INTEGRATION CIRCUIT AND METHOD FOR PROVIDING AN OUTPUT SIGNAL The field of the present application concerns an integration circuit or integration stage and a method for providing an output signal.

An integration circuit is a building block often employed in delta sigma analog-to-digital converters. Said delta sigma converters in turn are often used in temperature sensors.

A state of the art integrator circuit as shown in Figure 5 has an amplifier A with a capacitance C coupled in the amplifier's feedback loop such that amplifier A and

capacitance C form an inverting integrator. A first and a second current I I , 12 are alternately supplied to the input of the amplifier A. The amplifier A provides an output signal Vout representing an integration of a current ratio or current difference between the first and the second current

I I , 12. While one current is not switched to the input of the amplifier, i.e. while this current is idle, it is switched to a terminal R to which a reference voltage VR is supplied. A parasitic capacitance Cpl, Cp2 builds up at each input which receives the first or the second current I I , 12. Voltage swings across these parasitic capacitances Cpl and Cp2 lead to charge being injected into the amplifier A, thus

distorting the actual current ratio or current difference. Whereas voltage transients across the parasitic capacitance Cpi at the amplifier's A virtual ground node do not imply charge errors on average, the parasitic capacitances Cpl and Cp2 each are merely connected to the amplifier A for half a clock cycle of clock signal elk shown on the right side of Figure 5 and are recharged to the voltage VR of the terminal R during each idle phase. Consequently, during each active phase, i.e. while the respective current II or 12 is

connected to the input of the amplifier A, a respective charge difference resulting from the difference between the voltage VI, V2 at the input of the respective current and the voltage Vr at the terminal R is injected into the amplifier A. This charge is stored on the respective parasitic

capacitance Cpl, Cp2, thereby causing respective error currents. Each error current therein amounts to the quotient of a respective charge difference and a length of the period of the clock signal elk according to the following equations:

AQ1,2 = Cpl,2 * (VR- Vl,2)

Therein, Il,2err denotes the error current of first or second current II, 12, AQ1,2 denotes a difference in charge stored in first or second parasitic capacitance Cpl, Cp2,

respectively, Cpl, 2 denotes the first or second parasitic capacitance Cpl, Cp2, VR denotes the reference voltage VR and VI, 2 denotes the voltage VI or V2.

Whereas the error charge packages injected into the amplifier A are constant, the input currents II, 12 are continuous in time, so that the impact on the overall integrated charge depends on the period of the clock signal elk.

Two common solutions exist to address the above-mentioned problem. One possibility is to reduce the frequency of clock signal elk. This reduces the error contribution of the charge packages. However, this increases a total conversion time if this integrator is used e.g. in an analog-to-digital

converter. Longer conversion time also implies into higher power consumption. Another remedy consists in introducing a cascode switching scheme at the sensitive nodes, i.e. the input nodes of the currents II and 12, in order to reduce the voltage swing. However, offset between the cascodes still gives rise to residual swing. Moreover, when using the integrator in low supply voltage implementations, there may not be enough headroom for the cascade switches.

It is therefore an objective of the present application to overcome the shortcomings of the prior art.

The objective is solved by the subject-matter of the

independent patent claims. Embodiments and developments are the subject-matter of the dependent claims. In one embodiment an integration circuit has a first input terminal configured to receive a first input signal, a second input terminal configured to receive a second input signal, an output terminal to provide an output signal as a function of the first and the second input signal, a first and a second amplifier, and a capacitor. Each of first and second amplifiers is switchably connected between the first and the second input terminal and the output terminal. The capacitor is switchably coupled in a feedback loop either of the first or of the second amplifier such that the capacitor and one of the first and the second amplifiers form an inverting

integrator providing the output signal. The integration circuit is prepared to be operated in a first and a second subphase. In each of first and second subphases one of the first and the second input signals is supplied to the

inverting integrator and the respective other one of first and second input signals is supplied to the respective other one of the first and the second amplifier. Therein the output signal is provided as a function of a time continuous integration of a difference between or a ratio of the first and the second input signal.

For example, in the first subphase the first input signal is supplied to the first amplifier and the second input signal is supplied to the second amplifier. The first amplifier and the capacitor form an inverting integrator and provide the output signal, in this example. Consequently, in the first subphase in this example the output signal is provided as an integration of the first input signal. In the second subphase in this example, the second input signal is supplied to the first amplifier and the first input signal is supplied to the second amplifier. The output signal is consequently provided as a function of an integration of the second input signal. Therefore, during the phase in which an input signal does not contribute to the output signal, i.e. its idle phase, this input signal is switched to the amplifier which does not provide the output signal. By this, the virtual ground voltages of the first and the second amplifiers are matched. The virtual ground at the input of the amplifier which does not provide the output signal exhibits the same swing during the idle phase of this signal as the amplifier which forms the inverting integrator and provides the output signal during the previous active phase of this signal.

Thus, a voltage swing at the first and the second input terminals is greatly reduced. There is no need to increase the supply voltage or to reduce the clock frequency as in state of the art implementations.

In an exemplary embodiment the second amplifier is matched to the first amplifier.

In a development the second input signal is complementary to the first input signal. Therein, first and second input signals each comprise a current signal.

In a further development, each of the first input signal, the second input signal and the output signal is a single-ended signal . In other words, the signals used in the integration circuit, for example the first input signal, the second input signal and the output signal are not implemented as differential signals, but rather as single-ended signals. First and second input signals are each realized as a time continuous signal.

In a further development the integration circuit is prepared to be operated in a third and fourth subphase. In each of third and fourth subphases with respect to the first and second subphases, the respective other one of the first and the second amplifier forms the inverting integrator. In each of third and fourth subphases one of the first and the second input signals is supplied to the inverting integrator and the respective other one of the first and the second input signals is supplied to the respective other one of the first and the second amplifiers with respect to the preceding third or fourth subphase. In an example, in the third subphase, the first input signal is supplied to the first amplifier and the second input signal is supplied to the second amplifier which forms the inverting integrator and provides the output signal.

Therefore, in the third subphase the output signal is provided as a function of an integration of the second input signal. In the same example in the fourth subphase the first input signal is supplied to the second amplifier which forms the inverting integrator and the second input signal is supplied to the first amplifier. The output signal in the fourth subphase is provided as a function of an integration of the first input signal. In the third and fourth subphases the other amplifier forms the inverting integrator with respect to the amplifier which is used in the inverting integrator in the first and second subphases. In the described example, in the first and second subphases the first amplifier forms the inverting integrator and in the third and fourth subphases the second amplifier forms the inverting integrator.

By this swapping of roles of the amplifiers a residual error resulting from mismatch, for example in offset between the first and the second amplifier, is cancelled on average. Consequently, voltage swings which occur in the prior art implementations on the first and the second input terminals are reduced to nearly zero on average. Hence, charge errors are eliminated.

In a further development a transition between the first and the second subphase and vice versa and a transition between the third and the fourth subphase and vice versa is defined by interchanging a connection of the first and the second input terminal to the inverting integrator or the respective other one of the first and the second amplifier with respect to a connection of the first and the second input terminal to the inverting integrator or the respective other one of the first and the second amplifier during a preceding subphase.

When operation of the integration circuit passes over from first to second subphase or from second to first subphase or from third to fourth subphase or from fourth to third

subphase the switchable connection between the first and the second input terminal and the first and the second amplifier is swapped or exchanged. Consequently, if in a subphase the first input terminal is coupled to the first amplifier and the second input terminal is coupled to the second amplifier, in the next or subsequent subphase the first input terminal is coupled to the second amplifier and the second input terminal is coupled to the first amplifier. In a refinement the integration circuit is prepared to be operated during a predefined measurement period which is divided in a first and a second phase. During the first phase of the measurement period a sequence having the first and the second subphases is repeated. During the second phase of the measurement period a sequence having the third and the fourth subphases is repeated.

A length of the measurement period is predefined according to the application in which the integration circuit is used.

In a further development a transition from the first to the second phase is defined by changing over a connection of the capacitor in the feedback loop of the first or of the second amplifier from the first to the second amplifier or vice versa such that with respect to the preceding first or second phase the other amplifier and the capacitor form the

inverting integrator.

This means that if during the first phase of the measurement period the first amplifier together with the capacitor form the inverting integrator, the connection of the capacitor is changed over to the second amplifier such that in the

following or subsequent second phase of the measurement period the second amplifier together with the capacitor form the inverting integrator.

Consequently, the role of forming the inverting integrator is exchanged once during the predefined measurement period from first to second amplifier or vice versa.

In a refinement the changing over of the connection of the capacitor is effected such that in the first phase the first amplifier and the capacitor form the inverting integrator which is prepared to provide the output signal and in the second phase the second amplifier and the capacitor form the inverting integrator or vice versa. In a further development the first amplifier has an inverting input terminal which is switchably connected either to the first or to the second input terminal as a function of the first, the second, the third and the fourth subphase. The first amplifier further has a non-inverting input terminal which receives a reference signal. The second amplifier has an inverting input terminal which is switchably connected either to the first or to the second input terminal as a function of the first, the second, the third and the fourth subphase. The second amplifier further has a non-inverting input terminal which receives the reference signal.

The reference signal is represented for instance by a reference voltage. Said reference voltage is dimensioned such that the integration circuit works at the correct operating point. For example, the reference voltage amounts to half of a supply voltage of the integration circuit. In a further development the feedback loop of the first amplifier extends between an output of the first amplifier and its inverting input terminal. The feedback loop of the second amplifier extends between an output of the second amplifier and its inverting input terminal.

In a refinement the amplifier which does not form the

inverting integrator in a respective subphase is configured in unity feedback forming a dummy amplifier. The amplifier forming the dummy amplifier is configured in full series feedback by connecting its output to its negative input terminal. This amplifier can also be denoted as a voltage follower. The dummy amplifier can also be called a "replica amplifier". The amplifier forming the inverting integrator and the dummy amplifier are always connected to one of the input terminals.

In one embodiment a delta sigma converter has an integration circuit as described above and a comparator circuit which is coupled to the output terminal of the integration circuit. The comparator circuit is prepared to provide a comparator signal as a function of comparison of the output signal of the integration circuit with a threshold signal. Due to the fact that first and second input terminals are complementary to each other, each of them representing for instance a current, the output signal, e.g. a voltage, at the output terminal of the integration circuit increases and decreases linearly in a triangular form. The threshold signal is chosen such that it is approximately in the middle between maximum and minimum value of the output signal. Consequently, the comparator signal changes its value each time the output signal of the integration circuit assumes the value of the threshold signal.

In an exemplary embodiment the delta sigma converter as described above is used in a temperature sensor.

The temperature sensor is for instance implemented as a current domain temperature to digital converter. Therein a frontend of the temperature sensor, which may be implemented in bipolar technology, generates a first current proportional to absolute temperature and a second current which is

complementary to absolute temperature. These two currents are used as first and second input signals for the integration circuit of the delta sigma converter. A digital pulse

modulated signal is provided by the integration circuit containing the information about a ratio between absolute temperature and complementary temperature. Said signal is passed through digital post processing and finally represents the temperature in a digital form. By using the integration circuit as described above, faster switching is possible when compared to state of the art integrators which decreases the time needed for conversion of the temperature. In one embodiment a method for providing an output signal has the following steps:

- in a first subphase feeding a first input signal to a first amplifier forming an inverting integrator, feeding a second input signal to a second amplifier forming a dummy amplifier and providing an output signal by integrating the first input signal, and

- in a second subphase feeding the first input signal to the second amplifier forming the dummy amplifier, feeding the second input signal to the first amplifier forming the inverting integrator and providing the output signal by integrating the second input signal, therein the dummy amplifier is configured in unity feedback, and

- the output signal is provided as a function of a time continuous integration of a difference between or a ratio of the first and the second input signal.

During the first and the second subphases the first amplifier forms the inverting integrator which provides the output signal and the second amplifier forms the dummy amplifier. Consequently, first and second input signals are switched from the first to the second amplifier and vice versa. By this, an occurrence of error currents is greatly reduced.

In a development the method further comprises the following steps : - in a third subphase feeding the first input signal to the first amplifier forming the dummy amplifier, feeding the second input signal to the second amplifier forming the inverting integrator and providing the output signal by integrating the second input signal, and

- in a fourth subphase feeding the first input signal to the second amplifier forming the inverting integrator, feeding the second input signal to the first amplifier forming the dummy amplifier and providing the output signal by integrating the first input signal.

Therein, first and second subphases are repeated during a first phase of a predefined measurement period. Third and fourth subphases are repeated during a second phase of the predefined measurement period.

The method can be implemented in the integration circuit described above, for example.

The text below explains the proposed integration circuit and corresponding method in detail using exemplary embodiments with reference to the drawings. Components and circuit elements that are functionally identical or have the

identical effect bear identical reference numbers. In so far as circuit parts or components correspond to one another in function, a description of them will not be repeated in each of the following figures. Figure 1 shows an embodiment example of the proposed

integration circuit; Figure 2 shows an exemplary timing diagram for the proposed integration circuit;

Figure 3A shows an embodiment example of the proposed

integration circuit during the first subphase;

Figure 3B shows an embodiment example of the proposed

integration circuit during the second subphase;

Figure 3C shows an embodiment example of the proposed

integration circuit during the third subphase

Figure 3D shows an embodiment example of the proposed

integration circuit during the fourth subphase;

Figure 4 shows the use of a delta sigma converter in a

temperature sensor according to the proposed principle . Figure 1 shows an embodiment example of the proposed

integration circuit. The integration circuit has a first and a second input terminal Inl, In2, a first amplifier Al, a second amplifier A2, a capacitor C and an output terminal Out. The first input terminal Inl is configured to receive a first input signal II which comprises, for example, a first input current. The second input terminal In2 is configured to receive a second input signal 12 which comprises, for

instance, a second input current. The first amplifier Al is switchably connected between the first and the second input terminal Inl, In2 and the output terminal Out. The second amplifier A2 is switchably connected between the first and the second input terminal Inl, In2 and the output terminal Out. In detail, the first input terminal Inl is switchably connected to an inverting input terminal of the first

amplifier Al by a first switch SI. The first input terminal Inl is further switchably connected to an inverting input terminal of the second amplifier A2 by means of a second switch S2. The second input terminal In2 is switchably connected to the inverting input of the first amplifier Al by a third switch S3. The second input terminal In2 is further coupled to the inverting input of the second amplifier A2 by means of a fourth switch S4.

The capacitor C is switchably coupled in a feedback loop either of the first amplifier Al or of the second amplifier A2 by means of a first chopping switch CS1 and a second chopping switch CS2. The amplifier out of first and second amplifiers Al, A2 which has the capacitor C switched into its feedback loop by means of chopping switches CS1 and CS2 forms an inverting integrator. The respective other amplifier out of first and second amplifiers Al, A2 is configured in unity feedback with its output being connected to its inverting input by means of chopping switches CS1, CS2 thereby forming a dummy or replica amplifier. An output of the amplifier out of first and second amplifiers Al, A2 which forms the

inverting integrator is switched to the output terminal Out of the integration circuit by means of the second chopping switch CS2 for providing the output signal Out. The first and the second chopping switch CS1, CS2 are concurrently

controlled by a second clock signal clk2.

A chopping switch is also called a swapping switch,

implementation of which is known to those skilled in the art.

First, second, third and fourth switches SI, S2, S3, S4 are controlled by a first clock signal clkl and the inverted first clock signal clkl . In detail, the second and the third switch S2, S3 are controlled by the first clock signal clkl, whereas the first and the fourth switch SI, S4 are controlled by the inverted first clock signal clkl .

The non-inverting inputs of the first and the second

amplifiers Al, A2 each receive a reference signal Vref.

Figure 1 also shows parasitic capacitances which build up at the input terminals Inl, In2. Namely, a first parasitic capacitance Cpl builds up at the first input terminal Inl with respect to a reference potential terminal 10. Similarly, at the second input terminal In2 a second parasitic

capacitance Cp2 builds up with reference to the reference potential terminal 10. Operation of the integration circuit will be described below with reference to Figure 2 and

Figures 3A to 3D.

Figure 2 shows an exemplary timing diagram for the proposed integration circuit. The first and the second clock signals clkl, clk2 are depicted with respect to time t. A measurement period of the integration circuit is made up of two phases, namely a first phase PI and a second P2. First and second phases PI, P2 are determined by the second clock signal clk2. The first phase PI further has a number of first and second subphases SP1, SP2 determined by the first clock signal clkl. The second phase P2 has a number of third and fourth

subphases SP3, SP4 which also are determined by the first clock signal clkl.

Figure 3A shows an embodiment example of the proposed

integration circuit during the first subphase. During the first subphase SP1 as of Figure 2 the first amplifier Al is configured as inverting integrator Int, whereas the second amplifier A2 forms a dummy amplifier. Under control of the first clock signal clkl, the second and the third switches S2, S3 are closed, whereas the first and the fourth switches SI, S4 are opened. Consequently, the second input signal 12 is provided to the inverting integrator Int and the first input signal II is provided to the dummy amplifier A2. The output signal Vout is provided as an integration of the second input signal Inl . A first voltage VI drops at the first input terminal Inl. A second voltage V2 drops at the second input terminal In2.

Figure 3B shows an embodiment example of the proposed

integration circuit during the second subphase. The first amplifier Al continues to form the inverting integrator Int together with the capacitor C. Under control of the first clock signal clkl first and fourth switches SI, S4 are closed and second and third switches S2, S3 are opened.

Consequently, the first input signal II is supplied to the inverting integrator Int. The second input signal 12 is supplied to the dummy amplifier A2. The output signal Vout is provided as a function of an integration of the first input signal 11. It can be seen that instead of switching one of the input terminals Inl, In2 which is currently not used for

integration, i.e. it is idle, to a reference potential terminal as in state of the art implementations described in the opening part of this application, according to the proposed principle the input terminal whose input signal is not currently integrated is switched to the dummy amplifier. Because of a matching between the first and the second amplifiers Al, A2 and consequently matched virtual ground voltages, a voltage swing at the input terminals Inl, In2 is greatly reduced. There is no need to increase a supply voltage nor to decrease a frequency of the first clock signal clkl. The proposed integration circuit can be operated with a clock frequency of approximately 1 MHz, for example.

A voltage swing at an input node therein denotes the

difference in voltage at this node which occurs when

switching over from the first subphase SP1 to the second subphase SP2 and vice versa.

Figure 3C shows an embodiment example of the proposed

integration circuit during the third subphase SP3. Under control of the second clock signal clk2 the integration circuit is operated in the second phase P2 according to

Figure 2 and the roles of inverting integrator and dummy amplifier are swapped between first and second amplifier Al, A2 using the chopping switches CS1, CS2 of Figure 1.

Consequently, during the third subphase SP3 the second amplifier A2 forms the inverting integrator Int together with the capacitor C and provides the output signal Vout. The first amplifier Al forms the dummy amplifier configured in unity feedback. Under control of the first clock signal clkl the switches are operated such that the first input signal II is supplied to the dummy amplifier Al and the second input signal 12 is provided to the inverting integrator Int. The output signal Vout is provided in function of an integration of the second input signal 12. Therein, clkl represents the inverted second clock signal clk2.

Figure 3D shows an embodiment example of the proposed

integration circuit during the fourth subphase. In the fourth subphase SP4 as of Figure 2 the second amplifier A2 continues to form the inverting integrator Int. Under control of the first clock signal clkl, the first input signal Inl is supplied to the inverting integrator Int and the second input signal 12 is supplied to the dummy amplifier Al . The output signal Vout is provided as a function of an integration of the first input signal II.

It can be seen that during switchover from the third to the fourth subphase SP3, SP4 the input signals of first and second amplifiers Al, A2 are swapped. Thereby a voltage swing across the parasitic capacitances Cpl, Cp2 of Figure 1 is greatly reduced. Furthermore, by swapping the roles of inverting integrator Int and dummy amplifier between first and second amplifiers Al, A2 when switching over from the first phase PI to the second phase P2 in the operation of the integration circuit, a residual swing which is caused by mismatch of the operational amplifiers and potential

differences in offsets is further reduced. This residual error is cancelled on average.

Figure 4 shows the use of a delta sigma converter in a temperature sensor according to the proposed principle. The delta sigma converter Conv has an integration circuit IC according to the proposed principle described above and a comparator circuit Cmp which is coupled to the output of the integration circuit IC. The comparator circuit Cmp compares the output signal Vout of the integration circuit IC to a threshold signal Vth and therefrom provides a comparison signal Sc.

The comparison signal Sc is provided to a counter component Ctr which provides a number of zero crossings of the comparison signal Sc. After some further filtering a

temperature signal Stmp is provided.

When the delta sigma converter Conv is used in a temperature sensor a bipolar frontend FE is applied to provide the first and the second input signals II, 12 as is known to those skilled in the art.

The bipolar frontend FE has a proportional to absolute temperature, PTAT, sensor which provides the first input signal II in the form of the first input current. The bipolar frontend further has a complementary to absolute temperature, CTAT, sensor which provides the second input signal 12 in the form of the second input current. The integration circuit IC within the delta sigma converter Conv provides a digital pulse modulated signal that represents the ratio of first to second input signal II, 12 in the form of the comparison signal Sc. After applying digital post-processing in the form of the counter Ctr, the measured temperature is digitally represented in the temperature signal Stmp.

It is to be understood that any feature described in relation to any one embodiment may be used alone, or in combination with other features described, and may also be used in combination with one or more features of any other of the embodiments, or any combination of any other of the

embodiments unless described as alternative. Furthermore, equivalents and modifications not described above may also be employed without departing from the scope of the carrier signal generation circuit and method which are defined in the accompanying claims. Reference list

10 reference potential terminal

Inl, In2, Out input/output terminal

II, 12, Vout, Vref, Vth signal

VI, V2, VR voltage

Al, A2, A amp1ifier

SI, S2, S3, S4 switch

CS1, CS2 switch

Int inverting integrator

C capacitor

clkl, clk\ , clk2, clk2 clock signal

Cpl, Cp2, Cpi parasitic capacitance

SP1, SP2, SP3, SP4 subphase

PI, P2 phase

IC integration circuit

Conv delta sigma converter

Sc, Stmp signal

Ctr counter

Cmp comparator

Inv inverter

FE front end