Invention relates to a measurement and evaluation system for biosignals,
in particular integrated circuits for such signals.
the body functions
of living things is a variety of different measuring and Auswertevorrichtungen
used. The use of these systems affects hospitals, general practitioners, but
also the so-called home monitoring, in which preferably monitors in the
Patients will be installed to this in their usual environment
allow. There are a very large number of possible uses
and applications of medical monitoring and diagnostic systems,
Short-term analysis method, but also for long-term monitoring
are provided. In the focus of most monitoring and diagnostic systems
Vital signs such as the electrocardiogram (ECG) or the
Electroencephalogram (EEG). The ECG and the EEG apply to so-called
bioelectric signals emitted by the body due to muscular or
be generated and over
Electrodes on or in the body
can be tapped or measured. From the very specific
Prescribed signals are derived parameters that diagnostic
and prognostic statements about
to enable the health of patients. In the field of bioelectric
Signals, e.g. ECG and EEG, there are a variety of applications,
such as. the long-term ECG, the stress ECG, and many more
Parameters such as blood pressure, pulse rate, or blood oxygen content
Central to the present invention are extracorporeal monitoring systems for primary and
chronic diseases of the cardiovascular system, hypertension
and heart rhythm disorders,
miniaturized monitoring systems
and recording the aforementioned anomalies. The monitoring systems include,
especially clinical monitoring, home monitoring,
emergency medicine and sleep diagnostics. The classic monitoring systems
derive the biosignals by means of sensors from the body, transmit the signals via cable
to a measuring or evaluation station, in which the information content
the signals are evaluated and displayed. Problematic
doing so that the patient is bound by the cables to the evaluation station
is. To enable the patient more mobility, numerous
suggested in which the data is carried in a close to the body
electronic preprocessing unit (handset) processed so far
that these are e.g. be stored and / or by radio or similar. to a
Transfer base station
This generally creates the need for a portable electronic
Unit (handset), which provide the necessary preprocessing steps, such as
Conversion, in particular also analog-to-digital conversion can make
and still largely autonomous. Because electrical energy only over short
The wireless transmission of distances and small amounts by electromagnetic waves leads to the
Autonomy of the handsets usually for the use of accumulators / batteries.
The size and that
The weight of the batteries depends on the power consumption or
the energy consumption of portable, mobile electronics.
Depending on the requirements of the functionality of the electronics, the weight
and the size of the handsets
again essentially dominated by the accumulators or batteries.
So there is a general need, the handsets so small
to shape, and a special need, the power consumption or
the power consumption of the handset in terms of size and
some of the required parameters only rarely occur
Events are to be determined, or statistically only via long-term studies
There is also a need to identify the devices that
Measuring and evaluation systems
be used to optimize for long-term use,
i.e. these as mobile as possible,
so small and light, as well as reliable, robust, skin-friendly and
to design comfortably. Long-term monitoring in the sense of the present
Invention is a monitoring,
at least seven, but better fourteen days without interruption
and from an autonomous energy source (such as a
Battery or a battery).
increasing integration of electronic circuits on semiconductor chips
offers great opportunities for this.
So not only are the structures constantly being reduced, which are on
can be mapped to the semiconductors, at the same time increases the possible number
the functions (complexity)
per unit area, and
the power requirement of the functional unit decreases.
From the German patent application DE 100 28 460 A1
(Schwarz GmbH) is a signal processing circuit having a plurality of signal processing blocks, each having a signal input for receiving each of an analog signal and a signal output known, each of the signal processing blocks has a coupled to the signal input amplifier unit for amplifying an analog signal supplied thereto. The amplifier unit is suitable for detecting and amplifying bioelectrical signals and has a variable amplification of the analog signal which can be applied to its input. In addition to the variable gain a reference formation of a plurality of analog input signals is provided, which provides a reference signal, which in turn is subtracted from the input stages directly from the input signal.
It is an object of the invention to provide electronic circuits
represent the recording and processing of biological vital signs
improve the human and animal body. It is one
Another object of the invention to provide an electronic device which
smaller and lighter than conventional devices and vital signs
nonetheless in a clinically usable and approved manner.
The object is achieved by a
Measuring system for monitoring
solved by biosignals,
which at least one handset, which is worn on the body, and a
Base station for receiving measurement data. In accordance with the invention
the mobile part has a plurality of channels for processing the biosignals,
wherein a channel at least a first amplifier stage with an input
Biosignal has and a second input for receiving a
Compensation signal, wherein the compensation signal using the
Biosignals is determined, wherein the compensation signal by means of
the first amplifier stage
is subtracted from the biosignal. A measuring system with a handset,
which is designed according to the above criteria, has a
Variety of advantages.
Difficulty with numerous biosignals, which of the human
or animal body
are more or less static potentials,
which due to electrolytic processes or the like
occur between the electrodes and the human tissues occur.
These potentials are referred to as DC potentials or DC offsets,
because they are compared to the useful signals only very slightly in time
They are typically below 1 Hz and are dependent on
Useful signal z.T. significantly lower (e.g., below 0.016 Hz). on the other hand
the potentials certainly have values that are a few orders of magnitude
to be taller
the useful levels. To provide an electronics, which in the
Being able to process these signals are therefore two fundamental
Approaches known. First, the supply voltage
in spite of the big ones
even to be able to reinforce the relatively small user level. After that, the DC offsets
by means of high passes
the signals are amplified further.
However, the problem here is that low-power small
and portable devices the lowest possible supply voltage
should have. Furthermore
the high-pass filters have a low cut-off frequency to the
Not also suppress useful signals, which in turn causes large components,
in particular capacities
which preclude miniaturization. It is therefore
a special advantage if the DC offsets already at the beginning of a
Signal processing suppressed
A known variant consists of several biosignals,
which also from the body
be tapped to form an average and this from the input signal
deducted. The mean or reference value represents a mean
DC offsets of the other electrodes. Accordingly, the
Reduce DC offset this way, but only in a limited way
because it varies from electrode to electrode. Try other methods again
provide a predictive value of the biosignal in a control loop,
which is subtracted at the entrance. The input signal is not only
reduced by any offsets, but also in the information content
reduced, so compressed. Depending on how the control loop (predication loop)
is set, different compression values of the
Reach biosignal. The disadvantage of this is the complicated structure
the control loops and the required reconstruction of the input signals
after compression or coding. The present invention provides
now a method ready, which is both easy to implement
is, as well as a good oppression
the DC offsets without any reconstruction of the biosignals
is required. This succeeds by having an input signal
is converted into a digital signal. Subsequently, will
a simple arithmetic operation on the digital value of the input signal
applied to determine a long term average and the thus obtained
Intermediate value is converted back into an analogue signal
and subtracted from the input signal. It is advantageous that the
digital intermediate value is already available in digital form and
can be used to analyze the signal. On the other hand will
only the DC offset of the input signal is determined to subtract from this
to become. This avoids reconstruction of the input signal.
Preferably, the calculator is a digital counter or a digital integrator.
This is a very easy way to offset the
Subtract the input signal. A big time constant, like here
needed to determine the offsets
is digitally much easier to realize than analog
Wise. This saves a lot of space.
It is therefore particularly preferred that a compensation signal is provided for the first amplifier stage, which takes into account the signal components in a low-frequency range, which are usually suppressed by high-pass filters after the preamplifier stage. Such a measure must take into account that the subtracted signal components do not lead to a later re construction of the signal must be made. The only reconstruction step that can advantageously still be used here is the consideration of the low-frequency component as the digital value of the arithmetic unit for the output signal during the evaluation of the signal. However, a further reduction of the data signal taking into account the entropy is advantageously not carried out.
is also preferred that for determining the compensation signal
in the handset an analog-to-digital converter is provided for digitizing
of the biosignal, an arithmetic unit for calculating an intermediate value
from the digitized biosignal and a digital-to-analog converter
to go back
of the intermediate value in the compensation signal.
Task is also according to the invention
solved a measuring system,
which is an analog gain stage
and a digital calculator includes and preferably as one
Integrated circuit is designed. The digital calculator
is typically a microcontroller or digital signal processor
(DSP). The microcontroller or the DSP takes in particular
an evaluation of the measurement signals before and can additionally the various functions
This concerns in particular a recognition of certain events,
which can be used for medical diagnosis. So
can e.g. the ECG signal with respect to
certain artifacts or deviations of the ST segment or ST elevation / subsidence
be evaluated. This also applies to the pulse or blood oxygen saturation.
The biosignals or the vital parameters are of the calculated values
the digital signal processor or the microcontroller e.g. to
filter limit values or the signals with pattern signals
to compare. Deviations from given values, which are reflected in
Are determined, and e.g. to a suitable location
reported. It is considered advantageous in such systems,
only the events that exceed a certain threshold or thresholds,
in one for that
store designated memory. This allows a significant data reduction
a continuous record of all accumulated data.
Continue to take over
the microcontroller unit or the digital signal processor the
Preparation of data for transmission
or the conversion into a format for a specific interface.
This can be a serial or a parallel interface, in particular
come here also wireless transmission methods,
like Bluetooth, Zigbee or similar
into consideration. In the long-term monitoring
It should be remembered that the energy consumption of radio transmission
is to be kept low. In particular, the data is from the digital
Signal processor transmitted to a base station or transmitted wirelessly,
which can carry out further processing steps on the data, such as
e.g. to decode them, to cache them and in a suitable
Format in further, e.g. in a network coupled evaluation
Handset of a measuring system according to the invention
preferably has a power consumption of less than 6mW per
Channel on (favorable values
also at 1 mW or 2mW per channel). It is important to take into account
that a not inconsiderable proportion of the total power consumption
the processing and transmission
the digital data is omitted,
and another big one
Share on the analog preamplification.
Especially with a transmission
By means of Bluetooth, the power consumption for the data transmission (RF part) one
critical factor. According to the invention, it has been recognized that a particular
the power consumption for
the analog amplification
and digitizing the data and digital post-processing
as well as the transmission
the data is present when the power consumption of a preprocessing channel
until the analog-to-digital conversion is less than 6 mW. Even lower values
can be achieved, with the input-related noise of a
Especially to consider
an advantageous embodiment of the present invention
in the handset functional units housed, which perform self-test functions
It is particularly preferred that the self-test functions
by means of a pseudo-noise source, a sine-wave generator or square-wave generator
can be triggered externally or internally in the handset, a self-test,
so either a pseudo noise on the different inputs or
also at specific critical or important points of signal processing
can be issued. Of particular interest are the
analog inputs to
Recording or amplification
the biosignals, but also the compensation input or the input of the
Analog to digital converter.
Likewise comes a self-test of the digital interfaces by means
a sine or square wave signal into consideration. Digital filters or
Detection functions for
certain events can
also be tested this way. It is also advantageous
for purposes of self-test in a memory a pattern of one to
expect expected useful signal, and this as a test signal for
Self-test at specific points of the handset to feed.
According to the invention, the measuring system has specific characteristics in order to be suitable for medical use. This includes an input-related inherent noise of a channel of less than 4.5 μVss (also <5 μVss) in a Frequenzbe In addition, the temporal correlation of the digital samples per channel to each other should not be worse than 0.5 msec. This value is based on a typical sampling rate for an ECG signal of 2 kHz. Normally, far lower sampling rates are used for ECG signals (eg 500 Hz or 1 kHz), but it is not possible to record the electrical pulse of a pacemaker. Therefore, it is particularly advantageous to work for ECG signals at a sampling rate of 2 kHz and above. In addition, for clinical applications it is necessary to offer an output signal of at least 12 bits with a sampling rate of 500 Hz. Advantageously, first sampled at 2 kHz and the sampling rate is lowered later to 500 Hz. As a result, the pulses of the pacemaker can first be recorded and, if necessary, filtered out, in order to lower the sampling rate. The ECG should be digitally resolved at the input with a resolution of at least 4.5 μV. This results in an input range of just under 20 mV for the useful signal. For applications in the context of electrocardiography, an input signal range for the useful signal of at least 10 mV is to be preferred. This makes it possible to record the ECG signal even with the usual fluctuations occurring and dissolve sufficiently.
For the purposes of the present application, a channel has a first amplification stage
and a second amplifier stage (post amplifier), a
Low pass (anti-aliasing low pass) on. The cutoff frequency of the low pass
is advantageously set up switchable. Hereby leave
the characteristics of the filter for
optimize different applications. Will the low pass filter with
fixed cut-off frequency implemented, divorced under certain circumstances
Applications because the useful frequency range does not expand later
On the other hand, low cutoff frequencies require implementation
an anti-aliasing low-pass large components or large areas
the implementation as an integrated circuit. It is therefore special
Analog-to-digital converter to use, or analog-to-digital converter
to use at a high sampling rate.
has the measuring system according to the invention
at least ten uniform
on. Such a channel number is for
a variety of application suitable. On the other hand, for long-term monitoring
a system with only 3-5
advantageous. There is a difference between the number of
Electrodes that are on the body
located, e.g. in the ECG or EEG measurement, and the number
of processing channels,
which ever an analog preamp and possibly a post-amplifier for
Provide processing of the signals and the analog-to-digital conversion,
which may be for several
can be done sequentially or by an analog-to-digital converter
per channel. Of course, according to the invention, not only bioelectrical
Signals, such as those of the ECG or the EEG per channel amplified and
but also signals of a completely different nature, such as
Pressure signals or similar ..
it is preferred that the preamplifiers
in a measuring system according to the present
Invention have two operational amplifiers connected as voltage followers,
which are coupled so that they have two high-impedance inputs to
Coupling a biosignal and a compensation signal and a
have another input for coupling a reference signal,
where the reinforcement
of the reference signal to the output is about 1 and the gain of the
first reference signal to the output is negative. This inventive arrangement
the conventional one
Instrumentation amplifier principle
the advantage that an operational amplifier compared to the usual instrument amplifier
three operational amplifiers
can be saved. This saves chip area in integrated solutions, though
also otherwise complexity,
and therefore costs. About that
In addition, the noise characteristics of the circuit are improved. Yet
The circuit offers two high-impedance inputs for the useful signal and the first
Reference signal. The first reference signal is preferably the
Output signal of the digital-to-analog converter, which the output value
of the digital integrator or calculation value in the feedback loop. by virtue of
The noise properties of the resistors used may be 1 kOhm
up to a maximum of 6 kOhm for
do not exceed certain applications (EEG) depending on position.
the entrance for
the second reference signal offered an input, which with the
Value 1 reinforced
becomes. As a result, the reference formation, usually as summation of
two or three electrode signals, much easier.
is particularly preferred that the measuring system according to the invention in the form of a
Plaster or similar
compact on the body
is attached. Here, the arrangement of the electrodes on the
given in a certain frame, what the use in particular
medical laymen at home.
Preferably, the digital-to-analog converter is constructed in the handset according to the R-2R principle, the inputs of which are designed to receive the digital intermediate values of the arithmetic unit and in response to the digital Zwi between at least three potentials are switched, wherein one of the potentials is selected taking into account the probability of occurrence level of the biosignal. The intermediate value is typically exactly the average potential between the positive supply voltage and ground and is referred to as virtual ground (VGND). The advantage of this embodiment is that frequently occurring values lie exactly in the virtual mass range. Conventional digital-to-analog converters according to the R-2R principle must always switch back and forth in this area between positive and negative supply voltage, as a result of which unwanted non-linearity of the converter occurs due to parasitic effects and non-idealities of the resistors. This can be avoided by the arrangement according to the invention. According to the invention, the aforementioned aspect can be refined such that the input level or the probability of occurrence of the intermediate values is taken into account so that the switching potentials of the R-2R digital-to-analog converter are favorably selected with respect to the statistically most frequently occurring values. With three possible potentials, this can lead to the medium potential not being chosen exactly in the middle between the other two potentials.
The present invention will be explained below with reference to the figures. It
1 a block diagram of an integrated circuit according to the present invention,
2 a second block diagram of a device according to the invention,
3 a simplified circuit diagram of a reference stage according to the present invention,
4 another block diagram of one for an analog input stage according to the present invention,
5 a simplified circuit diagram of a preamplifier according to the present invention,
6 a simplified block diagram of offset suppression according to the present invention,
7 a simplified block diagram of the digital part according to the present invention,
8th another simplified block diagram of the digital part according to the present invention,
9 a simplified circuit diagram of a digital-to-analog converter according to the present invention, and
10 a simplified circuit diagram of a modified preferred embodiment of a digital-to-analog converter according to the present invention.
1 shows a simplified block diagram of a device according to the invention. The device according to the invention comprises electronic circuits which are designed to advantageously process biosignals. For this purpose, an electronic device according to the invention, for example, 10 channels 1 - 10 on which each a biosignal 101 - 110 , in particular a bioelectric signal, can process. A variety of biosignals can be considered as signals, such as the signal of an ECG electrode, the signal of an EEG electrode, or the signals from electrodes that are used for evoked potentials (EP) as well as for EOG, EMG, VEP, SEP, ( B) AEB, ERP, EKP, ENG ... are used. The evoked potentials are another field of application of the electroencephalogram, whereby an electrical impulse (potential) in the nerves is stimulated (evoked) by a stimulus. For example, there are visually, acoustically or somatosensitively evoked potentials. The device according to the invention comprises in addition to the ten input channels 1 - 10 a digital part 12 and a reference level 11 , The digital part 12 serves to control the analog part by means of the signals 114 - 124 in particular to provide clock signals, the digital output data 113 - 123 of analog-to-digital converters, to filter and in a suitable data format at the interface 111 issue. The digital part 111 . 112 also provides the interface for communication with other electronic components. In addition, the digital part generates 12 test signals 113 - 123 and performs self-tests of parts of the circuit or the whole circuit. The reference level 11 provides a reference signal 304 (Sum signals from up to three input signals 301 . 302 . 303 ), as typically used in the processing of bioelectric signals. The channels 1 - 10 are coupled to the digital part in a variety of ways. In addition, the channels provide the input gins, to which, for example, electrodes or sensors or other coupling elements are connected in order to connect the corresponding sensors. The reference stage is also typically coupled directly or indirectly with sensors in order to process the signals generated there.
2 shows another, simplified block diagram of a channel 1 to 10 with more details. Accordingly, one of the channels points 1 to 10 according to a preferred embodiment, an amplifier stage 20 , an analog-digital wall ler 21 , and a digital integrator 22 , as well as a digital-to-analog converter 23 on. The analog amplifier stage 20 It serves the signals directly or indirectly provided by sensors 201 . 202 absorb and process further, in particular to reinforce as low noise. These signals are sent to the analog-to-digital converter after amplification 21 which converts the signals into digital signals. Here are basically two principles into consideration. The analog-to-digital converter shown here 21 can either be an analog-to-digital converter 21 be that for all channels or several of the channels 1 to 10 is used (multiplexing), or it is an analog-to-digital converter 21 that only signals the signals of a single channel 1 to 10 or a single analog preamplifier digitized. Accordingly, there are the analog-to-digital converter 21 different requirements. The selection of a suitable analog-to-digital converter depends on the requirements of the biosignals that are to be processed here. The characteristics of such an analog-to-digital converter 21 These include resolution, dynamic range, signal-to-noise ratio, linearity, clock rate, power consumption, and more. The digital output signals 601 of the analog-to-digital converter 21 be in the present here advantageous embodiment according to the invention in an integrator 22 summed up and as an intermediate value 603 output to the digital-to-analog converter, digital-to-analog converted and in this form as a compensation signal 200 returned to the entrance level. If the integrator 22 and the other components are suitably dimensioned, then the present configuration can easily compensate for spurious signals. Thus, the amplitudes of the input signal can be reduced by already in the input stage a portion of the signal is subtracted, which from the digital-to-analog converter 23 provided. The circuit thus meets a circumstance that occurs in many bioelectric signals. The mostly metallic electrodes, and also gel electrodes, which are used to pick up the bioelectrical signals from the skin of the person to be examined, generate a constant voltage due to electrolytic processes, the so-called DC offsets. These variables, which only change very slowly, reach amplitudes which are often much larger than the actual useful signals (biosignals) that are actually of interest: due to the DC offsets, the useful signals in the analogue amplifiers 20 not strengthened enough. Therefore, according to conventional technique. the input signal is only slightly amplified before the DC offsets are eliminated or strongly suppressed by means of analogue high-pass filters. However, if low-frequency components of the useful signal are also of interest, the cut-off frequencies of the analog high-pass filters used should be selected such that they do not unnecessarily impair the frequency spectrum of the useful signals. This leads to a very low cutoff frequency, which results in large components. In particular, the required large capacity usually prevent a small design of a corresponding electronics.
In the present case, it is proposed in an integrator 22 (or a counter) the digital output signals of the analog-to-digital converter 21 first add up, and then back into an analog signal in the digital-to-analog converter 23 to convert and the converted output signal back to the analog amplifier 20 supply. Furthermore, a filter stage 204 provided which certain unwanted signal components of the analog compensation signal 200 after digital-to-analog conversion in the digital-to-analog converter 23 suppressed. The control lines 24 . 25 . 26 . 27 and 28 serve the various control purposes in the blocks 20 . 21 . 22 and 23 , In particular, this is the setting of the preamplification in the analog preamplifier stage 20 , the setting of the cut-off frequency of a low-pass filtering, which is also in the analog preamplifier stage 20 takes place, the adjustment of the sampling rate in the analog-to-digital converter 21 by means of the control line 25 , the corresponding adaptation of the digital integrator to the different boundary conditions with the control line 24 Finally, the adaptation of the digital-to-analog converter to different requirements with the control line 28 , At the entrances 201 and 202 The analog preamplification stage is typically the input for sensor signals from electrodes, either directly or through a network or other preamplification and conversion stages of the analog preamplifier stage 20 and also reference signals as described below.
3 shows a simplified circuit diagram of a reference stage 11 according to the present invention. Reference stages in the illustrated form are commonly used to form the average of two or three signals picked up by the body. Accordingly, three electrode signals 301 . 302 . 303 on the preamplifier units 30 . 31 and 32 passed, which usually have the same gain. The resistors 33 . 34 and 35 are located at the outputs of the preamplifier stages 30 . 31 and 32 and form at their output a summing point in which the three input signals 301 . 302 and 303 are summed up. This output signal is optionally connected to a further amplification stage consisting of the operational amplifier 36 and two resistors 37 and 38 built up. This is the classic voltage follower, its gain by the resistors 37 and 38 is determined. For some applications, the accuracy of the gain factor is determined by means of the resistors 37 and 38 is set, critical. Therefore, such resistors are preferably not integrated, or at least arranged so that a subsequent calibration of the resistance ratio is possible. The sum signal is used as reference signal 304 output.
4 shows the simplified block diagram of an analog input stage 20 as they interact with other blocks in 2 is shown. Shown are a preamplifier 40 (first gain stage), a post amplifier 41 , a low pass filter (Anti Aliasing Filter) 42 and various control signals 43 . 44 . 45 . 46 as well as a shutdown signal (power down) 47 , The preamplifier may preferably choose between a gain of 20 and 80 be switched. The post amplifier has a gain of 1 or 4 depending on the setting of the control signal 45 on. The cutoff frequency of the anti-aliasing low-pass filter 42 can be changed between a value of 3 kHz and 15 kHz. With the aforementioned settings, to which the device according to the invention is not limited, optimal results can be achieved for numerous applications. These include, in particular, the electrocardiogram and the electroencephalogram as well as the evoked potentials in electroencephalography. The output signal 207 gets to the analog-to-digital converter 21 ,
5 shows a simplified circuit diagram of a first preamplifier stage 40 according to a preferred embodiment of the present invention. Shown are two operational amplifiers 50 and 51 , which at their positive inputs the biosignal (input signal U EL ), for example, of electrodes applied to the body and the compensation signal U DC to the inputs 501 and 502 receive. The preamplifiers are connected as a voltage follower, wherein in each case an adjustable resistance ratio of the resistors R52 to R53 and the resistors R54 to R55 is used. The operational amplifiers are biased, via an operating current, over the access lines with respect to an optimal operating point 56 and 57 is supplied, set. The Operationsver stronger work preferably after the "chopper principle". In this case, the input signal is inverted at a certain clock rate and reset, so that the polarity alternates with a high frequency. This leads to a shift of the useful signal spectrum in a higher frequency range. After passing through the input amplifier, the so-chopped input signal is transformed back again, ie correctly assembled again with a reversed clock. The transformation into the higher frequency range for the passage through the preamplifier stage allows a lower noise component in the useful signal spectrum of the input signal. Typically, frequencies above 30 kHz are used for chopping according to the "chopper principle". It is also preferable that the resistors R52 and R55 be set equal. Likewise, the resistors R53 and R54 are preferably selected to be equal. This results in a gain ratio with respect to the input 502 from 1 + (R52 / R53). At the reference entrance 503 is typically the reference signal associated with the reference stage according to 3 is created, created. This indicates with correctly adjusted resistance ratio with respect to the output 504 a gain of 1 on. The output voltage Ua at the output pin 504 then results in Uref + (U DC - U EL ) (1 + R52 / R53), when the electrode signal U EL (biosignal) on the Eingangsgin 502 the amplifier stage is supplied while that of the D / A converter 23 output compensation signal 200 (U DC ) via the electrode 501 to the preamplifier stage 40 arrives. By this measure, a further inversion of the compensation signal is unnecessary, and this is also deducted from the input signal.
6 shows an integrator stage according to one aspect of the present invention. The integrator 22 according to 2 can in the in 6 realized in a simplified representation reproduced form. The input signal 601 comes from the output of the analog-to-digital converter 21 and is in the multiplier 60 with one in a register 62 stored value multiplied. Subsequently, the values thus multiplied are sent to a summer 61 output and continuously summed there, the result of the summation being stored in a register. The one in the register 62 stored value can by means of a control signal 63 be set. Here are several principles into consideration, such as loading a fixed value or increasing a counter to the register value in the register 62 adapt. From a signal theory perspective, it is in the register 62 stored value by the time constant of the integrator. The signal 603 is the output signal of the summer 61 which according to 2 is output to the digital-to-analog converter.
7 shows a simplified block diagram of a digital part according to an aspect of the present invention. The blocks 70a to 70j are decimators, which preferably comprise test signal generators which are provided when an analog-digital converter according to the ΣΔ principle is used as the analog-to-digital converter. An analog-digital converter according to the ΣΔ principle operates on the principle of oversampling and is basically known. ΣΔ AD converters operate according to an over-sampling principle and therefore require digital filters (so-called decimators) which re-arrange the signal spectrum around the high-frequency noise component and increase the signal word width. The exits 701 - 710 the decimators 70a to 70j be connected to a digital block 71 which preferably prefers a synchronous serial interface 111 has, over which the device according to the invention can communicate with other electronic components. Decimators are also with a clock generation block 72 for providing control signals 711 for different oversampling values and sampling rates, as well as with a configuration block with I 2 C interface 112 coupled via which the configuration of the digital part can be made from the outside. Via control signals 713 Analog components are configured. The in the clock generation block 72 generated clock signals 711 . 712 are also output to those other blocks such as the analog-to-digital converters or the integrators of each channel.
8th shows an example of one of the decimators 70a to 70j in a simplified block diagram. This includes a digital sine and / or square wave generator 80 , a pseudo-noise source 81 , Selection circuits (multiplexer) 82 and 84 , as well as a decimator 83 , which operates preferably according to the module N-method or according to a multi-phase principle. The control signal 85 regulates the frequency of the digital sine wave generator and square wave generator 80 , The genes 80 and 81 are preferably used for self-test purposes or to calibrate the other circuit parts. The decimators 83 work with a variable decimator rate and are adaptable to different sampling rates. The decimator gives a digital signal 701 off and receives an output signal 801 a ΣΔ modulator.
9 shows a digital-to-analog converter 23 in a conventional R-2R arrangement. The resistors R are typically designed taking into account non-idealities with very small tolerances to each other. This is achieved, for example, by the use of so-called unit resistors R in integrated circuits, which all have a layout that is as identical as possible. This is a significant advantage of the R-2R digital-to-analog converter. In accordance with a digital data signal (intermediate signal 603 ), which from the integrator 22 to the digital-to-analog converter 23 is passed, the switches S1 to Sn are switched to either VSS or VDD. This results in the Ausgangsgin 200 an analog output signal corresponding to the digital input word. Disadvantageous in the 9 The arrangement shown is that in a middle range of digital intermediate values 603 (around the origin or midpoint between VDD and VSS) in which the digital signal is typically most non-linearity occurs. Is the digital signal changing? 603 according to a particular digital coding (2's complement) to a LSB (Lowest Significant Bit) from 0 to -1, so a switching operation of (almost) all switches S1 to Sn-1 and a switching operation of the switch Sn in reverse Direction required. This causes large nonlinear effects in the very area where the signal moves most frequently.
To remedy this disadvantage, an improved arrangement 23 according to 10 proposed. Here, the switches S1 to Sn are switched between three possible values. These are VDD, VSS and VGND (Virtual Mass). VGND is preferably the mean between VDD and VSS. In this arrangement, the change of an LSB around the center of the value range causes only a change in a switch S1 to Sn. The influences due to non-idealities of the components, in particular the resistors R, mismatch and fluctuations of the voltage supply VDD, VSS are therefore significantly lower. In particular, the nonlinearity in the range of interest for these applications is much lower and shifted to a range of numbers where signal levels rarely occur.
circuits described above, or essential parts thereof,
are preferably as integrated circuits on a semiconductor substrate