EP0985881B1 - Flame monitoring system - Google Patents

Description

The invention relates to a flame monitoring system mentioned in the preamble of claim 1 Type, and a method for monitoring a flame according to the preamble of claim 9.

To monitor oil, gas or coal dust flames, i.a. also such flame monitoring systems or methods used that the intensity fluctuations of the flame in exploit the infrared spectral range. The advantage of such systems is that they are all types of fuel cover, and thus no fuel-specific monitoring types for multi-fuel burners - e.g. Detection of UV radiation for gas and visible radiation for heavy oil - may be required. Disadvantages of IR monitoring are, however, that both the slow changes in intensity of one afterglowing furnace wall - the so-called streak frequencies - as well as the rapid changes of in Usually, light sources operated with AC voltage can simulate a flame. If there is artificial light coming into the burner chamber during operation or even during burner system maintenance, IR monitoring would be present to fake a flame.

The filtering of the streak frequencies specified in various publications with up to 3 Hz be, is relatively easy to carry out by means of high passes, which by the combustion process generated flame frequencies above 10 Hz can not be trimmed. problematic and it becomes more complex, however, when the harmonics of the mains frequency are filtered out must be suppressed. This method also means loss of information from the flame, especially when the network frequency is subject to large tolerances or different nominal frequency ranges must be covered. The for flame monitoring devices Relevant European device standard EN298 also allows the possibility that a corresponding fastening system of the flame sensor its shutdown is achieved when it is off the attachment is removed. In any case, the ambient light security is also when viewing Ensure first and second errors according to EN298. This should be the case with the latter method extremely difficult to fulfill because the functionality e.g. of a limit switch only the actual distance of the flame sensor from its mounting can be checked.

It is therefore an endeavor to establish immunity to mains frequency-modulated extraneous light sources to reach electronically, be it by inherently fail-safe circuits or by cyclical Test of monitoring systems designed for continuous operation of course during the Burner operation.

EP 320 082 A1 describes a flame monitoring circuit in which only the Evaluation of the alternating light component of a flame as a measure for a fail-safe Flame detection is used. However, this solution only offers security against flame simulation, as long as the safety-relevant ambient light mentioned there is concerned Acts like light. Light from mostly AC-powered external light sources, on the other hand, leads very well to fake a flame and thus to an unsafe burner operation. In addition, there is a risk that an internal component error in the IC despite the missing flame Control of the fuel valve is maintained. For this reason alone, the Use on burners in continuous operation.

EP 334 027 A1 discloses a solution that is suitable in this regard, but the effort is a consequence the full two-channel system is disproportionately high and immunity to network frequencies Alternating light signals are achieved with frequency selective arrangements, the disadvantage of which Loss of flame signal information has already been mentioned.

A solution that eliminates this deficiency is shown in EP 229 265 A1. There will be grid frequency harmonic signals locked with high selectivity, so that the loss of information the flame signal is kept very low. However, the applicability to burners in the Continuous operation questionable because of an internal component error e.g. the flip-flop - with flame simulation as Consequence - is not determined in operation, and also immunity to network frequencies Alternating light signals could only be detected when the burner is at a standstill.

The object of the present invention is to provide a flame monitoring system and a method for Monitoring a flame to create immunity to harmonics Has input signals with the least loss of flame signal information and for use with Burners in continuous operation is suitable.

The stated object is achieved according to the invention by the features of independent claims 1 and 9 solved. Advantageous refinements result from the subclaims.

The present invention solves the problem in that a flame sensor first of all Flame outgoing radiation converts into a flame signal, which in turn by a Flame signal amplifier is converted into an output signal. A frequency selective arrangement that is arranged in parallel to the flame signal amplifier, also receives the flame signal itself and checks for the presence of periodic signals. Will the presence of non-periodic signals detected by the frequency selective arrangement, the Flame signal amplifier activated while in the detection of periodic signals or in the In the absence of a flame signal the flame signal amplifier is deactivated. About that it is also possible to overwrite the flame signal with a test signal so that the Input of the flame signal amplifier as well as the input of the frequency selective arrangement the test signal itself can be applied, so that failures within the Flame monitoring circuit, for example the failure of individual components can be detected.

The frequency-selective arrangement has a frequency detector which detects the presence non-periodic flame signals are detected and the flame signal amplifier via appropriate switching means activated or deactivated accordingly. This can be done in different ways become.

On the one hand, it is possible to first amplify the flame signal and convert it into a square-wave signal convert, whereby any reference signal can be used for this conversion. This square wave signal then serves as a control signal for a bipolar current or voltage source in turn feeds an integrator, so that the output signal of the integrator with periodic input signals of the frequency detector fluctuates around a constant mean value. In other Expressed in words, the bipolar current or voltage source charges the integrator depending on the fluctuation range of the input or the flame signal, so that with periodic input signals the average integration value is approximately zero.

The frequency-selective arrangement also has a coupler or a switch that first determines whether the output signal of the frequency detector, that is, the integrated Input signal within a defined switching threshold around a certain average remains in order to then operate a switch which activates the flame signal amplifier accordingly or deactivated. If the frequency detector detects that a purely periodic signal is present, then guarantees the above Switching threshold that residual fluctuations in the integrated signal around the constant mean value or slight deviations around the zero value are disregarded, which, depending on the cutoff frequency of the integrator, is also caused by purely periodic input signals can be.

Another possibility is that the frequency detector receives the input signal, ie For example, the flame signal is integrated over previously defined periods and the frequency-selective arrangement uses the integrated output signal to actuate a switch, which in turn activates or deactivates the flame signal amplifier. By integrating over this defined periods, it is possible to use a very narrow, ie narrow-band filtering discrete To make frequencies that are usually multiples of the network frequency, so here Extraneous light components that follow the AC voltage of the mains frequency are filtered out sharply, so that all other frequencies, ie flame signals in particular, are almost loss-free can be detected. It makes sense to use the frequency detector after each integration of the defined periods back to their original state, otherwise a drift the integrator output voltage could lead to flame simulation, which is a component fault when tested would be recognized.

The flame signal can be overwritten with a periodic test signal, so that the Frequency detector then evaluates the test signal, which is the verification of the circuit as such enables and the failure of individual components is detected.

The frequency detector activates the switch so that the flame signal amplifier at one non-periodic flame signal provides a valid output signal while in the detection of periodic input signals at the frequency detector of the flame signal amplifier is deactivated, so that no valid signal is supplied at the output of the flame signal amplifier.

The periodic test signal is advantageously applied at regular time intervals to always To have certainty about the proper functioning of the flame monitoring circuit.

Fig. 1

ein Flammenüberwachungssystem mit geeigneten Schaltelementen,

Fig. 2

die Signalverläufe des Flammenüberwachungssystems nach Fig. 1,

Fig. 3

ein anderes Flammenüberwachungssystem mit geeigneten Schaltelementen,

Fig. 4

ein Testsignal des Flammenüberwachungssystems nach Fig. 3,

Fig. 5

die Signalverläufe des Flammenüberwachungssystems nach Fig. 3, und

Fig. 6

ein vereinfachtes Flammenüberwachungssystems nach Fig. 3.

Preferred exemplary embodiments of the invention are explained in more detail below with reference to the drawings. Show:

Fig. 1

a flame monitoring system with suitable switching elements,

Fig. 2

1 the signal profiles of the flame monitoring system according to FIG. 1,

Fig. 3

another flame monitoring system with suitable switching elements,

Fig. 4

3, a test signal of the flame monitoring system according to FIG. 3,

Fig. 5

the waveforms of the flame monitoring system of FIG. 3, and

Fig. 6

a simplified flame monitoring system according to FIG. 3.

1 shows a flame monitoring system. The flame radiation picked up by a sensor 1 and converted into an electrical signal, the signal voltage U ₁ , is first amplified in a first input amplifier 2 with high-pass behavior and fed to the input of a Schmitt trigger 3. The signal voltage U ₁ is based on a mass m. With the signal voltage U ₂ at the output of the Schmitt trigger 3, firstly a bipolar current source 4 is now driven, which charges a first integrator 5 positively or negatively with respect to a reference voltage U _Ref . The polarity and duration of the respective charging cycles depend on the state of the output of the Schmitt trigger 3 and thus directly on the signal voltage U _{1 of} the sensor 1. The integrator 5 has a low-pass behavior, the cut-off frequency of the low-pass typically being around 80 Hz.

The signal voltage U ₂ at the output of the Schmitt trigger 3 is secondly processed by means of a circuit 6 for controlling an n-channel JFET 7 (junction field effect transistor) operating as a switch. The circuit 6 is designed as a charge pump consisting of two capacitors and two diodes, which transforms the alternating output signal U _{2 of} the Schmitt trigger 3 into a direct voltage signal U _{3 of} negative polarity. The DC voltage signal U ₃ is fed to the control input of the JFET 7 via a second switch 8 controlled by the output signal U _{4 of} the integrator 5. The control input of the JFET 7 is connected to the reference voltage U _Ref via a capacitor 9 for smoothing the control voltage. In the example, the second switch 8 is designed as the light-receiving side of an optocoupler 10, the light-transmitting side of which the signal voltage U _{4 is} supplied to the output of the integrator 5 via a rectifier 11.

The rectifier 11 and the downstream optocoupler 10 represent a load for the integrator 5. The integrator 5 is now charged or discharged at irregular intervals on the one hand by the current source 4 according to the state of the output of the Schmitt trigger 3. On the other hand, the integrator 5 is loaded insofar as the amount of the signal voltage U ₂ at its output is above the switching threshold of the optocoupler 10. At frequencies of the signal voltage U ₁ , which are below the limit frequency of the low-pass filter of the integrator 5, the charging current supplied by the current source 4 for the integrator 5 is significantly greater than the discharge current due to the load by the rectifier 11 and the optocoupler 10, so that the Integrator 5 can be charged to a comparatively large positive as well as negative potential. At frequencies of the signal voltage U ₁ , which lie above the limit frequency of the low-pass filter of the integrator 5, on the other hand, the discharge current due to the load by the rectifier 11 and the optocoupler 10 is significantly greater than the charging current supplied by the current source 4, so that the signal voltage U ₂ remains at the output of the integrator 5 below the switching threshold of the optocoupler 10.

The signal voltage U ₁ is now secondly fed to a second input amplifier 12 with high-pass behavior, rectified by means of a second rectifier 13 and fed to a second integrator 14. If the JFET 7 blocks, then the signal voltage U _{1 is} amplified by the second input amplifier 12 and the voltage U ₅ at the output of the second integrator 14 has a value which is different from the potential of the ground m. If, however, the JFET 7 is in the conductive state, then the signal voltage U ₁ at the input of the amplifier 12 becomes ineffective, so that the voltage U ₅ at the output of the integrator 14 assumes the potential of the ground m.

2 shows the voltage signals U ₁ , U ₂ and U ₄ in the event that only radiation emanating from the flame falls on the sensor 1. At the output of the Schmitt trigger 3, pulses 15 of different lengths occur. As long as a pulse 15 is present, the integrator 5 is charged by the current source 4, the integrator 5 is discharged in the pauses between the pulses 15. According to the above description, the signal voltage U ₂ is mostly above the switching threshold 16 of the optocoupler 10. As can be seen from the figure, however, the optocoupler 10 is switched on and off at irregular intervals. Thanks to the smoothing of the output signal of the optocoupler 10 by the capacitor 9, however, the JFET remains in the blocking state, so that the flame signal U ₁ reaches the second input amplifier 12 and the voltage U ₅ at the output of the second integrator 14 has a value, the "flame." available "means.

If the sensor 1 (FIG. 1) is released from its holder and placed next to the burner, the light coming from a neon tube, for example, the fundamental frequency of which is approximately 100 Hz, then striking the output of the Schmitt trigger 3 a signal voltage U ₂ , which consists of a regular sequence of pulses 15, the duty cycle of which is 1. The pulses 15 charge and discharge the integrator 5 by means of the current source 4 for periods of the same length, so that the signal voltage U ₄ at the output of the integrator 5 is already a triangular voltage after a short time, the peak values of which due to the low-pass behavior of the integrator 5 below the switching threshold of the Optocoupler 10 lie. The optocoupler 10 then remains permanently switched off and the JFET 7 becomes conductive. As a result, the flame signal is no longer amplified by the second input amplifier 12 and the voltage U ₅ at the output of the second integrator 14 takes on the value of the mass m, which means "flame not present".

FIG. 2 also shows the profile of the signal voltage U ₄ in the event that the sensor 1 was released from its holder at the time t ₁ (FIG. 1). The switching thresholds 16 of the optocoupler 10 are also shown. The signal voltage U ₄ , which happens to have a high value at time t ₁ , so that the JFET blocks, gradually decreases due to the low-pass behavior of the integrator 5 and, finally, is no longer able to control the optocoupler 10 ,

A control input is also shown in FIG. 1, via which a test signal T can be superimposed on the signal voltage U ₁ . Such a test signal T is, for example, a 100 Hz signal which simulates a light source operated with alternating current. If the test signal T is applied from the point in time t ₁ , the output signal U _{4 of} the integrator 5 runs against the reference voltage U _Ref due to the damping of the coupler 19, that is to say the rectifier 11 and the optocoupler 10, after falling below the switching threshold 16 and after Expiry of the time period Δ _t the output voltage U ₅ at the output of the flame signal amplifier 40 assumes the value of the mass m. 2 there is thus an output signal which, despite the sensor being strongly illuminated with artificial light, gives the evaluation “flame not present”.

However, an output signal is often required that not only reports the presence of a flame, but also represents a measure of the strength of the flame radiation detected by the sensor. For this The reason is the actual flame signal amplifier 40 as a purely analog processing channel with the Blocks 12, 13 and 14 built.

1. Die Meldung, ob ein gültiges Flammensignal U₁ vorliegt, das heißt, ob sich die Frequenz des Eingangssignals und damit das Ein-/Aus-Verhältnis des Schmitt-Triggers 3 laufend ändert.

2. Der Nachweis, daß der Analogwert U₅ am Ausgang des Integrators 14 zu Null wird, wenn vom Flammensensor 1 ein Signal mit konstanter Frequenz bzw. kein Signal mehr geliefert wird, wobei dieser Nachweis als Folge der Einspeisung einer Testspannung U_T erbracht werden muß.

Blocks 18, 19 and 6 and 17 have two different tasks to perform here:

1. The message as to whether a valid flame signal U ₁ is present, that is to say whether the frequency of the input signal and thus the on / off ratio of the Schmitt trigger 3 is constantly changing.

2. Evidence that the analog value U ₅ at the output of the integrator 14 becomes zero when the flame sensor 1 supplies a signal with a constant frequency or no signal, this verification having to be provided as a result of the injection of a test voltage U _T ,

The solution according to FIG. 1 is not only limited to the blocking of certain frequencies, but in principle forms the mean value 0 at every constant frequency at the integrator 5. However, the instantaneous voltage reaches more or less depending on the frequency of the input signal U ₁ and on the time constant of the integrator 5 high values, so that a periodic pulse control of the coupler 19 is possible under certain system conditions. Here it is advisable to supplement the integrator 5 with a series resistor to form a simple RC low-pass filter and to design the current source 4 as a voltage source, for example. a bipolar voltage source, so that for Schmitt trigger pulses with a duty cycle of 1 there is only a moderately increasing attenuation above the cut-off frequency of 6 decibels per octave. However, the more the duty cycle deviates from 1 even at the higher frequencies, the less this damping has an effect. Depending on the time course of the flame signal U ₁ , 5 voltages arise at the capacitor of the integrator, which change more or less frequently in amplitude and polarity. Even if it is assumed that the radiation frequency of mains-operated light sources is twice the mains frequency, for example 100 Hz, the cut-off frequency of the simple low-pass filter mentioned above must be set very low in order to distinguish between the useful signal of the flame and the interference signal of, for example, 100 Hz with sufficient accuracy can. In order to achieve a larger bandwidth for the useful signal with the same signal-to-noise ratio, z. B. the interposition of a low-pass higher order at the output of the input amplifier 2 advantageous.

However, in order to obtain the bandwidth for the flame signal, which is independent of the mains frequency harmonics Interference signals is an infinitely narrow band block for these interference frequencies required.

Fig. 3 shows a solution that is specifically designed for blocking defined harmonic network frequencies, that is, for example 50 Hz, 100 Hz, 150 Hz, etc. Here, the mean value is newly formed over each network period and read out in such a way that sensor signals which are harmonic to the network frequency always lead to the readout value zero, while signals with a different frequency provide values whose magnitudes are different from zero, in order to thus detect a valid flame signal U ₁ . With this principle, the integration time is directly dependent on the current network frequency, which enables a clear distinction between useful and interference signals.

The input amplifier 20 with low-pass characteristic is used for preamplification of the sensor signal U ₁ with simultaneous damping of high-frequency interference voltages. This is followed by a further amplifier 21 with a high-pass characteristic, in which the low streak frequencies are damped as mentioned above.

The output signal of this amplifier 21 is used in three different ways for different purposes further processed. In the averager 22, the integration is carried out over a respective network period. The averager 22 is integrated after each integration interval by means of the switch shown in FIG. 22 is reset to zero. Immediately before this RESET, the current one Value of the integrator is read out by closing switch 23 and via the two-way rectifier 24 connected as a trigger pulse to the input of the monoflop 25. With the differentiator 26 from the leading edge of the monoflop pulse the control pulse for the RESET switch of the integrator or averager 22 won.

In order to control the readout switch 23, the mains ripple voltage becomes 30 in the Schmitt trigger ΔU generates a trigger pulse for the monoflop 29, which in turn is the readout switch 23 operated synchronously. The dependence of the RESET pulse for the integrator 22 on the leading edge of the monoflop 25 - and not directly from the control pulse for the readout switch 23 - should ensure that the content of the integrator 22 is always read out before it is triggered by the RESET pulse is deleted.

Analogously to the principle according to FIG. 1, the output signal U _{4 of} the integrator 22 is also used indirectly here to release the sensor signal U _1, which in this case is preamplified, for further processing.

For this purpose, the preamplified sensor signal U ₁ is first fed to the Schmitt trigger 28, the output pulses of which are used to generate a negative voltage with the aid of the charge pump 6. As in FIG. 1, the negative voltage serves to block the self-conducting JFET 7, as a result of which the input of the active filter stage 33 is released for the preamplified sensor signal U ₁ . This stage in turn has a high-pass characteristic to further attenuate the streak frequency signals. In the following two-way rectifier 34 with an integration capacitor, the analog output voltage U _{5 is} obtained from the preamplified sensor signal U ₁ .

For the test for switching off the output signal U ₅ when sensor signals U ₁ harmonic to the mains frequency occur, by raising the mean value of the amplifier supply voltage U _S from the operating value U _B to the test value U _T, the threshold of the tens diode 31 is exceeded and the test switch 32 is closed, as a result of which the supply voltage U _T superimposed mains ripple voltage .DELTA.U superimposed on the sensor signal U ₁ and thus a mains frequency interference signal is coupled in (cf. FIG. 4). The forced overwriting of the sensor signal by the line ripple voltage leads to the fact that the values averaged over the integrator 22 over each line period become zero, so that ultimately the switch 17, i.e. the JFET 7 is conductive and the output signal U _{5 must} also become zero.

4 shows the switching of the supply voltage U _S from operation U _B to test U _T and vice versa. This switching can also be controlled with a microprocessor system. The error detection takes place according to the same principle as described for FIG. 1.

The amplifier supply voltage U _S is equal to U _B plus ΔU. The phases operation and test are shown in FIG. 4 in such a way that the test voltage is present between the times t 'and t ".

5 shows the contents of the integrator or averager 22. In the case of different sensor signals U ₁ , different output signals U ₄ with different values a, b, c and d are available for reading out at the output of the integrator 22. From the course of the integrator voltage U _{4 it} can be seen that the zero-point-symmetrical interference signal must always deliver the result zero if integration is carried out over constant periods ΔT. These are advantageously network periods or multiples of the corresponding network period. It does not matter at what point in time the integration interval begins. The time from the start of the readout to the end of the reset, that is to say to the beginning of the next integration interval, can be kept so short compared to the network period duration ΔT, that is to say the interval itself, that the "measurement error" can be neglected despite integration in each of the successive network periods.

3 can also be varied. 6 shows such an example Variant of the circuit of Fig. 3. So it is z. B. possible for the control of the charge pump 6 share the output signal of the Schmitt trigger 30 in order to save the Schmitt trigger 28. In addition to saving components, this variant would have the advantage of being more uniform and thus more reliable gate voltage generation for the self-conducting JFET 7 due to constant Pump frequency. However, an existing property would then be lost, including one has a certain advantage, namely that the charge pump with its high-pass character an additional Security against the detection of streak signals as described above, if as in Fig. 3 from Useful signal is dependent, offers.

It is also conceivable that the active filter stage 33 can be omitted if the attenuation of the streak frequencies in the high-pass amplifier 21 is already sufficient to prevent a flame simulation.

The Schmitt trigger 30 is also not necessary because the monoflop 29 is directly from the Line ripple voltage ΔU can be controlled.

Another variant with savings on the Schmitt trigger 28 would be the activation of the Charge pump 6 from the monoflop 25. As a result, the transistor 27 could be omitted, so that the Discharge time constant of the charge pump 6 can be dimensioned so small that the test in the available time can be carried out.

Claims (12)

1. A flame monitoring system comprising
      a flame sensor (1) which converts the radiation emanating from the flame into a flame signal (U₁),
      a flame signal amplifier (40) which converts the flame signal (U₁) into an output signal (U₅), and
      a frequency-selective arrangement (6, 17, 18, 19) which detects the presence of periodic signals in the flame signal (U₁),
      characterised in that
      the frequency-selective arrangement (6, 17, 18, 19) activates the flame signal amplifier (40) when there is a non-periodic flame signal (U₁), and
      the frequency-selective arrangement (6, 17, 18, 19) deactivates the flame signal amplifier (40) when there is a flame signal (U₁) with periodic signals or no flame signal (U₁) or a test signal (T).
2. A flame monitoring system according to claim 1 characterised in that the frequency-selective arrangement (6, 17, 18, 19) has a frequency detector (18) which detects the presence of non-periodic flame signals (U₁) and appropriately activates or deactivates the flame signal amplifier (40) by way of switching means (6, 17, 19).
3. A flame monitoring system according to claim 1 or claim 2 characterised in that the frequency detector (18) converts the flame signal (U₁) into a rectangular signal (U₂) and uses the rectangular signal (U₂) as a control signal of a bipolar current/voltage source (4) for feeding an integrator (5) so that the output signal (U₄) of the integrator (5) fluctuates about a constant mean value (U_Ref) with periodic flame signals (U₁).
4. A flame monitoring system according to claim 2 or claim 3 characterised in that the frequency-selective arrangement (6, 17, 18, 19) has a coupler (19) which, when the output signal (U₄) of the frequency detector (18) remains within a defined switching threshold (16) around the constant mean value (U_Ref) activates a switch (17) which deactivates the flame signal amplifier (40).
5. A flame monitoring system according to claim 1 or claim 2 characterised in that the frequency detector (18) integrates the flame signal (U₁) over defined periods and the frequency-selective arrangement (6, 17, 18) uses that integrated output signal (U₄) for the actuation of a switch (17) which activates or deactivates the flame signal amplifier (40).
6. A flame monitoring system according to claim 5 characterised in that the frequency detector (18) is resettable to its initial state after each integration over one of the defined periods.
7. A flame monitoring system according to claim 5 or claim 6 characterised in that the defined periods are multiples of the mains periods.
8. A flame monitoring system according to one of the preceding claims characterised in that the flame signal (U₁) is overwritable with a periodic test signal (T) and the frequency detector (18) evaluates the test signal (T).
9. A method of monitoring a flame, wherein
      the radiation emanating from the flame is converted into a flame signal (U₁) and same is converted into an output signal (U₅), and
      the presence of periodic signals in the flame signal (U₁) is detected by a frequency-selective arrangement (6, 17, 18, 19),
      characterised in that
      the flame signal (U₁) is converted into an output signal (U₅) when there is a non-periodic flame signal (U₁), and
      the flame signal (U₁) is converted into a zero signal (U_m) when there is a flame signal (U₁) with periodic signals or no flame signal (U₁) or a test signal (T).
10. A method according to claim 9 characterised in that periodic flame signals (U₁) are detected by means of integration of the flame signals (U₁) over given periods and/or around a constant mean value (U_Ref).
11. A method according to claim 9 or claim 10 characterised in that if the integrated flame signal (U₁) is nearly zero or remains within a defined switching threshold (16) around the constant mean value (U_Ref). the flame signal (U₁) is converted into the zero signal (U_m).
12. A method according to one of claims 9 to 11 characterised in that a periodic test signal (T) is applied in regular time intervals and the occurrence of the zero signal (U_m) is checked.

Images