Note: Descriptions are shown in the official language in which they were submitted.
CA 02494917 2005-02-07
Air Purification Device
The invention relates to an air purification device for
reducing pollutants in the air, comprising an ioniser,
io which is exposed to an air flow and may be acted on, from a
driver stage, by ionisation power, wherein the air supplied
by the air flow may be ionised as a function of the
ionisation power, and comprising a gas sensor for measuring
pollutant concentrations.
is
It is in principle known to treat room air or breathing air
with what are known as ionisers, in order to reduce
pollutants. Pollutants and odorous substances usually form
complex and large molecules, which are broken down by the
20 ioniser into small molecular fragments. At the same time,
radicals, in particular oxygen radicals, form as a result
of the ionisation, and these radicals can then oxidise with
the broken-down fragments. The ioniser operates on the
basis of a controlled gas discharge that takes places
2s between two electrodes and a dielectric located
therebetween. The gas discharge is a barrier discharge,
the dielectric acting as a dielectric barrier. Individual
discharges that are limited with respect to time and are
preferably distributed homogeneously over the entire
3o electrode surface are thus attained. It is characteristic
of these barrier discharges that the transition into a
thermal arc discharge is prevented by the dielectric
barrier. The discharge is interrupted before the high-
energy electrons (1 - 10 eV) that arise during the striking
CA 02494917 2005-02-07
- 2 -
process release their energy to the surrounding gas as a
result of thermalisation.
In the household sector, in particular, various
s applications for an air purification device of this type
have been proposed in the past. For example, it is known
from DE 198 10 497 A1 to provide an air purification device
of this type in a toilet, in order to eliminate odours.
For this purpose, suitable suction devices comprising air
to ducts on the upper flushing rim of the WC-bowl or in a
hollow channel in the toilet seat lead the contaminated air
to the ioniser, in order to reduce the odour contamination.
One problem in the operation of the ioniser is that of
is activating the ioniser at an ionisation power suitable to
requirements. If the ioniser is acted on by too little
ionisation power, the ionisation is unsatisfactorily low,
whereas if ionisation is too high, too many ions and
radicals, which leave the operator with the impression of
2o the odour of a pungent corrosive or cleansing agent, are
sometimes released. In this operating state, in addition
to the formation of ions, there is also the production of
ozone, the excessive production of which is likewise
undesirable.
In order to solve this problem, WO 98/26482 describes an
air purification device comprising an ioniser, the supply
voltage of which is controlled via a gas sensor. The gas
sensor is a metal oxide semiconductor sensor, the
3o resistance of which decreases as the concentration of
specific gases (generally oxidisable gases or vapours, such
as hydrogen sulphide, hydrogen, ammonia, ethanol or carbon
monoxide, for example) increases: The variation in
resistance is thus a measure of the contamination of the
3s air with specific pollutants. According to WO 98/26482, as
CA 02494917 2005-02-07
- 3 -
the pollutant concentration rises, the ionisation power by
which the ioniser is acted on is increased in a sensor-
controlled manner up to a maximum value. In other words,
at a low pollutant concentration measured by the gas
s sensor, the ioniser is acted on by a correspondingly low
ionisation power, whereas at a high pollutant concentration
measured by the gas sensor, the ioniser is also activated
at a correspondingly high ionisation power. In order to
supplement this sensor control, WO 98/26482 also describes
to the use of an additional ionisation sensor and/or ozone
sensor. Since it is a prerequisite that the air quality
sensor in the sensor control measures the pollutant
concentration of the supplied air and is thus arranged, in
terms of the flow technology, upstream of the ioniser, the
is purpose of the additional ionisation sensor and/or ozone
sensor is to identify an ozone concentration, which is
still undesirable, in the purified air, in order then
optionally to correct the ionisation power as appropriate.
2o A sensor control corresponding to WO 98/26482 is also
described in DE 43 34 956 A1. DE 43 34 956 A1 proposes a
tin oxide gas sensor that detects the oxidisable room air
components. If this gas sensor identifies a relatively
high degree of room contamination, then the ioniser is also
25 activated at a relatively high ionisation power. The use
of a moisture sensor and a flow sensor is also proposed, in
order to increase the ionisation power even if a relatively
large volume of air or a relatively high air moisture
content is measured.
One drawback of the control methods known from WO 98/26482
and DE 43 34 956 A1 is the fact that the gas sensors used
have a limited measuring range and also a comparatively
slow reaction time. As a result of the limited measuring
range, sensor control of the ionisation power is not
CA 02494917 2005-02-07
- 4 -
possible in the peripheral regions of the measuring range.
If, for example, the pollutant concentration is below the
lowest measurement value of the gas sensor, the ioniser is
either switched off or the ionisation power continues to be
s operated at a predetermined minimum value. In the event of
rapidly varying pollutant concentrations, the slow reaction
time of the sensor also means that the ioniser is only
activated according to requirements after a certain delay.
In the elimination of odours in a toilet, for example, this
to delay is disadvantageous, since precisely in the event of a
precipitous increase in odorous substances, immediate
elimination of the odorous substances by means of the
ioniser is desirable.
i5 The object of the invention is therefore to provide an air
purification device that allows the air to be purified
according to requirements even if pollutant concentrations
vary rapidly and/or take on extreme values.
2o This object is achieved by an air purification device
having the features of Claim l and a method for reducing
pollutants having the features of Claim 18.
A fundamental feature of the invention consists in the fact
25 that the driver stage, the ioniser and the gas sensor
cooperate with a controller in a closed loop control
circuit such that the output signal of the gas sensor
substantially corresponds to a predetermined desired value.
In other words, whereas according to the prior art, a
3o sensor control is proposed wherein the sensor
characteristic curve extends as a function of the measured
pollutant concentration, the invention describes a
fundamentally different path. In accordance with the
invention, the gas sensor is only operated at a specific
3s operating point, which is determined by the desired value
CA 02494917 2005-02-07
- 5 -
of the control circuit. The gas sensor therefore always
supplies as an output signal a value that substantially
corresponds to the desired value, whereas the controller is
responsible for adjusting in the ioniser precisely that
ionisation power that maintains the output of the gas
sensor at said desired value.
Nevertheless, in order to achieve this aim, there has to be
a certain feedback between the gas sensor and the ioniser.
to However, the need for this feedback and the relation
between the feedback and the arrangement of the gas sensor
with respect to the air flow and with respect to the
ioniser were not previously recognised in the prior art
either. The arrangements of the gas sensor described in
i5 the prior art merely relate to arrangements that, in terms
of flow technology, are located upstream of the ioniser, so
that the control circuit effect in accordance with the
invention cannot occur.
2o In contrast to this, the invention is also based on the
recognition that the gas sensor is arranged with respect to
the air flow and with respect to the ioniser such that, in
an open loop control circuit, a variation in the output
signal of the gas sensor owing to a precipitous variation
2s in the pollutant concentration in the air supplied by the
air flow may be compensated by a variation in the
ionisation energy such that the output signal of the gas
sensor may be returned to its original value. The feedback
between the ioniser and gas sensor must therefore be
3o produced by means of the arrangement of the gas senor with
respect to the air flow and with respect to the ioniser
such that the effect of the ioniser and the effect of the
pollutant concentration contained in the air may be
superimposed on the gas sensor.
CA 02494917 2005-02-07
- 6 -
There is an open loop control circuit, in the terms of the
invention, if an electric feedback between the output
signal of the gas sensor and the controller is interrupted.
s There is a precipitous variation in the pollutant
concentration as a test function for the open loop control
circuit, in the terms of the invention, if, at a specific
moment, the pollutant concentration in the air supplied to
the ioniser by means of the air flow changes from a first
to constant value by a specific jump height to a second
constant value. In a practical experimental arrangement,
this means that optionally provided circulation air of the
air flow has to be interrupted, so that the pollutant
concentration in the air flow supplied to the ioniser
is remains, as a prerequisite, constant before and after the
precipitous variation in the pollutant concentration, and
is not additionally influenced by the air flow drawn off by
the ioniser.
2o Typical variations in the pollutant concentration are
preferably taken as a basis in the jump amplitude of the
precipitous variation in the pollutant concentration:
Typical variations in the pollutant concentration in the
air flow may be determined for the respective application
2s in that the expected variations in the pollutant
concentration are plotted on a histogram, according to
their estimated frequency. Any cases that fall within +/-
~ of a frequency maximum, for example, may be taken to
be typical. If, for example, the air purification device
3o is intended to reduce the odour of cigarette smoke in a
room, the expected air contamination resulting from
cigarette smoke relative to normal air contamination is
taken as the basis for the typical variation in the
pollutant concentration. In accordance with the invention,
35 the gas sensor must now be arranged with respect to the air
CA 02494917 2005-02-07
flow and with respect to the ioniser such that said
variation in the pollutant concentration in the air flow
may be compensated once more by a variation in the
ionisation energy, such that the output signal of the gas
s sensor may be returned to its original value, which, in the
example, corresponds to the original value of the normal
air contamination. Therefore, the greater the expected
influence of the variation in the pollutant concentration,
the closer the gas sensor also has to be arranged to the
io ioniser. If, on the other hand, only small variations in
the pollutant concentration are expected, then the gas
sensor should not be arranged too close to the ioniser, as
otherwise the output signal of the gas sensor may easily
enter the limitation. In any case, however, the gas sensor
is must maintain a specific minimum proximity from the ioniser
so that there is sufficient feedback between the ioniser
and gas sensor to compensate the variations in the
pollutant concentration that occur and thus to maintain the
output signal, in accordance with the invention, in the
2o region of a predetermined desired value.
A further recognition of the invention consists in the fact
that commercially available gas sensors for measuring
pollutant concentrations may be used as the measuring
2s element of the control loop. It has been found that even
an excessive production of ozone that is harmful for human
beings may thus be prevented by means of the ioniser, so
that the ionisation sensors or ozone sensors that are
otherwise used for this purpose are not strictly required.
In the method in accordance with the invention for reducing
pollutants in the air, the desired value is adjusted, using
the air purification device in accordance with the
invention, to a specific pollutant concentration, air
3s containing pollutants is supplied to the ioniser, and air
CA 02494917 2005-02-07
with a reduced pollutant content is drawn off from the
ioniser. According to a preferred embodiment, it is
provided that, in the circulation air mode, all or part of
the drawn-off air is fed back to the ioniser, in order to
s increase the efficiency of the air purification.
A fundamental advantage of the invention consists in the
fact that the efficiency of the air purification device is,
in principle, not limited by the measuring range of the gas
io sensor. Since the gas sensor is, in accordance with the
invention, operated at an operating point determined by the
desired value, even variations in the pollutant
concentration that exceed the measuring range of the gas
sensor may be treated by the air purification device. In
is the case of a conventional sensor control, in contrast, the
output signal of the gas sensor would enter the limitation
and would thus also limit the activation of the ioniser or
the driver stage. The limitations of the air purification
device are therefore, in principle, only conditioned by the
20 limitation of the ionisation power. The ionisation power
may, however, be additionally increased by means of
suitable measures, such as by connecting further ionisers
and/or blowers to increase the flow speed of the air flow,
for example. This opens up a broad range of possible
25 applications, from the household sector to the industrial
purification of large volumes of air, for the air
purification device in accordance with the invention.
A further advantage of the invention consists in the fact
ao that a suitable configuration of the controller facilitates
a transient response of the closed loop control circuit,
the transient time of which is below the time constant of
the gas sensor. This may be achieved, for example, by
means of a differential content in the controller, whereby
3s even in the case of small variations in the output signal
CA 02494917 2005-02-07
- 9 -
of the gas sensor, large control variables are produced in
the driver stage.
According to a preferred embodiment, it is provided that
s the driver stage comprises a high-voltage transformer, on
the secondary side of which an oscillating high voltage may
be generated. The ionisation power supplied to the ioniser
may be influenced primarily by the peak value of the
oscillating high voltage and/or by the pulsing of the
to oscillating high voltage. Preferably, the driver stage
comprises a circuit for pulse width modulation, with which
the high-voltage transformer may be activated on the
primary side and the peak value and/or the pulse ratio of
the secondary-side oscillating high voltage may be
is adjusted. In a series circuit comprising the high-voltage
transformer and resonator, which is supplied on the input
side with D.C. voltage, the pulse width-modulated signal
may be rectified and supplied to the input of the
resonator. The resonator, in turn, supplies an oscillating
2o voltage to the primary side of the high-voltage
transformer, so that the peak value on the secondary side
of the high-voltage transformer is thus proportional to the
pulse width ratio. In addition or alternatively, it may be
provided that the high voltage delivered on the secondary
2s side is pulsed. This means that the ioniser is only acted
on by a specific number of full waves before the
oscillating high voltage is then interrupted once more.
The ionisation power thus supplied in the medium is also
proportional to the pulse width ratio. The pulse width
ao ratio may be obtained from the same pulse width modulation
signal as applied at the input of the resonator, or else a
further pulse width modulation signal is generated for this
purpose.
CA 02494917 2005-02-07
- 10 -
According to a further preferred embodiment, it is provided
that the secondary-side oscillating high voltage may be
adjusted with a peak value in the range from 1 kV to 10 kV
and with a frequency in the range from 10 kHz to 50 kHz.
According to a preferred embodiment, the ioniser comprises
a glass tube, the inner wall of which is lined with a
perforated metal sheet as a first electrode and the outer
wall of which is surrounded by a wire mesh as a second
io electrode, the oscillating high voltage of the driver stage
being applied between the first electrode and the second
electrode. In order to disinfect or purify the gas flowing
around the ionisation tube, the high-voltage transformer is
activated such that, in the event of a gas discharge,
i5 radicals, preferably oxygen radicals, are generated. At a
peak value of 1 to 10 kV, the high-voltage transformer is
conventionally operated at an A.C. voltage in the range
from approx. 10 kHz to 50 kHz, preferably in the range from
kHz to 30 kHz. If a gas flows around an ionisation tube
20 of this type, a gas discharge, which results in an
ionisation of the flowing gas, takes place. The gas
discharge is a barrier discharge, which takes place by
means of the glass tube acting as a dielectric barrier.
Individual discharges that are limited with respect to time
2s and are preferably distributed homogeneously over the
entire electrode surface are thus attained. It is
characteristic of these barrier discharges that the
transition into a thermal arc discharge is prevented by the
dielectric barrier. The discharge is interrupted before
3o the high-energy electrons (1 - 10 eV) that arise during the
striking process release their energy to the surrounding
gas as a result of thermalisation. Alternatively, any
other configuration of the ioniser is, of course,
conceivable, such as a tabular arrangement, for example, or
CA 02494917 2005-02-07
- 11 -
else a combination of a tube arrangement and a tabular
arrangement.
According to a preferred embodiment, it is also provided
s that the gas sensor comprises a metal oxide sensor, the
resistance of which varies in the event of reactions with
gases. The metal oxide is applied to a substrate, which is
maintained at a predetermined temperature using a heating
element. Preferably, a gas sensor, which does not exhibit
to any variation in resistance relative to the varying
pollutant concentration in the air, is used. It has been
found that particularly reliable control of the pollutant
concentration is possible using gas sensors of this type.
The metal oxide may, for example, comprise tin oxide.
According to a further preferred embodiment, it is provided
that, with respect to the air flowing around the ioniser,
the air inlet opening in the gas sensor is at a distance of
approx. 0.5 cm to 5.0 cm, preferably approx. 1.0 cm to 2.0
2o cm, from the surface of the ioniser. It has been found
that at these distances, the modulation range of the gas
sensor may usually be reconciled with the modulation range
of the ioniser and the value range of conventional
pollutant concentrations.
According to a further preferred embodiment, it is provided
that the desired value may be manually adjusted on the
device. The operator is thus able, in a normal pollutant
concentration of the air, to determine the operating mode
of the device that he finds most comfortable. The
arrangement of the gas sensor is particular preferably
selected such that the predetermined desired value
corresponds to a central range with respect to the total
modulation range of the output signal of the gas sensor.
ss Since, in other words, the control circuit ensures, in
CA 02494917 2005-02-07
- 12 -
accordance with the invention, that the pollutant
concentration measured by the gas sensor substantially
corresponds to the desired value, the gas sensor is thus
operated in a range that allows maximum modulation in the
s transient process of the closed loop control circuit.
According to a further preferred embodiment, it is provided
that the air flow is generated by means of convection,
which, in the case of small domestic devices, for example,
io may result from the heating of air supplied to the
electrical components of the device.
According to a further preferred embodiment, a ventilator
is provided for generating the air flow. It has been
i5 recognised that the air flow may also have an influence on
the functioning of the control circuit. If, for example,
the gas sensor is located, on the flow side, upstream of
the ioniser, then there is less coupling between the
ioniser and gas sensor, at the same distance of the gas
2o sensor from the surface of the ioniser, in comparison to an
arrangement in which the gas sensor is arranged, on the
flow side, downstream of the ioniser.
According to a further preferred embodiment, it is
2s therefore provided that an additional controller
additionally controls the flow rate of the air flow such
that the output signal of the gas sensor substantially
corresponds to a predetermined desired value. In
particular, it has proven expedient that the additiona l
3o controller is connected as soon as a limitation occurs in
the control circuit comprising the ioniser, driver stage,
gas sensor and controller. In this case, the additional
controller must act to allow the limitation that has
occurred to be suitably compensated.
CA 02494917 2005-02-07
- 13 -
The functioning of the control circuit is, of course, to a
large extent dependent on what type of controller is used.
Once the transfer response of the remaining control circuit
elements, i.e. the ioniser, the driver stage and the gas
s sensor, has been determined using suitable identification
methods, the controller may, in principle, be configured
according to the available control engineering methods.
The most obvious examples of convention control circuit
elements are a P-controller, a PI-controller or a PID-
lo controller. The simplest of these is the P-controller;
however, in principle, this requires a control deviation
between the predetermined desired value and the pollutant
concentration measured by the gas sensor, in order to be
able to issue a control variable. Nevertheless, if the
15 amplification factor of the P-controller is selected high
enough, the control deviation may be disregarded. However,
a high amplification factor of the P-controller is only
permissible if there is still sufficient signal/noise
distance in the output signal of the gas sensor. If, on
2o the other hand, the signal-noise distance in the output
signal of the gas sensor is no longer sufficient for the
use of a P-controller, a PI-controller may be used. On
account of its integrative response, the PI-controller is
able to supply a continuous control variable even if the
2s control deviation disappears. In other words, if a PI-
controller is used, the disappearance of the control
deviation may, in principle, be attained in the case of a
transient control circuit. In order to accelerate the
transient response of the control circuit, a differential
3o element is conventionally attached to the PI controller,
thus producing a PID-controller. In the event of rapid
variations in the pollutant concentration or the desired
value, the differential response of the PID-controller may
cause limitations to appear in the control circuit
3s elements. In this case, it is advantageous to provide the
CA 02494917 2005-02-07
- 14 -
above-mentioned connection of an additional controller for
the flow rate. If the ionisation power of the ioniser
therefore reaches the upper limit, the additional
controller may instead provide an increase in the flow rate
s of the air flow.
In addition to the conventional P-controller, PI-controller
and PID-controller types, other controllers, such as a
rule-based fuzzy controller or a state controller, for
to example, may, of course, also be provided. A rule-based
fuzzy controller or a state controller is particularly
suitable if, in addition to the measured pollutant
concentration, further measurement variables are to be
processed by the controller. In principle, it is
is conceivable to improve the control circuit response using
additional sensors, such as a moisture sensor and/or an
ionisation sensor and/or an ozone sensor.
According to a further preferred embodiment, it is provided
2o that a calibration element calibrates the gas sensor to the
desired value if a pollutant concentration corresponding to
the desired value is supplied to the gas sensor.
Preferably, the ionisation energy supplied to the ioniser
is disconnected during the calibration of the gas sensor,
zs in order to prevent disruptive reactions of the ioniser in
the calibrating operation. Alternatively, however, the
ioniser may also be activated in the calibrating operation
with a predetermined continuous ionisation power, at which
the ioniser may be operated in order to constantly maintain
3o minimum pleasant room air conditions.
The tolerances of a gas sensor, which are conditioned by
the manufacturing process, may be compensated by
calibrating the gas sensor. When the above-mentioned tin
35 oxide gas sensors were used, it was noted that the
CA 02494917 2005-02-07
- 15 -
tolerances substantially lead to an absolute displacement
of the characteristic curve, whereas the relative variation
in the sensor signal as a function of the gas concentration
remains approximately the same in all of the gas sensors.
s In this case, the calibration element may comprise a simple
adder, which in the calibrating operation adds a
corresponding voltage to the output voltage of the gas
sensor. In this case, it is also necessary that a
pollutant concentration that is specified by the user to be
to "clean air" is supplied to the gas sensor during the
calibrating operation. The aim is to determine, in the
calibrating operation by means of the calibration element,
the additional voltage that is required to make the control
deviation approximately zero .
The invention will be described below in greater detail
using various embodiments, with reference to the
accompanying drawings, in which;
zo Fig. la shows a block diagram of the transfer response of
a gas sensor with a jump function as the input;
Fig. lb shows the response function at the output 150
with a jump amplitude of 1;
Fig. lc shows the response function at the output 150
with a jump amplitude of 2.5;
Fig. 2a shows a block diagram of the transfer response of
3o a sensor control with a jump function as the
input
Fig. 2b shows the response function at the output 250
with a jump amplitude of 1;
CA 02494917 2005-02-07
- 16 -
Fig. 2c shows the response function at the output 250
with a jump amplitude of 2.5;
Fig. 3a shows a block diagram of the transfer response of
s an open loop control circuit with a jump function
of the pollutant concentration;
Fig. 3b shows the response function at the output 350
with a jump amplitude of 1:
io
Fig. 4a shows a block diagram of the transfer response of
an open loop control circuit with a jump function
of the ionisation power;
i5 Fig. 4b shows the response function at the output 450
with a jump amplitude of 1;
Fig. 4c shows the response function at the output 450
with a jump amplitude of -1;
Fig. 5a shows a block diagram of the signal flow of a
closed loop control circuit;
Fig. 5b shows a block diagram of the transfer response of
2s a closed loop control circuit with a jump
function of the pollutant concentration;
Fig. 5c shows the response functions at the outputs 550
and 551 with a jump amplitude of 1;
Fig. 6a shows a block diagram of the transfer response of
a closed loop control circuit with a jump
function of the desired value and a subsequent
jump function of the pollutant concentration;
CA 02494917 2005-02-07
- 17 -
Fig. 6b shows the response functions at the outputs 650
and 651 with jump amplitudes, in each case, of 1:
Fig. 7 shows the sensitivity characteristics of a tin
s oxide gas sensor;
Fig. 8 shows a perspective .illustration of an air
purification device in accordance with the
invention;
io
Fig. 9 shows a block diagram of the air purification
device in accordance with the invention according
to Fig. 8; and
15 Fig. 10 shows a flow diagram of the control algorithm of
the controller from Fig. 9.
Fig. 1a shows a block diagram of the transfer response of a
gas sensor with a jump function as the input. The series
2o connection of two PT1-elements 111, 112 and a limitation
element 113 was accordingly taken as a model for the
transfer response of a gas sensor 110. The input function
is a precipitous increase in the pollutant concentration
101, wherein the corresponding response function may be
25 traced at the output 150. The following parameters were
taken as a basis:
PT1-elements 111, 112: time constant = 10.0 sec, transfer
value = 1.0
Limitation 113: Upper limit = 2.0, lower limit = -2.0
It was thus assumed that the output signal of the gas
sensor may be activated in a range from -2.0 volts to 2.0
vOltS.
CA 02494917 2005-02-07
- 18 -
Fig. 1 shows the response function at the output 150 with a
jump amplitude of 1. The gas sensor thus responds, as
expected, in a delayed manner to a jump function of the
s pollutant concentration and, after approx. 60 seconds,
exponentially approximates the jump amplitude of 1.
Fig. lc shows the response function at the output 150 with
a jump amplitude of 2.5. Once the value 2.0 has been
to reached, the limitation 113 enters into effect so that,
after approx. 30 seconds, the response function remains
constant at the value 2.0 and cannot further approximate
the jump amplitude of 2.5.
is Fig. 2a shows a block diagram of the transfer response of a
sensor control with a jump function as the input. The
basic structure of a sensor control according to the prior
art, such as that according to WO 98/26482 or according to
DE 43 34 956 A1, for example, comprises a gas sensor 210
2o with a subsequent driver stage 220. As in Fig. la, the gas
sensor 210 comprises two PT1-elements 211, 222 and a
limitation 213, the parameters also corresponding to those
of Fig. 1a. A P-element 221 with a limitation 222
connected downstream was taken as a basis for the model of
2s the driver stage 220. The following parameters were
assumed:
P-element 221: transfer coefficient = 250.0
so Limiter 222: upper limit = 500 V, lower limit = -500.0 V
This means that, according to Fig. 2a, the output voltage
of the gas sensor 210 is converted by means of the driver
stage 220 with the factor 250 into a high voltage:
as however, for the sake of simplicity, the offsets that occur
CA 02494917 2005-02-07
- 19 -
in practice were not taken into account. Conventional
output voltages of a gas sensor connected in a voltage
divider are, for example, in the range from 1 V to 5 V and
are converted by means of the driver stage into a high
s voltage of, for example, 1000 V to 2000 V. However, for
the model of the control circuit, these offsets are of no
further importance and may easily be attached at any time,
if required.
to In order to examine the transfer response of the sensor
control according to Fig. 2a, it was, in turn, assumed that
there is at the input a precipitous increase in pollutant
concentration 201, which is recorded at the output 250 of
the driver stage 220.
is
Fig. 2b shows the response function at the output 250 with
a jump amplitude of 1. In order also to be able to
illustrate the jump amplitude in Fig. 2b, it was, however,
enlarged by the factor 250. As expected, according to Fig.
20 2b, the same response function appears as in Fig. lb, but
in this case extended by the factor 250, as a result of the
driver stage 220 connected downstream.
Finally, Fig. 2c shows the response function at the output
zs 250 with a jump amplitude of 2.5, the jump amplitude having
once more been enlarged, for reasons of illustration, by
the factor 250. As a result of the increased jump
amplitude of 2.5, the limitations 213 and 222,
respectively, enter into effect, so that, after approx. 30
3o seconds, the response function according to Fig. 2c remains
constant at 500 V.
The illustrated transfer response according to Fig. 2a,
Fig. 2b and 2c corresponds substantially to known sensor
35 controls for air purification devices comprising ionisers.
CA 02494917 2005-02-07
- 20 -
In contrast, the invention proposes the construction of a
closed loop control circuit, wherein the effects of the
pollutant concentration and the air ionisation from the
ioniser are superimposed and compensated on the pollutant.
s sensor. A block diagram of the signal flow of a control
circuit closed in this manner is illustrated in Fig. 5a and
will be explained below in greater detail. In order to
analyse individual components of the control circuit, a
block diagram of the transfer response of an open loop
to control circuit with a jump function of the pollutant
concentration is illustrated in Fig. 3a.
The basic construction of the open loop control circuit
according to Fig. 3a comprises a controller 340, a driver
is stage 320 connected downstream and the subsequent ioniser
330. In accordance with the invention, the effects of the
ioniser 330 and the pollutants contained in the air flow
are now to be superimposed at the input of the gas sensor
310. The model for this circumstance is formed, in the
2o block diagram according to Fig. 3a, by means of the
summation point 303, on which both a jump function of the
pollutant concentration 301 and, via the transfer path 332,
the ioniser 330 exert an influence. The parameters of the
gas sensor 310 are identical to the parameters indicated in
2s Fig. la. Since, according to Fig. 3a, the response of the
gas sensor is initially to be viewed in isolation, in the
case of a jump function of the pollutant concentration, the
parameters of the remaining control circuit elements are,
for the time being, insignificant, and will therefore only
3o be explained at the appropriate point in the following
figures.
Fig. 3b shows the response function at the output 350 with
a jump amplitude 1. Since, according to Fig. 3a, an open
as loop circuit was taken to be a prerequisite, the response
CA 02494917 2005-02-07
- 21 -
function according to Fig. 3b results exclusively from the
precipitous variation in the pollutant concentration and
therefore corresponds to the response function according to
Fig. lb.
s
Fig. 4a shows a block diagram of the transfer response of
an open loop control circuit with a jump function of the
ionisation power. As in Fig. 3a, the open loop control
circuit comprises, once more, a controller 440, a driver
io stage 420, an ioniser 420 and a gas sensor 410. In this
case, only the ioniser 430 has an effect on the summation
point 403, without a further additional influence from the
pollutant concentration, which is now kept constant in the
air flow supplied to the ioniser.
is
In order to examine a jump function of the ionisation power
in the block diagram according to Fig. 4a, the summation
point 405, on which the jump function 404 exerts an
influence, was inserted between the controller 440 and the
2o driver stage 420. The parameters of the blocks 411, 412,
913 of the gas sensor 910 are identical to the parameters
of the gas sensor 110 according to Fig. 1a. The parameters
of the blocks 421, 422 of the driver stage 420 are also
identical to the parameters of the driver stage 220
2s according to Fig. 2a. The ioniser 430 was modelled by
means of a simple P-element 431 having the following
parameter:
P-element 431: transfer coefficient = -0.004
The output of the ioniser directly exerts an influence on
the summation point 403, via the path 432, without any
delay. It was therefore assumed, in this case, that the
gas sensor 410 is arranged in immediate proximity to the
3s ioniser 430. In the event of a greater distance between
CA 02494917 2005-02-07
- 22 -
the ioniser 430 and gas sensor 410, a dead time element,
for example, may be inserted on the path 432. The transfer
response of the P-element of 431 therefore corresponds to a
conversion of the variation in high voltage at the output
s of the driver stage 420 into a variation in the pollutant
concentration to be measured by the gas sensor 410.
Fig. 4b shows the response function at the output 450 with
a jump amplitude of 1. An increase in the input voltage at
io the driver stage 420 by 1 volt therefore results in a
decrease, also of 1 volt, of the output voltage of the gas
sensor, the time function, in this case, resulting, once
more, from the transfer response of the two PT1-elements
412, 413. The opposing response may be explained in that
is an increase in the ionisation power is accompanied by a
reduction of pollutants in the air flow. Accordingly, Fig.
4c shows the response function at the output 950 with a
jump amplitude of -1. An opposing response may also be
identified in this case, since a decrease in the ionisation
ao power results in an increase in the pollutant concentration
in the air flow.
The measurements on the open loop control circuit according
to Fig. 3a, Fig. 3b and Fig. 4a, Fig. 4b and Fig. 4c,
2s respectively, indicate how the arrangement in accordance
with the invention of the gas sensor with respect to the
gas flow and with respect to the ioniser may easily be
identified. Fig. 3b shows the output signal of the gas
sensor, in the case of an open loop control circuit, on the
3o basis of a variation in the pollutant concentration in the
air flow. As a result of this variation, the output signal
on the gas sensor rises from 0 V to 1 V.
In accordance with the invention, the gas sensor now has to
3s be arranged with respect to the air flow and with respect
CA 02494917 2005-02-07
- 23 -
to the ioniser such that, in an open loop control circuit,
this variation may be compensated by a variation in the
ionisation energy such that the output signal of the gas
sensor may be returned to its original value. Fig. 4b
s shows the output signal of the gas sensor in an open loop
control circuit, in the event of a variation in the
ionisation energy and, at the same time, a constant
pollutant concentration in the air flow supplied to the
ioniser. In this case, the output signal of the gas sensor
l0 450 changes from 0 V to -1 V if the voltage of 1 V is
increased at the input of the driver stage. The
arrangement of the gas sensor with respect to the air flow
and with respect to the ioniser that is simulated in this
case therefore corresponds precisely to the desired effect,
is such that the variation, illustrated in Fig. 3b, in the
output signal of the gas sensor, on account of a
corresponding variation in the ionisation energy according
to Fig. 4b, may be compensated. In practice, experiments
corresponding to Fig. 3a and Fig. 4a may be carried out in
20 order to verify said compensation effect on the open loop
control circuit.
The response of the closed loop control circuit will now be
explained in greater detail. For this purpose, Fig. 5a
zs shows, in the first place, a block diagram of the main
signal flow of the closed loop control circuit. The closed
loop control circuit comprises the above-described control
circuit elements, i.e. a gas sensor 510, a controller 540,
a driver stage 520 and an ioniser 530. The driver stage
30 520, for its part, comprises a voltage source 525, a pulse
width modulator 526, a resonator 527 and a high-voltage
transformer 528.
A D.C. voltage supplied by the voltage source 525 is
ss converted by the pulse width modulator 526 into pulses
CA 02494917 2005-02-07
- 24 -
exhibiting a pulse width ratio determined by the controller
540 and a clock rate determined by a clock generator (not
shown in greater detail). In the event of smoothing of
these pulses, a D.C. voltage is produced that is
s proportional to the pulse width ratio and is supplied to a
resonator 527. The resonator 527 is wired to the
subsequent high-voltage transformer 528 such that, on the
one hand, when a D.C. voltage is supplied, it automatically
starts to oscillate at a working frequency in the range
to from approx. 25 kHz to 35 kHz and, on the other hand, it
supplies a secondary-side oscillating high voltage, the
peak value of which is approximately proportional to the
input voltage of the resonator 527 or to the adjusted pulse
width of the pulse width modulator 526. The oscillating
is high voltage, supplied by the high-frequency transformer
528, exhibiting peak values in the range from l.O kV to 2.0
kV, for example, is applied to the two electrodes of the
ioniser 530.
2o The air 500 to be purified flows around the ioniser 530,
the gas sensor 510 being arranged, on the flow side,
downstream of the ionisation tube 530. In the case of the
closed loop control circuit, all or part of the air flow
may be fed back using the circulation air mode. The gas
2s sensor 510 supplies its output signal to the controller
540, which carries out a desired/actual value comparison on
the basis of the desired value 547 and adjusts the pulse
width ratio of the pulse width modulator 526 in accordance
with the basic control algorithm.
Fig. 5b shows a block diagram of the transfer response of a
closed loop control circuit with a jump function of the
pollutant concentration. The closed loop control circuit
according to Fig. 5b is developed from the open loop
3s control circuit according to Fig. 3a in that the output
CA 02494917 2005-02-07
- 25 -
signal 550 of the gas sensor is fed back to the controller
540 via the branch 514. The blocks of the gas sensor 510,
of the driver stage 520 and of the ioniser 530 with the
associated parameters are identical to the indicated
s parameters of the gas sensor 310 according to Fig. 3a or
the driver stage 420 and the ioniser 430 according to Fig.
4a, so that reference may be made, in this regard, to the
description according to Fig. 3a and according to Fig. 4a.
to The construction of the controller 540 will now be
described in detail. The desired value 547 is guided in
the controller to the subtraction point 546. The control
deviation thus determined reaches the subsequent PID-
controller via the P-element 541. The PID-controller
is comprises, in turn, a P-element 542, a DTl-element 543 and
an I-element 544, the outputs of which are integrated with
the summation point 545 to form the output 551. The output
551 supplies the control variable, which is used as the
input for the driver stage 520. The parameters of the
2o controller 540 were defined as follows:
Desired value 547: desired value = 0
P-element 541: transfer coefficient = -1
P-element 542: transfer coefficient = 2
DTl-element 543: transfer coefficient = 8, time constant =
2 secs.
I-controller 544: transfer coefficient = 0.21/second,
corresponding to an integration constant of 5 secs.
The closed control loop response is now examined with
3s reference to the jump function 501, which corresponds to a
CA 02494917 2005-02-07
- 26 -
precipitous variation in the pollutant concentration in the
air flow. In this case, the time signals are illustrated
at the output of the gas sensor 550 and at the output of
the controller 551.
Fig. 5c shows the response functions at the outputs 550 and
551 with a jump amplitude of 1.
It is clear from the output signal of the gas sensor 550
to that, despite a precipitous variation in the pollutant
concentration, the control circuit is able to return the
output signal 550 to the desired value 547. Once the
output signal has been increased to approx. 0.25, after
approx. 40 seconds, the output signal reaches its original
is value once again and then, within a further 40 seconds,
approximates the desired value once more, with a small
overshoot. The output variable 551 of the controller 540,
on the other hand, ensures that the driver stage 520 is
acted on by an adequate input variable, so that the
2o variation in the pollutant concentration that has occurred
may be compensated at the summation point 503. After
approx. 25 seconds, the control variable 551 has reached
its maximum value and from then on approximates the end
value 1.0, which corresponds to an input voltage of 1.0 V
2s at the input of the driver stage 520. It may be inferred
from Fig. 5c that the transient response of the closed loop
control circuit is substantially determined by the time
response of the gas sensor 510, provided that no additional
delays occur on the path 532 between the ioniser 530 and
3o gas sensor 510. The time constant of the gas sensor may be
determined using an arrangement as shown in Fig. la. The
time constant of the recorded jump function 150 corresponds
approximately to the time in which the jump function 150
reached the value (1 - 1/e), if it is assumed that the
CA 02494917 2005-02-07
_ 27 _
total transfer response of the gas sensor is approximated
by means of an individual PT1-element.
If, on the other hand, the path 532 between the ioniser 530
s and gas sensor 510 were to exhibit a delay (as a result of
the flow rate of the air flow, for example, if the gas
sensor is arranged at a distance from the ioniser), a
secondary condition may be established for this delay time,
in order not to slow down the transient response of the
to closed loop control circuit unnecessarily. It may
accordingly be stipulated as a secondary condition that, in
an open loop control circuit and at a constant pollutant
concentration, the delay time of the output signal of the
gas sensor is to be below the above-defined time constant
is of the gas sensor, in the event of a variation in the
ionisation energy. In the present case, the time constant
of the gas sensor 510 may be determined from the time
function according to Fig. lb to be approx. 20 seconds. In
order to optimise, with respect to time, the transient
2o response of the closed loop control circuit, the gas sensor
should thus fulfil the additional secondary condition, with
respect to the air flow and with respect to the ioniser,
such that the delay time of the path 532 is also below 20
seconds. Generally, this secondary condition is easy to
2s fulfil, in that the gas sensor is arranged suitably close
to the ioniser.
Fig. 6a shows a block diagram of the transfer response of a
closed loop control circuit with a jump function of the
3o desired value and a subsequent jump function of the
pollutant concentration. The only difference between the
block diagram according to Fig. 6a and the block diagram
according to Fig. 5b is that a jump function 648 is now the
desired value and the precipitous variation in the
3s pollutant concentration 601 only takes place after a
CA 02494917 2005-02-07
- 28 -
certain dead time 602. 100 secs, were taken as a parameter
for the dead time. Otherwise, the block diagram according
to Fig. 6a corresponds to the block diagram according to
Fig. 5b, so that, as far as the remaining components are
s concerned, reference may be made to that part of the
description.
The closed loop circuit according to Fig. 6a is therefore
first of all acted on by a variation in the desired value
io 648 and is then, once the dead time 602 has expired,
additionally acted on by a variation in the pollutant
concentration 601. In Fig. 6b, the corresponding response
functions are illustrated at the outputs 650 and 651. The
dot-dash line at value 2 also indicates the limitation,
is which corresponds to the limitation of the driver stage
620, taking into account the transfer coefficient of the P-
element 621.
Owing to the differential content 643 of the cohtroller
20 640, the precipitous increase in the desired value 648
initially results in a high control variable 651. After 60
seconds, the control circuit has then built up to the new
desired value, so that there is now the output signal with
the value -1.0 at the output 650 of the gas sensor. After
2s 100 seconds, there is then an additional cut-in of the
precipitous variation in the pollutant concentration, after
which the control variable 651 rises once more, in order
this time to maintain the output signal 650 of the gas
sensor at the value -1. What is revealing, in this case,
3o is the interpretation of the regions 623 and 624. As a
result of the limitation 622 of the driver stage 620, the
control variables above the value 2.0 or below the value -
2.0 may not be transmitted to the ioniser 630. It is
therefore expedient, as stated above, to provide additional
as measures in these regions, in order to supply a higher
CA 02494917 2005-02-07
_ 29 _
ionisation power, by connecting an additional blower and/or
by connecting further ionisers, for example.
Fig. 7 shows the sensitivity characteristics of a tin oxide
s gas sensor. The diagram plots the relative resistance
variation, based on air, of the tin oxide element as a
function of the pollutant concentration of various
pollutants. As the line 701 shows, the tin oxide gas
sensor is insensitive to air or oxygen. However, as the
to pollutant concentration increases, the sensor exhibits
marked sensitivity to HZS, hydrogen, ammonia, ethanol and
C0. For the household sector, it has been found that
stable control may, in particular, be attained if the
control is adjusted to the sensitivity curve 702 of CO.
is
Fig. 8 shows a perspective illustration of an air
purification device in accordance with the invention. The
air purification device 801 is configured as a table device
with a pedestal 802 and a cover 803. An ionisation tube
20 804, which is constructed in the above-described manner, is
fastened to the pedestal, as the ioniser. A gas sensor
805, which, in accordance with the invention, is arranged
with respect to the ioniser 804 such that, in an open loop
control circuit, a variation in the output signal of the
2s gas sensor owing to a precipitous variation in the
pollutant concentration in the air supplied by the air flow
may be compensated by a variation in the ionisation energy,
is also fastened to the pedestal. The air flow enters and
leaves the housing through the air slits 806 formed in the
ao cover 803. A suitable ventilator may also be provided on
the pedestal 802 or even outside the device in order to
assist the air flow. An LED display 807 and an operating
potentiometer 808, for operating the device, and an
electrical supply line 809, for supplying power, are
35 provided at the edge of the pedestal.
CA 02494917 2005-02-07
- 30 -
The function of the air purification device 801 will be
explained with reference to Fig. 9, which shows a block
diagram of the air purification device in accordance with
s the invention according to Fig. 8. The calibrating
operation, in which the gas sensor is calibrated to a
predetermined pollutant concentration, will be described
first of all. This calibration is generally necessary
because commercially available gas sensors exhibit various
io characteristic curves and would thus elicit different
control circuit responses. However, with the use of tin
oxide gas sensors, it has been noted that the relative
variation in the output signal of the gas sensor, in the
event of the gas concentration varying, is almost constant,
is and that only an absolute displacement of the output
signal, at a given gas concentration, between various gas
sensors may be observed. In the case of the control in
accordance with the invention, the fact that the sensor is
only operated in a small operating range anyway; so that
2o the sensor characteristic curve may be linearised around
this operating range once the operating point has been
calibrated, may also be utilised.
For the calibrating operation, the change-over switch 901
zs is initially brought into the position l, so that the
ionisation tube 904 is not acted on by ionisation power.
Instead, the control deviation is supplied to the
calibration element 912. A constant pollutant
concentration, which, depending on the respective
ao application, corresponds to "clean air", and hence to the
targeted desired value, is then introduced into the air
flow 906. On the device, the operating potentiometer 808
is brought into the targeted desired value position, so
that the desired value 908 thus adjusted is at the
ss comparison point 909. If calibration has not yet taken
CA 02494917 2005-02-07
- 31 -
place, a control deviation 910 will then be observed at the
output of the comparison element. The adding element 911
and the calibration element 912 are now also provided for
the purposes of calibration. The calibration element 912
s receives the control deviation 910 as an input from the
change-over switch 901, and then increases or decreases the
output voltage 913 such that the control deviation 910 is
set to zero. The voltage level 913 thus determined may,
for example, be stored in a memory, so that it is still
to available after a power failure. This type of calibration
may optionally be repeated a plurality of times, wherein
varying pollutant concentrations 906 may also be taken into
account.
is Running operation, for which the change-over switch 901 is
moved into the position 2, so that the controller 902
receives the control deviation 910 as the input variable,
will now be described. The driver stage 903 supplies the
ionisation tube 904 with ionisation power as a function of
2o the output of the controller 902. The control algorithm of
the controller 902 corresponds to an integration
controller, the function of which is illustrated by the
flow diagram according to Fig. 10. First of all, it should
be assumed that the controller supplies a previously stored
2s initialisation variable, which corresponds to a low
ionisation power, at the output. As long as the pollutant
concentration 906 corresponds to the previously adjusted
desired value, the control deviation 910 remains unchanged
at zero, so that the controller does not take any action.
3o If the pollutant concentration 906 now increases, this
increase in the pollutant concentration is detected by the
gas sensor 905, which results in an enlargement of the
control deviation 910. The controller 902 then increases
the control variable 914 as a function of the control
as algorithm, so that the ionisation tube 904 is acted on by a
CA 02494917 2005-02-07
- 32 -
greater ionisation power via the driver stage 903. This
process continues, in accordance with the invention, until,
as a result of the increased ionisation power, the output
signal of the gas sensor 905 is returned to its original
5 value and the control deviation 910 is thus set to zero
once more. The corresponding functioning ensues if,
conversely, the pollutant concentration 906 is reduced once
more.
to The user may use the display 907 to monitor the control
variable 914. Large control variables indicate a high
ionisation power, and hence air that is highly contaminated
with pollutants, whereas low control variables correspond
to the pollutant contaminations determined during the
is calibrating operation. In the air purification device
according to Fig. 8, the display 907 is configured as an
LED display 807. In this case, it is expedient to adjust
the present value range of the control variable 914 to the
display range of the LED display 807. This may be done in
2o that, in a predetermined time window, the value range is
detected between the smallest and the largest control
variable and, between these, the values of the control
variable are divided onto the LED display 807 in a linear
or correspondingly scaled manner (i.e. logarithmically, for
25 example).
The control algorithm will be explained in detail with
reference to Fig. 10, which shows a flow diagram of the
control algorithm of the controller from Fig. 9. In the
3o step 1001, the desired value and the supplied measurement
value of the gas sensor, which, as explained above, was
optionally corrected by a calibration value, are first of
all compared. In the steps 1002 and 1003, it is then
initially checked whether there is a positive or a negative
35 control deviation. If this is the case, in the step 1004
CA 02494917 2005-02-07
- 33 -
or 1005, a waiting timer, which is used to eliminate
disturbance variables, is started. In the step 1006 or
1007, it is then checked whether there is still the control
deviation. If this is the case, the control variable 914
s is raised or lowered.
CA 02494917 2005-02-07
- 34 -
List of Reference Numerals
s
101 Jump function, pollutant concentration
110 Gas sensor
l0 111 PT1-element
112 PT1-element
113 Limitation element
150 Output, gas sensor
201 Jump function, pollutant concentration
is 210 Gas sensor
211 PTl-element
212 PT1-element
213 Limitation element
220 Driver stage
20 221 P-element
222 Limitation element
250 Output, gas sensor
301 Jump function, pollutant concentration
303 Summation point
2s 310 Gas sensor
311 Limitation element
312 PT1-element
313 PT1-element
320 Driver stage
30 321 P-element
322 Limitation element
330 Ioniser
331 P-element
332 Transfer path
ss 340 Controller
CA 02494917 2005-02-07
- 35 -
350 Output, gas sensor
403 Summation point
904 Jump function, ionisation power
405 Summation point
s 410 Gas sensor
411 Limitation element
412 PT1-element
413 PT1-element
420 Driver stage
l0 421 P-element
422 Limitation element
430 Ioniser
431 P-element
432 Path, ioniser - gas sensor
i5 440 Controller
450 Output, gas sensor
500 Air flow
501 Jump function, pollutant concentration
503 Summation point
20 510 Gas sensor
511 Limitation element
512 PT1-element
513 PT1-element
514 Feedback branch of the control circuit
2s 520 Driver stage
521 P-element
522 Limitation element
525 Voltage source
526 Pulse width modulator
30 527 Resonator
528 High-voltage transformer
530 Ionisation tube
531 P-element
532 Path, ioniser - gas sensor
35 540 Controller
CA 02494917 2005-02-07
- 36 -
541 P-element
542 P-element
543 DT1-element
544 I-element
s 545 Summation point
546 Subtraction point
547 Desired value
550 Output, gas sensor
551 Output, controller
l0601 Jump function, pollutant concentration
602 Dead time
603 Summation point
610 Gas sensor
611 Limitation element
is612 PT1-element
613 PT1-element
620 Driver stage
621 P-element
622 Limitation element
20623 Limitation of the ionisation power
624 Limitation of the ionisation power
630 Ioniser
631 P-element
632 Path, ioniser - gas sensor
25640 Controller
641 P-element
642 P-element
643 DT1-element
644 I-element
30645 Summation point
646 Subtraction point
648 Jump function, desired value
650 Output, pollutant sensor
651 Output, controller
ss701 Sensitivity curve, air or oxygen
CA 02494917 2005-02-07
- 37 -
702 Sensitivity curve, pollutants
801 Air purification device
802 Pedestal
803 Cover
s 804 Ionisation tube
805 Gas sensor
806 Air slit
807 LED display
808 Operating potentiometer
l0 809 Electrical supply line
901 Change-over switch
902 Controller
903 Driver stage
904 Ionisation tube
is 905 Gas sensor
906 Air flow, pollutant concentration
907 Display
908 Desired value
909 Comparison point
20 910 Control deviation
911 Adding element
912 Calibration element
913 Voltage level
914 Control variable
25
