or steps. It is thus to be interpreted as specifying the presence of the stated features, integers, steps or components as referred to, but does not preclude the presence or addition of one or more other features, integers, steps or components, or groups thereof. Thus, the scope of the expression “a device comprising means A and B” should not be limited to devices consisting only of components A and B. It means that with respect to the present invention, the only relevant components of the device are A and B.
Reference throughout this specification to “one embodiment” or “an embodiment” means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, appearances of the phrases “in one embodiment” or “in an embodiment” in various places throughout this specification are not necessarily all referring to the same embodiment, but may. Furthermore, the particular features, structures or characteristics may be combined in any suitable manner, as would be apparent to one of ordinary skill in the art from this disclosure, in one or more embodiments.
Similarly it should be appreciated that in the description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. This method of disclosure, however, is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.
Furthermore, while some embodiments described herein include some but not other features included in other embodiments, combinations of features of different embodiments are meant to be within the scope of the invention, and form different embodiments, as would be understood by those in the art. For example, in the following claims, any of the claimed embodiments can be used in any combination.
In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In other instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Definitions:
Where in the present invention reference is made to a “chopper”, what is meant is a block having an input port for receiving an input signal X(t), and a control port for receiving a control signal m(t), in particular a periodic block wave having alternating values +1 and −1, and adapted for providing an output signal Y(t), which can be expressed in mathematical terms as the product: Y(t)=m(t) X(t). The output signal is of the same nature as the input signal, e.g. both are voltages, or both are currents. The input and/or output signal may be a differential signal, in which case the actual signal is defined by the difference in voltage or current at two nodes of the port. The input and/or output signal may be a single-ended signal, in which case the actual signal is defined as the voltage or current at one node representative for the input port, relative to a constant reference voltage (e.g. ground) or reference current. It is known in the art that choppers can be implemented in most cases with not more than four switches. Examples are provided e.g. in [Enz96]
Where in the present invention reference is made to a “Zero-Order-Hold” circuit (abbreviated as ZOH), what is meant is a circuit having an impulse response function f(t) which is equal to 1.0 for a predefined time interval T, and is equal to 0 elsewhere, without amplitude quantization other than by physical quantities such as for example integer numbers of photons or electrons. Typically a sample-and-hold circuit or zero-order-hold circuit is periodically sampled at a frequency fs, in which case the predefined time-interval T may be equal to the sampling period Ts=1/fs, but that is not absolutely necessary, and T may also be chosen as another fraction of the sampling period Ts, preferably at least 50% of the sampling period, for example 50% or 75% thereof. This means that an Analog-to-Digital Converter (ADC), even a 10-bit ADC or a 12-bit ADC or a 16-bit ADC does not qualify as a “zero-order-hold circuit” as defined herein.
In this document, the terms “zero-order-hold circuit” and “sample-and-hold circuit” are used as synonyms.
Where in the present invention, it is stated that block B (or component B) is arranged “downstream” of block A (or component A), what is meant is that an (or the) output of block A is connected to an (or the) input of block B, either directly (i.e. without any intermediate blocks or components), or indirectly (i.e. with only a few, e.g. only one or only two components or blocks in between, such as e.g. one summing element, or two summing elements connected in series, or a resistor divider, etc.). Stated in other words, it thus means that block A and block B are connected in series, with optionally a few (e.g. only one or only two) other blocks or components in between. As an example, the first transconductance G1 of
Where in the present invention, it is stated that block A is arranged “upstream” of block B, what is meant is that block B is arranged “downstream” of block A.
Unless explicitly mentioned otherwise, the term “feedback signal” (without prefix first or second) means “first feedback signal”, and the term “feedback path” (without prefix first or second) means “first feedback path”.
Unless explicitly mentioned otherwise, the term “sampling frequency” (without prefix first or second) means the sampling frequency fs applied to the sample-and-hold block of the first feedback path. In fact, the only embodiment shown wherein two different sampling frequencies are used is that of
The terms “feedback path” and “feedback loop” are used as synonyms.
In this document, the expression “second feedback loop” and “offset feedback loop” are used as synonyms.
Where in the present invention reference is made to a “current source”, what is meant is a component capable of providing a current to a node, substantially independent on the voltage of that node. Typically a current source has a very large output impedance, e.g. at least 1 K Ohm, or at least 10 K Ohm. (thus a supply voltage node VDD or a ground node GND does not qualify as a “current source” because these nodes typically have a very small output impedance).
The term “low noise” is used in its well-recognized meaning in the field of amplifier circuits, more specifically in the fields of amplifier circuits implemented in semiconductor devices, such as e.g. sensor products comprising a sensor element and an amplifier circuit.
Many different types of amplifier circuits are known in the art, and they are used for a large variety of different applications for amplifying a large variety of signals, and the requirements (e.g. in terms of bandwidth, gain, noise, input level/power, output level/power, power efficiency, group delay, etc.) of the amplifier circuits can vary largely for each application. For example, for high quality audio signals, the useful bandwidth of the signal is typically 20 Hz to 20 kHz, but the input and output levels (or power) vary considerably between an amplifier in an hearing aids, an amplifier for a home-cinema system, and an amplifier for loudspeakers at a rock concert. Wireless communication systems (e.g. radio, satellite) typically have to amplify very small signals at relatively high bandwidth, but it will be clear that the requirements for deep space communication are much more severe.
The present invention is related to the field of amplification of low-level signals such as can be obtained for example from sensors, such as for example a single-ended or differential voltage signal provided by a thermocouple or a thermopile (indicative of a temperature to be measured, which may in turn be indicative for an amount of IR radiation received), or a differential voltage obtained from a Hall sensor (e.g. indicative of a magnetic field to be measured), or a differential voltage provided by a Wheatstone bridge (e.g. indicative of an absolute pressure or a pressure difference to be measured), etc.
In many prior art applications, additional post-processing is done on the amplifier output signals. This may include filtering (e.g. for suppression of upconverted offset and flicker noise), additional amplification, analog-to-digital conversion, etc. Analog post-processing blocks can be operating in continuous-time or in discrete-time. However, circuits operating in discrete time are sampling-based, and hence noise folding may occur. Sampling also occurs in any types of analog-to-digital conversion. In general, modern electronic systems tend to digitize information as quickly as possible, migrating additional functionality and signal processing tasks as much as possible to the digital domain.
The present invention is primarily aimed at “low-noise” amplification. It is known in the art that the noise performance of an amplifier depends on the current that the amplifier is allowed to consume. Low-noise amplifiers are specifically designed such that for a given maximum supply current, the (input-referred) amplifier noise is reduced, e.g. minimized. When the amount of amplifier noise is fixed, e.g. when this noise level is imposed by the application (e.g. depending on the environment conditions such as temperature), low-noise amplifier topologies reduce, e.g. minimize the needed supply current.
The inventors had the task of finding a circuit topology which can be largely, preferably completely integrated in a semiconductor device, and offers an improved noise performance, especially if no analog filter is used between an output of the amplifier and an input of an ADC following the amplifier circuit.
As a specific set of requirements, the amplifier circuit topology of the present invention should allow the circuit to be completely integrated, allow reading out a sensor element with a gain-factor of 1000, has a bandwidth (3 dB bandwidth) of at least 430 kHz, while adding only about 4 nV/√Hz of (input-referred) noise, and has an output which can be directly digitized (without requiring additional analog low-pass filter), but of course the present invention is not limited to circuits that satisfy this specific set of requirements.
an input block comprising a first chopper CH1 for providing a chopped input voltage signal Vin_ch indicative of a voltage signal to be amplified;
a first transconductance G1 arranged downstream of the first chopper CH1 and adapted for receiving a chopped voltage signal Vin_ch and adapted for converting said chopped voltage signal into a chopped current signal I_ch;
a second chopper CH2 arranged downstream of the first transconductance G1 adapted for providing a demodulated current signal I_demod;
a current integrator CD arranged downstream of the second chopper CH2, the current integrator CI1 comprising an integrating capacitor Cint for storing an integrated signal Vint and being adapted for providing a continuous-time signal Vct;
The first chopper, the first transconductance and the second chopper are not strictly required. In general, the input block may be adapted for receiving the voltage signal to be amplified. The integrator circuit is thereby arranged downstream of the input block and adapted for receiving the voltage signal or a processed version thereof (the processed version may for example be obtained via the first transconductance, the first chopper and the second chopper, however, this is not strictly required). The integrator circuit thereby comprises an integrating capacitor for storing an integrated signal and adapted for providing a continuous-time signal representative for the integral of the voltage signal.
The circuit moreover comprises a first feedback path comprising: a sample-and-hold block S&H arranged downstream of the integrator circuit (in this example the current integrator CI1) and adapted for receiving the continuous-time signal Vct and for sampling said signal using a sampling signal having a sampling frequency fs, thereby providing a sampled signal Vdt, and for providing a zero-order-hold voltage signal Vzoh proportional to said sampled signal Vdt; a first feedback block G2 arranged downstream of the sample-and-hold block S&H, the first feedback path being adapted for providing a first feedback signal Ifb proportional to the zero-order-hold signal Vzoh according to a predefined proportionality factor, the first feedback signal Ifb being provided to the integrator circuit (in this example the current integrator CI1) or to a component or to a node upstream of integrator circuit (in this example the current integrator CI1).
The circuit is sized, i.e. the values of the components are chosen such that the amplitude ratio Vct/Vin of the continuous-time signal Vct output of the block H1 and the voltage signal to be amplified Vin is more the 1, e.g. at least 2.0.
The first feedback block G2 is adapted for providing a feedback signal Ifb such that, for an input signal Vin equal to zero (and thus in this example also the chopped input signal Vin_ch equal to zero) a charge stored on the integrating capacitor Cint at the beginning of a sample period is linearly removed during the sampling period Ts of the sampling signal, in such a way that the absolute value of the charge is smaller at the end of the sampling period that at the beginning thereof. In some embodiments of the present invention the charge may be completely removed at the end of the single sampling period Ts.
In this example each chopper is adapted to be operated at a chopping frequency fc. Although not explicitly shown in the drawings, the signal m(t) stands for “modulation signal” and it has a frequency fc referred to as the “chopping frequency fc”.
In this example the sample-and-hold-block S&H is adapted to be operated at a “sampling frequency fs”. The sampling frequency fs is equal to the chopping frequency fc, or is an integer multiple thereof, the integer being larger than zero. This can be written in mathematical terms as: fs=N·fc, where N can be chosen as 1, 2, 3, etc. In the drawings the sampling signal is sometimes denoted by “tn=n·T” (see e.g.
One or more of the signals Vct, Vdt and Vzoh may be provided as output signal, but that is not absolutely required. The signal Vct is a continuous-time signal, hence can be used when the amplifier needs to interface to continuous-time post-processing. The signal Vdt and Vzoh are time-discrete signals, hence can be used when the amplifier needs to interface to sampling-based post-processing. Since an important advantage of present invention over prior art relates to a lower level of noise-folding upon sampling the amplifier output, the noise characteristics of the time-discrete signals will be described further.
The mandatory components of the amplifier circuit, the integrator CI1, the sample-and-hold block S&H and the first feedback block may be completely integrated in a semiconductor device. In embodiments which are comprising a first chopper CH1, a first transconductance G1, and a second chopper CH2, also these may be completely integrated in the semiconductor device. Preferably these components are located on a single die.
It is pointed out that this high-level block-diagram can be implemented for amplifying a single-ended signal, or can be implemented for amplifying a differential signal. The step of converting the high-level schematic to a practical chip-implementation (at component or register-level) falls outside the scope of this document.
All choppers CH1, CH2 of the circuit are adapted to operate with a chopping signal m(t) having a chopping frequency “fc”, and the sample-and-hold block S&H is adapted to operate with the sampling signal s(t) having a sampling frequency fs. An example of a sampling signal s(t) and a chopping signal m(t) are shown in
The parts shown in dotted line (in particular filter H1, filter F2, ADC, CPU, MEM) are optional.
The block H1 is optional. If present, the block H1 may comprise additional filtering, e.g. a passive first order filter. But this block may also comprise an integrator or even an amplifier.
In order not to obscure the description, unless explicitly mentioned otherwise, the invention will be described as if block H1 is not present, in which case the output of the integrator CI1 is assumed to be connected to an input of the sample-and-hold block.
The embodiment of
The integrated circuit may further comprise a digital controller or digital processor, e.g. a programmable microcontroller, or a hardcoded state-machine, connected to one or more memories, (e.g. RAM, ROM, FLASH, NVRAM, etc.) in a manner known in the art. The digital processor may be adapted for receiving the values Vq digitized by the ADC described above, and for performing calculations in the digital domain, such as for example for calculating an angular position based on the values obtained from a plurality of Hall sensors, which may be read-out by means of one or more amplifier circuits described above (see also
As already discussed above, the circuit of
How the Circuit of
Although the behavior of each individual component of the circuit of
Indeed, the forward path (from the voltage input node up to and including the current integrator CI1) operates in continuous-time, while the behavior of the feedback path (comprising the second transconductance G2) is based on a value Vdt sampled in discrete-time. Whereas the behavior (stability, amplification, etc.) of a continuous-time system with a continuous-time feedback loop is well understood in the prior art, the behavior of mixed-time circuits is not so well understood.
It is noted that the sample-and-hold block is considered as part of the “feedback path”, and not of the “forward path”, but the exact border between “forward path” and “feedback path” of a circuit containing a feedback loop is somewhat arbitrary, and is not relevant for the invention. To avoid confusion, the terms “downstream” and “upstream” are used to describe the relative position of components in the circuit.
In order to get a better feeling of what happens in the time-domain, different parts of the circuit will first be described separately. The behavior of the circuit as a whole, will be described later (in relation to
Behavior of the Sub-Circuit Containing CH1+G1+CH2:
The behavior of the partial circuit from the first chopper CH1, the first transconductance G1, and the second chopper CH2 is presumed to be well understood. In short, the first chopper CH1 modulates the input voltage signal Vin to move the baseband signal (i.e. DC and low frequencies) to higher frequencies (around the chopping frequency fc and odd multiples of fc) where there is no significant flicker noise. The transconductance G1 converts the voltage signal into a current signal, and inevitably adds an amount of noise. The second chopper CH2 modulates the noise added by the first transconductance G1 to higher frequencies (around fc) while demodulating the current signal (around fc) back to baseband. As a result, the noise introduced by the transconductance G1 is upconverted to higher frequency bands, away from the baseband. This behavior is known in the art.
Behavior of CI1:
Assuming first that the first feedback signal Ifb is zero, and that the initial state of the integrating capacitor (part of the current integrator CI1, but not explicitly shown) is zero, the behavior of the current integrator CI1 is, as the name suggests, to integrate the demodulated current I demod, thereby converting the current I demod into a voltage Vint. For example, a constant current I_demod would result in a linear increase with time of the charge stored on the integrating capacitor, and thus in a linear increase with time of the output voltage of the current integrator CI1, as illustrated by curve 93 and 94 in
Behavior of Sample & Hold Block:
The behavior of a sample-and-hold block is well known in the art, and can for example be implemented by a switch (operated at the sampling frequency) and a capacitor for storing charge, possibly followed by a buffer for allowing readout of the capacitor voltage without removing charge from the capacitor. It is pointed out that, in embodiments of the present invention, the sample-and-hold circuit stores continuous-amplitude signals, (also known as “analog signals”), not discrete-amplitude signals (also known as “quantized signals”). Thus an ADC and a DAC connected in series do not qualify as a S&H block of the present invention. The output signal of the sample-and-hold block is referred to herein as “zero-order-hold voltage” Vzoh. The input voltage of the Sample & Hold block may be the integrator voltage Vint, the signal Vct related to Vint through an optional filter block H1, or any of these signals scaled down (e.g. by means of a voltage divider). Indeed, if the gain of the circuit is for example 1000, the integrated signal needs to be attenuated by a factor 1000 in the feedback path. This attenuation may be implemented completely in the second transconductance G2, downstream of the sample-and-hold block, or may be partly attenuated upstream of the sample-and-hold block.
Behavior of the First Feedback Path:
The inventors came to the idea of providing a first feedback signal (directly or indirectly) upstream of the integrator circuit (e.g. current integrator CI1), and to size the feedback path in such a way that an absolute value of a charge present on the integrating capacitor Cint at time t1=n·Ts will be smaller than the absolute value of the charge present on the integrating capacitor at time t2=(n+1)·Ts, i.e. after exactly one period Ts=1/fs (assuming that the input signal Vin is zero). According to embodiments of the present invention, this functionality can be realized by providing a (first) feedback signal Ifb that is proportional to the zero-order-hold signal Vzoh. While the proportional feedback path may have one or more sign reversals at various places, the net effect is to provide negative feedback (schematically indicated by the minus-sign of the summing element S1 in
During normal operation of the circuit, when the input signal Vin and thus also I_demod is not zero, the feedback signal Ifb is subtracted from the demodulated signal I_demod in the summing element S1, and the combined current I_delta will be applied to the input of the current integrator CI1. The result being that the voltage over the integrating capacitor varies as the sum of signal 91 and signal 93 from time t=(n−1)·Ts to t=n·Ts, and hence gradually varies from the value Vn−1 to Vn. Then a new sample of Vzoh is taken, and the integrating capacitor is discharged (e.g. linearly discharged if H1 is absent) according to signal 92, while being charged according to signal 94 (not necessarily linearly), together moving the voltage corresponding to the charge on the capacitor from Vn to vn+1, etc. In embodiments of the present invention the discrete feedback may be such that an arbitrary state of the integrator at time n·Ts is partly or completely erased after one period Ts, and the new state of the integrator at time (n+1)·Ts is partly or solely determined by integration of the input signal over the sampling period Ts. It is also noted that a sudden change in the input leads to an immediate change in the current I_demod, which directly changes the direction in which the integration output signal Vint evolves. The integrator output therefore has a low latency (delay), substantially less than an integration interval Ts. For some applications, such a low-latency response is advantageous.
Behavior of the Entire Circuit:
The attentive reader will probably have recognized that signals Vn−1, Vn, Vn+1 do not change sign each period, because the first chopper CH1 is modulating and the second chopper CH2 is demodulating with the same chopper signal m(t), and thus cancel each other. Hence, there is no net translation in frequency from the amplifier input voltage Vin to the integrator output voltage Vint. In contrast, the offset and flicker noise of the first transconductance G1 has been upconverted to the chopping frequency fc, and is filtered out to some extend by the integration of CI1. It takes additional effort to recognize that the integrator output Vint is an amplified version of the input signal Vin, especially because the transfer from input to output is not time-invariant (i.e. an arbitrary shift in time of the input signal Vin may lead to an output signal which is not just the same output signal shifted over the same time). In spite of this, the system behaves as an amplifier with a low-frequency gain A0, which can be expressed by the mathematical formula:
A
0=(Ts·G1)/Cint [1]
where Ts=1/fs is the sampling period at which the sample-and-hold block is operated, G1 is the transconductance of the first transconductance, and Cint is the capacity of the integrating capacitor.
The most surprising effect comes from an analysis of the circuit behavior in terms of noise. In general, sampling in the time-domain leads to increased noise due to folding of noise components beyond the Nyquist frequency into the base band. However, contrary to what is expected, the sampling-based feedback path leads to an amplifier with less noise folding as compared to an amplifier with continuous-time proportional feedback, as will be demonstrated further by means of simulations and calculations. This circuit (if properly sized) thus may offer significant improvements in terms of noise reduction, which, as far as is known to the inventors, are unprecedented in the prior art (assuming the same current consumption).
Possible Method of Sizing the Components:
Although the behavior of some blocks of the circuit of
The considerations for deciding the chopping frequency are very similar to what is the case for prior art circuits. For a given input signal, the chopping frequency fc needs to be at least larger than the bandwidth fBW of the input signal, that is the ratio fc/fBW needs to be larger than one. In the prior art, when technically possible, a large fc/fBW ratio (e.g. a factor 10 or more) is often preferred, because this helps in the suppression of upconverted offset and flicker noise. However, there are often practical upper limits to the chopping frequency, for example:
to keep residual offset due charge injection of the chopper switches sufficiently low (increases with chopper frequency fc), or to have sufficient settling in view of parasitic RC effects (e.g. when chopping a sensor). For clarifying the rest of the sizing, the chopping frequency can be considered as given.
In preferred embodiments of the present invention, the sampling frequency fs is chosen equal to N·fc, where N is an integer value larger than zero, for example N=1 or N=2. The components of the blocks (G1, CI1, G2) can then be designed for example so as to obtain a predefined gain factor A0 larger than 2.0, for example 1000, and such that the charge on the capacitor is removed after one sample period Ts=1/fs (assuming the input signal is zero). The latter “state-erasing” condition can be expressed in mathematical form as:
Ts·G
2
/Cint=1 [2]
where Ts=1/fs is the sampling period, G2 is the transconductance of the feedback block, and Cint is the capacity of the integrating capacitor. This condition may be used to deduce the size of the integration capacitor as:
Cint=Ts·G2 [3]
Combining formula [1] and [2], allows the gain to be written as the ratio of the first and second trans conductance, namely:
A
0
=G
1
/G
2 [4]
where G1 is the first transconductance, and G2 is the second transconductance.
The attentive reader will realize that there are three variables: G1, G2 and Cint but at this point there are only two requirements, as expressed by formula [3] and formula [4]. The remaining requirement comes from the targeted noise-performance of the amplifier, which is dominated by the input transconductance G1. The skilled person can choose an appropriate G1 depending on the envisioned noise behavior. When G1 is chosen too small, the (input-referred) white noise of the amplifier circuit may not be sufficiently low. When G1 is chosen too large, the circuit consumes more current than needed.
When the sizing relies on conditions [3] and [4], this sizing strategy leads to the outcome that, when the input signal is zero at the beginning of an integration interval (where it is understood that this remains the case thereafter), any charge stored on the integrating capacitor (Cint) at the beginning of a sample period (Ts) will be linearly removed during one single sampling period (Ts) of the sampling signal, in such a way that the charge is completely removed at the end of the single sampling period (Ts). This has as implication that the output of the low noise amplifier circuit is not affected by the values of the input signals at times which are more than one single sampling period (Ts) in the past. This sizing leads to a low noise amplifier circuit having a “minimal memory”. It is minimal because obviously the values of the input signal during the integration interval all should have an influence on the output, so the “memory” cannot be smaller than one sample period (Ts). While the minimal-memory feature may be beneficial for some applications, it must be stressed that this is not mandatory, and other sizings are possible (or even advisable). Before describing these other sizings, it may be remarked first that conditions [3] and [4] also relate to a nominal sizing. It is well known that in practice component parameters drift (e.g. with temperature). The drift of the DC gain A0 as given by [4] may be minimized by aiming for a good matching of the transconductors G1 and G2. However, it is obvious for a skilled person that equation [3] certainly cannot hold exactly (at all times), because it involves two different types of components (a capacitor and a transconductor).
In view of the above-mentioned elements, equation [3] is generalized to include a scale factor c:
Cint=c·Ts·G2 [3′]
In this, the scale factor c may be representative for a deviation due to drift of component parameters, and/or it may be representative for a deliberate design strategy.
Based on the theory of sampled-data feedback systems, the transfer function of the low noise amplifier for this generalized case can be determined as:
From this transfer function, it can be deduced that the amplifier is stable for c values in the range from zero to two. The previously discussed case of “minimal memory” is for c equal to one. A plot of the amplitude transfer for various cases is plotted in
In embodiments of the present invention the parameter c is selected depending on the application at hand. Thereby a parameter c different from one may be selected. The parameter c may be in the range 0 to 2. In this, characteristics similar to the characteristics shown in
When the input signal is zero within an integration interval defined by two consecutive sampling instants, any charge (VC) stored on the integrating capacitor (Cint) at the beginning of the integration interval will be linearly changed during said integration interval. This is illustrated in
Variants of the Circuit of
Before describing the noise behavior of the circuit, a number of variants of the circuit shown in
different input blocks that provide a chopped voltage signal can be used, as will be illustrated in
different first feedback paths can be used, as will be illustrated in
optionally a second feedback path can be added, as will be illustrated in
Different Input Blocks:
In the embodiment of
In
In
In a variant (not shown) of the circuit of
In a variant (not shown) of the circuit of
In yet another variant (not shown) of
In yet another variant (not shown) of
In a variant of the circuit of
In a variant of the circuit of
In
In
In another embodiment (not shown), the input block further comprises a transducer TR of the kind that converts a physical (e.g. non-electrical)signal or excitation into a voltage signal, the transducer being arranged upstream of the first chopper CH1 for providing the voltage signal Vin to be amplified in response to the physical (e.g. non-electrical)signal or excitation.
The transducer may for example comprise a piezo-electrical material adapted for generating a voltage in response to a pressure exerted upon the transducer, or the transductor may for example comprise a photovoltaic element adapted for generating a voltage signal in response to a light incident to the photovoltaic element, or an antenna adapted for converting electromagnetic energy into a voltage signal, or in general, any kind of transducer adapted for converting an energy signal into a voltage signal. Similar to what is mentioned above, the transducer may be implemented on the same die as the amplifier circuit, or may be implemented on a separate die. Variants of embodiments of
Different First Feedback Path:
In
If the feedback signal is applied as a voltage signal Vfb, the feedback path contains a proportional scaler. If the feedback path is applied as a current signal Ifb, the feedback path contains a transconductance G2. If the feedback signal is applied downstream of the second chopper CH2 or upstream of the first chopper CH1, a non-chopped feedback signal is to be provided. If the feedback signal is applied between the first chopper CH1 and the second chopper CH2, a chopped feedback signal is to be provided, in which case the feedback path additionally comprises a third chopper CH3, chopped at the same frequency fc as the first and second chopper CH1, CH2. Hence, it can be understood why:
in the embodiment of
in the embodiment of
in the embodiment of
in
Optional Second Feedback Path:
An advantage of the second feedback path is that it may further reduce the amount of up-converted offset and flicker noise. The third chopper CH3 demodulates the up-converted noise to low frequencies, which is integrated by the filter H3, and fed back as an offset compensation signal Vfb2. Additionally, the amplifier output signal is upconverted to fc by CH3, which is largely blocked by the integration and the optional zero of H3.
In the specific embodiment shown in
The following options are contemplated regarding the second feedback path:
no second feedback path is provided (as shown e.g. in
the second feedback signal is provided as a (voltage or current) feedback signal to an internal node of the first transconductance G1 (not shown);
the second feedback signal is provided as a current signal to a current summing node or current summing element located between the first transconductance G1 and the second chopper CH2 (not shown).
the second feedback signal is provided as a chopped current to the current integrator CI1, in which case an additional chopper is present between the output of H3 and the current integrator CI1, the additional chopper being operated at the same chopping frequency fc as CH2. (not shown).
It is pointed out that all combinations of input blocks described above (in relation to
Another table Table 2, not explicitly shown, but similar to Table 1 (and with N being an integer larger than zero, preferably N=1 or N=2) can be formed, containing “
Another table Table 3, not explicitly shown, but similar to Table 1 (and with N being an integer larger than zero, preferably N=1 or N=2) can be formed, containing “
Another table Table 4, not explicitly shown, but similar to Table 1 (and with N being an integer larger than zero, preferably N=1 or N=2) can be formed, containing “
The attentive reader will recognize that the third chopper CH3 in the circuit of
It is also clear to the skilled person that in any of the previous embodiments which comprise a ZOH followed by a third chopper CH3, the order of these two components may be interchanged when the sample frequency fs=N·fc with N being an integer value equal to or larger than 2. In this case, the ZOH block provides a chopped output signal rather than an unchopped signal.
Sizing of the Second Feedback Path
The purpose of the second feedback path is to reduce the spectral components corresponding with up-converted offset and flicker noise in the output signal(s). Since the up-converted offset and flicker noise can be referred to the input (Vin), the effect of the second feedback loop on the overall transfer from input Vin to an amplifier output is to attenuate frequencies around the chopping frequency fc and odd multiples thereof. The choice of the particular filter H3, and the sizing of its parameters affects the depth and width of the notches. An example of a suitable filter H3 has the form: H3(z)=α(1+1/z)/(z−1), in which a is a scale factor to be sized, and in which the denominator represents the integration, and the numerator the FIR factor (1+1/z).
Because of the omnipresence of computing power, the task of sizing this filter may be found numerically, by means of simulation for various parameter values. But analytical techniques are also possible. The combined action of the feedback path comprising CH3, the filter H3(z), and CH2 can be modeled as a transfer H3(−z)=α(1−1/z)/(z+1), that is the action of chopping before and after the filter H3 can be understood as implementing a frequency-transformation (z→−z) of the H3 filter. This allows to derive an equivalent system-level model of the system in
From the discussion of
the sampler of the first feedback path operated at fs1=2·fc and the same or physically different sampler of the second feedback path also operated at fs2=2·fc (as in
Simulation Results:
The behavior of the circuit shown in
Behavior in the Time-Domain:
Behavior in the Frequency-Domain:
At the time of making the invention, the inventors could not predict the precise impact the sample-and-hold block (sampled at fs), and the first feedback signal (as proportional feedback signal with a very specific proportionality factor) would have on the overall circuit behavior, in particular, on its noise behavior. In order to study the effects of noise folding within the feedback path, transfer functions where determined, and the result is shown in
Plot 1801 shows a typical (prior art) first order transfer function of an amplifier circuit without sample-and-hold circuit and without the second feedback path (offset feedback). This transfer is shown as a first-order reference characteristic, with a DC-gain equal to A0, and a 3 dB bandwidth at f/fs=0.44.
Plot 1802 is an approximate transfer function of the circuit shown in
Plot 1803 is a simulated transfer function of the circuit shown in
The frequency-band of interest is the band from DC to the 3 dB bandwidth at fBW=0.44 fs. The 3 dB bandwidth is close to the so called “Nyquist frequency” corresponding to f=fs/2, or f/fs=0.5. In practice, the input signal (e.g. coming from a sensor) may have higher frequencies, e.g. white noise, but these frequencies are undesired. An “ideal” amplifier would amplify the band of interest fBW with amplification factor A0, and would block all frequencies higher than fBW. Such an ideal amplifier does not exist in practice, and can only be approximated. One approach used in the prior art is to use a chopped amplifier with a bandwidth significantly exceeding fBW, and add thereafter higher order continuous-time filters, e.g. a second order RC-filter or Butterworth filter or Elliptic filter or Chebychev filter. Although the design of such filters is well described in literature, their realization in integrated circuitry requires active components, such as op-amps, which leads to additional power consumption. These filters also necessarily operate in continuous-time, which often require tuning due to the large spread in component values.
The present invention however, takes a different approach, by adding a sample-and-hold block to the first feedback loop as describe above, which can easily be integrated in a semiconductor device. This results in the transfer characteristic 1802, which is somewhat higher than the reference curve in the pass-band, and has deep notches in the frequency spectrum at multiples of the sample frequency (i.e. f/fs=integer). The amplifier with the “proportional discrete-time feedback path” as proposed by the present invention thus provides an inherent filtering, which leads to far less noise folding than the first-order reference curve (when the amplified signal is sampled), because the area under the curve 1802 is much lower than the area under the curve 1801. In fact, the graph of
It is noted that noise folding due to sampling only occurs for frequencies above the Nyquist frequency. Therefore, unwanted signals in the frequency range from f=fBW to f=fs/2 can still be removed, e.g. in the digital domain or by means of switched-capacitor filters.
As discussed above, by sizing the components of the second feedback path, the frequency response of the system can be tuned. As can be seen e.g. in
Another advantage of providing the second feedback loop is that the location of the steep edge (notch at f=fs/2) is mainly or solely defined by the chopping frequency fc and sampling frequency fs, and not by component values (e.g. resistors, capacitors, inductors), which typically have much higher tolerances (e.g. more than 10% or even more than 30% on their absolute value).
In order to quantify the reduction in noise-folding, an excess factor can be defined, corresponding with the area under the curves in
Sampling Frequency:
In the circuits shown in any of
Preferred Embodiments: fs=1·fc or fs=2·fc
The case fs=2·fc presented above (e.g. illustrated by any of the circuits shown in
Another interesting case is when chopping is performed at the frequency fc=fs, (e.g. illustrated by any of the circuits shown in
The solution to obtain these advantages are very simple to implement in an integrated circuit, and comes at almost no addition cost, in contrast to higher order analog filters.
Once understood, the solution (or “teaching”) of the present invention can be relatively easily added to existing circuits, such as to the prior art circuit shown in
Optionally a fourth chopper CH4 and a fifth chopper CH5 are added upstream and downstream of the sample-and-hold block, but these choppers CH4, CH5 are not absolutely required. When present, these choppers modulate the offset and flicker noise of the sample-and-hold block away from the band of interest. However, since the sample-and-hold block operates on large (amplified) signals, the relative contribution of its offset and flicker noise can often be tolerated. All choppers CH1 to CH3 and when present also CH4, CH5 may be operated at the same chopping frequency fc.
Although not explicitly shown, it shall be clear to the skilled person that also in the embodiments of
The operation of the circuits illustrated in
While
Variants of this circuit (not shown) may comprise other devices that act as a “nullor”. Practical implementations of nullors comprise active devices that provide gain, such as any type of transistor (MOSFET, bipolar, etc.), and whose features as a nullor emerges as the result of negative feedback being applied to an input of the nullor. It is noted that the nullator aspect of OTA1 and OTA2 is sufficient to understand how they implement at the same time the function of low-noise transconductance G1 (converting an input voltage into a current), and of the current integrator (CI1).
an input block comprising a first chopper CH1 adapted for providing a differential chopped input voltage signal indicative of a differential voltage signal Vin+, Vin− to be amplified;
a first Operational Transconductance Amplifier OTA1 having a non-inverting input port “p1” connected to a first output of the first chopper CH1, the first OTA being adapted for providing a first output signal out1;
a second Operational Transconductance Amplifier OTA2 having a non-inverting input port “p2” connected to a second output of the first chopper CH1, different from the first output, the second OTA being adapted for providing a second output signal out2;
a second chopper CH2 adapted for receiving the first respectively second output signal out1, out2 from the first respectively second Operational Transconductance
Amplifier OTA1, OTA2, the second chopper having a first output connected to a first node N1 and a second output connected to a second node N2;
a first integrating capacitor Cint1 arranged between the first node N1 and a third node N3;
a second integrating capacitor Cint2 arranged between the second node N2 and a fourth node N4;
a third chopper CH3 having a first input connected to the third node N3 and a second input connected to the fourth node N4 and having a first output connected to an inverting input “n1” of the first Operational Transconductance Amplifier OTA1 and having a second output connected to an inverting input “n2” of the second Operational Transconductance Amplifier OTA2;
a resistor network connected to the third, fourth, fifth and sixth node N3-N6;
a sample-and-hold circuit (S&H) having a first input in1 connected to the first node N1 and a second input in2 connected to the second node N2 and a first output out3 connected to the fifth node N5 and a second output out4 connected to the sixth node N6;
wherein a potential difference between the first and second node N1, N2 forming a continuous-time voltage signal Vct being an amplified version of the differential input signal Vin+, Vin−, a potential difference between the fifth and sixth node N5, N6 forming a zero-order-hold voltage signal Vzoh of said continuous-time voltage signal Vct;
and wherein the values of the integrating Capacitors Cint1, Cint2 and of the resistors Rfb1, Rfb2, Rin are chosen such that an amplitude ratio of the continuous-time signal Vct and the differential voltage signal to be amplified Vin is at least 2.0;
and wherein each chopper CH1, CH2, CH3 is adapted to be operated at a chopping frequency fc, and the sample-and-hold-block S&H is adapted to be operated at a sampling frequency fs, wherein the sampling frequency fs is equal to the chopping frequency fc or to an integer multiple N larger than zero thereof, such as for example once or twice the chopping frequency.
The resistor network shown in
a first feedback resistor Rfb1 arranged between the third node N3 and a fifth node N5, and
a second feedback resistor Rfb2 arranged between the fourth node N4 and a sixth node N6, and
a third resistor Rin arranged between the third node N3 and the fourth node N4,
but the invention is not limited thereto, and other resistor networks, e.g. a ladder network may also be used.
It is noted that OTA1 and OTA2 may be considered as two halves of a single differential amplifier.
This circuit functions very much the same as the circuits mentioned above, except that each nullor (e.g. an OTA) performs the role of both a transconductance G1 and a current integrator CI1. It is an advantage of this embodiment that it may require less active components as compared to embodiments described above.
The working principles of this circuit are very much the same as described above, in relation to
an input signal Vin is applied to a first chopper CH1,
the first chopper CH1 separates the useful signal band (baseband) and the noise added by the OTAs (represented by Vn),
the Op-Amps or OTAs provides a current signal I_ch at its output, the magnitude of this current being such that a feedback voltage Vfb matches Vin (nullor operation)
the second chopper CH2 demodulates the current signal I_ch back to baseband, resulting in a demodulated current I demod being applied to the integration capacitors Cint1, Cint2
this application of the demodulated current I demod to the integration capacitors Cint1, Cint2 results in accumulation of the corresponding charge, the voltages on the integration capacitors thus providing an output of a current integration CI1 which also provides a continuous-time output voltage Vct of the amplifier
a sample-and-hold circuit downstream of the second chopper CH2 samples and holds the continuous-time output voltage Vct,
a feedback path provides a signal proportional to the zero-order-hold voltage Vzoh as a feedback signal to remove the charge stored on the integrating capacitors Cint1, Cint2 in a single sampling period fs (assuming a zero input signal Vin).
Also here, the sampling frequency fs is an integer multiple of the chopping frequency fc, thus fs=N·fc, where N is an integer in the range of 1 to 10, preferably N=1 or N=2. The proportional feedback path is implemented by a scaler, more in particular by a resistor divider. Although only three resistors are shown, it is clear that other resistor dividers, e.g. based on a resistor network may be used. Although not shown, it shall be clear to the skilled person that also here different input blocks can be applied (as illustrated in
Indeed, the inventors have also added a second chopper CH2, but arranged between the output of the amplifiers A1, A2 and a first (lower) terminal of the integrating capacitors C1, C2, and they added a third chopper CH3 between the other (upper) terminal of the integrating capacitors C1, C2 and the inverting inputs of the amplifiers, and they inserted a sample-and-hold circuit between the output of the amplifiers A1, A2 and the resistor network R1, R2, R3, and they specifically dimensioned the components of the feedback path such that a charge stored on the capacitors C1, C2 is removed after one sampling period TS=1/fs (assuming a zero input signal). As mentioned above, the sampling frequency fs is chosen equal to an integer multiple of the chopping frequency fc, thus:
fs=N·fc, with N being an integer larger than zero, preferably N=1 or N=2.
In other words, the semiconductor amplifier circuit 3100 comprises:
an input block comprising a first chopper CH1 adapted for receiving a differential voltage input signal in+, in− to be amplified, and for chopping said signal at a chopper frequency fc, a first output of the first chopper CH1 being connected to a gate GG1 of a first transistor T1, a second output of the first chopper CH1 being connected to a gate GG2 of a second transistor T2;
the first transistor T1 having a source SS1 connected through a first current source Ib1 to ground, and having a drain DD1 connected through a second current source Ib2 to a supply voltage vdd;
the second transistor T2 having a source SS2 connected through a third current source Ib3 to ground, and having a drain DD2 connected through a fourth current source Ib4 to a supply voltage vdd;
a first Operational Amplifier A1 and a second Operational Amplifier A2, a non-inverting input “p1” of the first Operational Amplifier A1 and a non-inverting input “p2” of the second Operational Amplifier A2 being connected to a constant or a common voltage VB;
a sample-and-hold block having a first input in1 connected to an output out1 of the first Operational Amplifier A1, and having a second input in2 connected to an output out2 of the second Operational Amplifier A2, and having a first and second output out3, out4 connected via a resistor network R1, R2, R3 to the source SS1 of the first transistor T1 and to the source SS2 of the second transistor for providing a differential feedback signal Vfb;
a second chopper CH2 having a first respectively second input connected to a first respectively second output of the first respectively second Operational Amplifier A1, A2, and having a first respectively second output connected to a first respectively second output node +OUT, −OUT, a potential difference between these nodes forming a differential output voltage Vout as a continuous-time voltage signal being an amplified version of the differential input voltage;
a first integrating capacitor C1 having one terminal connected to the first output node +OUT and another terminal connected to a first input of a third chopper CH3;
a second integrating capacitor C2 having with one terminal connected to the second output node −OUT and another terminal connected to a second input of a third chopper CH3;
the third chopper CH3 having a first output connected to the drain DD1 of the first transistor T1 and connected to the inverting input “n1” of the first Operational Amplifier A1, and having a second output connected to the drain DD2 of the second transistor T2 and connected to the inverting input “n2” of the second Operational Amplifier A2;
and wherein the values of the integrating Capacitors Cint1, Cint2 and of the resistors Rfb1, Rfb2, Rin of the resistor network are chosen such that an amplitude ratio of the differential continuous-time output signal and the differential voltage input signal Vin is at least 2.0;
and wherein each chopper is adapted to be operated at a chopping frequency fc, and the sample-and-hold-block is adapted to be operated at a sampling frequency fs,
wherein the sampling frequency fs is equal to the chopping frequency fc, or an integer multiple (larger than zero) thereof.
Although not explicitly shown, it shall be clear to the skilled person that also in this circuit different input blocks can be added (as illustrated in
In alternative embodiments (not shown) the first and second transistor T1, T2 may be PMOS-type MOSFETs, or bipolar transistors, or circuits that may act as a (more ideal) transistor such as Darlington-pairs, cascoded transistors, or in general a “current conveyor”, e.g. a “first generation current conveyor” or a “second generation current conveyor” (i.e. a three-terminal device with terminals designated X, Y, and Z, in which the potential at X equals whatever voltage is applied to Y, and the current flowing into X is mirrored at Z with a high output impedance).
Variants of
The embodiments of
A semiconductor amplifier circuit 2600, 2700, 2800, 3100, comprising:
an input block comprising a first chopper CH1 adapted for providing a chopped input voltage signal Vin_ch indicative of a voltage signal Vin to be amplified;
at least one amplifier OTA1, OTA2; AMP1, AMP2; A1, A2, each having an inverting input port n1, n2 and a non-inverting input port p1, p2 and an output port q1, q2, and each being adapted to provide an output signal out1, out2 on its output port q1, q2 such that a voltage on its inverting input port n1, n2 is substantially equal to a voltage on its non-inverting input port p1, p2,
and wherein either the output port of the first chopper CH1 is connected to the non-inverting input port p1, p2 of the amplifier (as illustrated e.g. in
a second chopper CH2 having an input port and an output port, the input port being connected to the output port q1, q2 of the at least one amplifier;
a third chopper CH3 having an input port and an output port, the output port being connected to the inverting input port n1, n2 of the at least one amplifier;
at least one integration capacitor Cint1, Cint2; C1, C2 connected between the output port of the second chopper CH2 and the input port of the third chopper CH3;
a sample-and-hold circuit S&H having an input port in1, in2 connected to the output port q1, q2 of the amplifier or to the output port of the second chopper CH2, directly and/or indirectly through one or more intermediate components CH4, and having an output port out3, out4 for providing a zero order hold voltage signal Vzoh;
a proportional feedback network Rfb1, Rfb2, Rin; R1, R2, R3 having an input port connected directly and/or indirectly through one or more intermediate components CH5 to the output port of the sample-and-hold circuit S&H, and being adapted for providing a proportional feedback signal Vfb; Vfb; Ifb1, Ifb2, and having an output port connected directly or indirectly through one or more intermediate components CH5; T1, T2 to the inverting input port n1, n2 of the at least one amplifier;
wherein a voltage of the at least one integration capacitor provides a continuous-time voltage signal Vct being an amplified version of the voltage signal to be amplified Vin;
and wherein the values of the proportional feedback network are chosen such that an amplitude ratio of the continuous-time signal Vct and the voltage signal to be amplified Vin is larger than 1.0 for a predefined frequency range (e.g. at least 2.0);
and wherein the value of the at least one integration capacitor is chosen such that, for a chopped input signal Vin_ch equal to zero, a charge stored on the integrating capacitor at the beginning of a sample period Ts is linearly removed during one single sampling period Ts of the sampling signal, in such a way that the absolute value of the charge is smaller at the end of the sampling period than at the beginning of the sampling period. It may for example be completely removed at the end of the single sampling period Ts;
and wherein each chopper CH1, CH2, CH3, and also CH4, CH5 if present, is adapted to be operated at a chopping frequency fc, and the sample-and-hold-block S&H is adapted to be operated at a sampling frequency fs, wherein the sampling frequency fs is equal to the chopping frequency fc (i.e. fs=1·fc) or the sampling frequency fs is equal to an integer number (N) larger than zero times the chopping frequency fc (i.e. fs=N·fc; N>0).
Further Variants
Although only shown in
Although the invention is described primarily for a circuit with a deadbeat response equal to a single sampling period Ts=1/fs, (meaning that the charge on the capacitor at the start of a sample interval is removed after exactly one period Ts), the circuits of
When in the disclosed embodiments choppers have the same chopping frequency fc, there may be differences in the phase of the modulation signals being used. Also, the signals controlling the chopper switches may be adapted to mitigate parasitic effects, such as charge injection, e.g. by introducing some zero-banding of transient spikes, but these are considered implementation details and are well known in the art of chip design.
In exemplary embodiments of the present invention which have been detailed with the inclusion of chopping, e.g. the embodiments of
This leads to embodiments of the present invention which can be described as follows:
A semiconductor amplifier circuit, comprising: an input block adapted for receiving a voltage signal to be amplified; at least one amplifier having an input port which comprises at least one input node and an output port which comprises at least one output node; at least one integration capacitor connected in a negative feedback configuration between the at least one output node and the at least one input node of the at least one amplifier. In such an embodiment the semiconductor amplifier circuit is configured such that, during operation, the voltage signal to be amplified directly or indirectly (i.e. through one or more intermediate components) arrives at the input port of the at least one amplifier. Embodiments of the present invention may comprise a connection for passing the voltage signal to be amplified to an amplifier input node which is also connected to the at least one integration capacitor. This connection may be an indirect connection comprising one or more intermediate components adapted for converting the voltage signal to be amplified into a current signal. Embodiments of the present invention comprise a sample and hold circuit having an input port connected to the output port of the at least one amplifier, directly and/or indirectly (i.e. through one or more intermediate components), and having an output port for providing a zero order hold voltage signal; a proportional feedback network having an input port connected directly and/or indirectly through one or more intermediate components to the output port of the sample and hold circuit, and being adapted for providing a proportional feedback signal, and having an output port connected directly and/or indirectly (i.e. through one or more intermediate components) to the input port of the at least one amplifier; wherein a voltage of the at least one integration capacitor provides a continuous-time voltage signal being an amplified version of the voltage signal to be amplified; and wherein the values of the proportional feedback network are chosen such that an amplitude ratio of the continuous-time signal and the voltage signal to be amplified is larger than 1.0 for a predefined frequency range (e.g. for frequencies of the voltage signal to be amplified from DC up to a frequency representative for the bandwidth of the semiconductor (amplifier) circuit); and wherein the value of the at least one integration capacitor is chosen such that, for an input signal equal to zero, a charge stored on the integrating capacitor at the beginning of a sample period is linearly removed during the sampling period of the sampling signal, in such a way that the absolute value of the charge is smaller at the end of the sampling period than at the beginning of the sampling period.
| Number | Date | Country | Kind |
|---|---|---|---|
| 1600774.2 | Jan 2016 | GB | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 15402502 | Jan 2017 | US |
| Child | 16176242 | US |
