Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
METHOD AND DEVICE FOR PROVIDING POWER TO A MAGNETRON
Document Type and Number:
WIPO Patent Application WO/1998/011655
Kind Code:
A1
Abstract:
The present invention relates to a method and a device for providing power to a magnetron from an alternating voltage source. According to the invention, the power supply is accomplished by said alternating voltage being rectified into a rectified voltage which is then converted into a high direct current for powering said magnetron, wherein the power provided to said conversion is controlled based upon a detected actual effect in the power supply means.

Inventors:
Lidstr�m, Kjell (Kritgatan 19, Bure�, S-930 15, SE)
Application Number:
PCT/SE1997/001520
Publication Date:
March 19, 1998
Filing Date:
September 09, 1997
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
IKL SKELLEFTE� AKTIEBOLAG (P.O. Box 192, Skellefte�, S-931 22, SE)
Lidstr�m, Kjell (Kritgatan 19, Bure�, S-930 15, SE)
International Classes:
H02M1/00; H02M1/42; H02M5/458; H05B6/66; (IPC1-7): H02M3/335; H05B6/68
Foreign References:
US5451750A
US4318165A
US5208432A
US4076996A
Other References:
PATENT ABSTRACTS OF JAPAN, Vol. 13, Nr. 429, E-823; & JP,A,01 159 994 (NEW JAPAN RADIO CO LTD), 22 June 1989.
Attorney, Agent or Firm:
AWAPATENT AB (P.O. Box 45086, Stockholm, S-104 30, SE)
Download PDF:
Claims:
CLAIMS
1. Method for providing power to a magnetron from an alternating voltage source, comprising the steps of: a) converting said alternating voltage into a rectified voltage; b) converting said rectified voltage, preferably by means of high frequency switching, into a high frequency alternating voltage; c) transforming said high frequency alternating voltage into a high frequency alternating high voltage; and d) rectifying said alternating high voltage into a rectified high voltage which is applied to said magnetron for providing power thereto; wherein said step a) comprises controlling the power provided to the step b) conversion based upon a detected actual power level.
2. Method as claimed in claim 1, wherein said actual power level is detected prior to step b) .
3. Method as claimed in claim 1 or 2, wherein said power controlling step comprises: comparing the detected actual power level with a set desired power level; and controlling the power provided to the converting step b) in such a way that the detected actual power level is kept essentially equal to the desired power level.
4. Method as claimed in any one of the preceding claims, comprising controlling the current provided from said alternating voltage source to be sinusoidal, or to have the same wave form as the voltage, and/or to be essentially in phase with the voltage.
5. Method as claimed in claim 4, wherein said current controlling step is performed integrated with said power controlling step.
6. Method as claimed in any one of the preceding claims, wherein said controlling step comprises control¬ ling, by means of high frequency switching, the power provided to the step b) conversion.
7. Method as claimed in any one of the preceding claims, wherein said actual power level is detected by detecting an actual current and an actual voltage providing said actual power level.
8. Method as claimed in claim 7, comprising the steps of: calculating the quotient between said desired power level and the mean value of said actual voltage as a measure of the desired current mean value; scaling the desired current mean value with the quotient between said actual voltage and the mean value of the actual voltage for providing an instantaneous desired current value; and relating the desired current value to said actual current and performing said controlling based upon this relation.
9. Method as claimed in any one of the preceding claims, comprising tuning the steps b) to c) so that each inversion switching in step b) is performed at zero current crossings and hence essentially without losses.
10. Method as claimed in any one of the preceding claims, wherein said inversion is performed with 100% duty cycle.
11. Device for providing power to a magnetron from an alternating voltage source, comprising: first rectifying means (1; 11; Brl) for rectifying said alternating voltage into a rectified voltage; inverter means (14; T2, T3) for inversion of said rectified voltage into a high frequency alternating voltage; means (15; Tr2) for transforming said high frequency alternating voltage into a high frequency alternating high voltage; and second rectifying means (16; D2, D3, C4, C5) for rectifying said alternating high voltage into a rectified high voltage to power said magnetron; means (5; 13; 19; R) for detecting an actual power; and means (4; 12; R, LI, Tl, Cl, Dl) for, prior to said inverter means, controlling the power provided to said inverter means based upon the detected actual power.
12. Device as claimed in claim 10, wherein said means for detecting an actual power are provided to detect said power prior to said inverter means.
13. Device as claimed in claim 11 or 12, wherein said control means are provided to compare the detected actual power with a set desired power and to control the power provided to said inverter means in such a way that the detected actual power is kept essentially equal to the desired power.
14. Device as claimed in any one of claims 11 to 13, wherein said control means (18; 19; R) are provided to control the current provided from said alternating vol¬ tage source to be sinusoidal, or to have the same wave form as the voltage, and/or to be essentially in phase with the voltage.
15. Device as claimed in claim 14, wherein said control means are provided to perform said current control integrated with said power control.
16. Device as claimed in any one of claims 11 to 15, wherein said control means comprise a high frequency switching converter (Tl),.
17. Device as claimed in any one of claims 11 to 16, wherein said detecting means are provided to detect an actual current and an actual voltage providing said actual power.
18. Device as claimed in claim 17, wherein said control means comprise processing means (M/D) provided to calculate the quotient between said desired power level and the mean value of said actual voltage as a measure of the instantaneous desired current mean value and to multiply this value with the quotient between said actual voltage and the mean value of the actual voltage to provide an instantaneous desired current value, wherein said control means are provided to relate the desired current value to said actual current and to perform said power control based upon this relation.
19. Device as claimed in claim 18, wherein said control means further comprise comparator means (C) for relating said desired current value to said actual current and to provide a signal which is to be used as a basis for said power control based upon this relation.
20. Device as claimed in any one of claims 11 to 19, comprising means (L2, C2, C3, C4, Tr2) for tuning said inverter means so that each inversion switching is performed at zero current crossings and hence essentially without losses.
21. Device as claimed in any one of claims 11 to 20, wherein said inverter means comprise two transistors (T2, T3) which are alternately active for providing said alternating voltage.
22. Device as claimed in any one of claims 11 to 21, wherein said inverter means (T2, T3) operate at a 100% duty cycle.
Description:
METHOD AND DEVICE FOR PROVIDING POWER TO A MAGNETRON

Technical Field of Invention

The present invention refers to a method and a device for providing power to a magnetron.

Technical Background and Prior Art

Magnetrons are devices which are used in microwave systems for converting electrical energy into microwaves. Magnetrons are for example used as microwave sources in radar equipment, microwave ovens and plasma lamps. The power supply to a magnetron essentially consists of the voltage which is applied between the anode and cathode (filament) of the magnetron, which is generally called the anode voltage. This anode voltage may be of the order of a few kilovolts. The current passing through the magnetron caused by this voltage is generally called the anode current.

In Fig. 1, there is shown a schematic diagram of a typical relationship between the anode voltage U and the anode current I of the magnetron. Except for extremely small currents, the magnetron shows a relatively low dynamic impedance. This means that very small changes in the anode voltage will give rise to very large changes in the anode current. To compensate for this, the anode current from a power supply unit is preferably provided at a relatively high, preferably inductive, generator impedance.

As the magnetron characteristics or conditions changes, for example because of the magnetron succes¬ sively being heated and the temperature hence changing during operation, the emission capacity of the filament slowly decaying after use for a longer period of time, the load changing or because of similar circumstances, the voltage to current relationship of Fig. 1 may be shifted in relation to the initial position. Depending on how the power supply is affected or controlled as a result of these changed conditions, the applied anode

voltage and/or the applied anode current may vary considerably.

A known technique for controlling the power supply to the magnetron is based upon the principle that the anode current is controlled under the assumption that the anode voltage is, or is being held, constant.

An example thereof is shown in Fig. 2, wherein a magnetron power supply, comprising a first, rectifying step and a second, high voltage generating step, supplies power to a magnetron, wherein the current is detected just before the magnetron and the high voltage generation is controlled based thereupon.

Much effort has been invested in finding a suitable method for stabilizing the magnetron power, but, due to the complexity of the problem, successes have been limited.

Hence, there exists a need for improved possibili¬ ties to control the power supply to a magnetron, espe¬ cially in applications showing difficult load conditions, such as microwave powered plasma lamps.

Objects of the invention

An object of the invention is to provide a more stable magnetron performance, especially during difficult load conditions.

Another object of the invention is to provide more accurate control of the power supply to a magnetron.

Another object of the invention is to provide a power supply which exhibits very small losses.

Summary of the invention

The above mentioned and other objects are achieved by a method and a device having the features defined in the accompanying claims. According to an aspect of the invention, power supply to a magnetron from an alternating voltage source is accomplished. The power supply comprises rectifying

the alternating voltage from the alternating voltage source into a rectified voltage which is then converted into a rectified high voltage for powering said magnet¬ ron. Furthermore, the power provided to said conversion is controlled based upon an actual power.

The invention is hence based upon the insight as to the advantage of controlling the power in a step prece¬ ding the conversion from rectified low voltage to recti¬ fied high voltage. Since the power control is performed prior to this conversion, the conversion itself need not comprise power controlling elements, whereby the conver¬ sion efficiency is increased considerably.

The inventors have hence come to the insight that the effect of load changes, mains power supply disturban- ces, temperature and the like, upon the operation of the magnetron can be suppressed by continuously making sure that the power which is provided by an initial power supply stage is constant.

Preferably, the detected actual power is compared to a set desired power and the power provided to said con¬ version is controlled in order to keep the detected actual power essentially equal to the desired power.

Instead of only using a current value or a voltage value as feedback for controlling the power supply, the inventors have come to the insight that increased control of the magnetron performance is achieved when the power control is based upon a "true" power value.

Furthermore, the inventors have realized the advan¬ tage of performing this power control based upon a power detected in association with the initial rectifying and, in any case, prior to the conversion into high voltage.

To avoid having to derive information as to the actual power level from the high voltage generating step or closer to the magnetron, where the high voltage levels make detection of a "true" power value difficult, the power is preferably detected prior to said conversion.

As the power detection as well as the power control based thereupon for keeping the power constant are per¬ formed prior to the conversion from rectified low voltage to rectified high voltage, this conversion, as stated, may be designed without the incorporation of controlling or feedback elements, i.e. the conversion may be given an optimal design showing small losses. This means that the power losses in the conversion step will be under good control. By controlling the power prior to an optimized conversion step, there is no need for knowledge about the specific circumstances closer to the magnetron, as these basically follows directly from the controlled power.

According to a preferred embodiment of the present invention, the current provided by said alternating voltage source is also controlled to be sinusoidal, or to have the same wave form as the alternating voltage, and/or to be essentially in phase with the alternating voltage. Thereby, the power supply unit wil] show a very high power factor, and it is ensured that no harmonics of the mains frequency will be sent back to the public supply network.

It has been found that this current and power cont¬ rol may be integrated into one single unit/function in a very preferable manner. According to yet another embodi- ment of the invention, the control hence comprises high frequency switching of the rectified signal for providing a controlled, constant and power factor corrected power. Preferably, an actual current and an actual voltage is detected, together providing said actual power. These actual current and voltage values may then be used for providing power control as well as current control as described above.

According to an especially preferred embodiment of the invention, the measured current and voltage is used in a processing operation, wherein the quotient between the desired power and the mean value of said actual vol¬ tage is calculated, resulting in a value of the desired

current mean value. This desired mean value is then multiplied with the quotient between said actual voltage and the mean value of the actual voltage, resulting in the voltage phase factor, to provide an instantaneously desired current value.

Based upon the processing described above, or simi¬ lar operations which provide a desired current value, the control is performed by comparing the desired current value with the actual current. The current and the vol- tage is then controlled so as to keep the power constant, and to increase or decrease the current based upon this comparison.

As mentioned above, the invention has the advantage that the conversion step may be designed to exhibit a high efficiency. According to a preferred embodiment, the conversion from rectified low voltage to rectified high voltage is performed by inverting the rectified voltage at a comparatively high frequency, after which the inverted voltage is transformed into an alternating high voltage which is then rectified into said rectified high voltage for powering the magnetron.

As these steps lack monitoring, control, feedback, and the like, incorporated circuits and functions may be tuned in a very preferable manner so that each inverting switching is performed at a zero current crossing and hence essentially without losses. The inversion is advan¬ tageously performed by the use of two transistors which are kept active alternately for providing the generally pulse shaped, alternating voltage. As neither the inver- sion frequency nor the on-off-ratio have to be varied for controlling the power supply, since this control has al¬ ready been accomplished in a preceding step, the inver¬ sion operation performed by the transistors may take place at a 100% duty cycle, thus allowing for said tuning, which in turn allows for the switching at zero current crossings. This is one of the primary factors which contribute to the high efficiency.

Further advantages, aspects and features of the pre¬ sent invention will become more clear from the descrip¬ tion below.

Brief Description of the Drawings

Exemplifying embodiments of the invention will now be described with reference to the accompanying drawings, wherein:

Fig. 1 schematically shows a typical relationship between the anode voltage and the anode current of a magnetron;

Fig. 2 shows a block diagram of a power supply unit using control according to prior art;

Fig. 3 schematically shows a block diagram of a power supply unit according to a first embodiment of the present invention;

Fig. 4 schematically shows a block diagram of a power supply unit according to a second embodiment of the present invention; Fig. 5 schematically shows a block diagram of a power supply unit according to a third embodiment of the present invention;

Fig. 6 schematically shows a block diagram of a power supply unit according to a fourth embodiment of the present invention;

Fig. 7 schematically shoes a detailed exemplifying embodiment of a device according to the present invention;

Fig. 8a to 8e schematically show signal wave forms present in the circuit of Fig. 7;

Fig. 8f schematically shows a simplified diagram of elements included in the circuit of Fig. 7; and

Fig. 8g and 8h schematically show signal wave forms present in the circuit of Fig. 8f.

Detailed description of preferred embodiments

I Fig. 1, there is shown a schematic diagram of a typical relationship between the magnetron anode voltage U and anode current I. Except for extremely small currents, the magnetron shows a comparatively low dynamic impedance.

In Fig. 2, there is schematically shown a block dia¬ gram of an example of a prior art magnetron power supply comprising a rectifying stage 1, which rectifies the alternating voltage from an alternating voltage source into a rectified voltage, and a voltage converting step 2, which converts the rectified voltage into a rectified high voltage for providing power to a magnetron 3. In the prior art power supply, the anode current is used as a feedback signal for a control stage incorporated in the voltage converting stage 2.

Fig. 3 schematically shows a block diagram of a power supply unit according to a first embodiment of the present invention. In Fig. 3, a power control step 4, 5 is provided between the rectifying stage 1 and the vol¬ tage converting stage 2. The power control step 4, 5 detects the power of the voltage being provided to the converting stage 2, compares this power with a set desired power Pset and continuously controls the power based thereupon. Since there is neither feedback nor control in stage 2, the converting stage may be designed to show an optimized high efficiency, as will be described more fully below.

In Fig. 4, there is shown a block diagram of a power supply unit according to a second embodiment of the invention. The power supply unit in Fig. 4 comprises a first rectifying stage 11, a power control stage 12, a preferably high frequency switching, inverting stage 14, a step-up transformation stage 15, and a second rectify- ing stage 16, and the magnetron 17 itself. A detecting/- controlling stage 13 detects the power provided to the inverting stage 12 and controls the power control stage

12 so as to keep said power equal to Pset, i.e. so as to keep the power provided to the following stage 14 constant.

As the stages 14, 15, and 16 lack monitoring, cont- rol, feedback, and the like, these stages are tuned so that each inverting switching takes place at a zero cur¬ rent crossing and hence essentially without losses. The inversion in stage 14 is achieved by the use of two transistors operating alternately. Since power control has already been accomplished in stage 12, 13, the transistors of stage 14 may operate at a 100% duty cycle. Inductances and capacitances present in the three cir¬ cuits are dimensioned so as to get a resonance frequency having a ratio, generally equal to 1/1, to the switching frequency in stage 14, thus allowing for said tuning with switching at zero current crosses.

In the embodiment of Fig. 5, the block diagram of Fig. 4 has been supplemented with a power factor control¬ ling stage 18 which makes sure that the current supplied from the AC mains follows the shape of the voltage wave form, which generally is sinusoidal, and that no harmo¬ nics are passed to the public power network. With this arrangement, both power control and power factor control are performed in the initial stages without any mter- ference with the tuned stages 14, 15, and 16.

In Fig. 6, there is schematically shown how the power controlling and the power factor controlling functions may be integrated into on single control unit 19. The control unit 19 in Fig. 6 detects the actual power at the output from the rectifying stage, but could 3ust as well detect the power at the input of stage 14, as described in the embodiments above.

A more specific exemplifying embodiment of a magne¬ tron power supply unit according to the invention will now be described with reference to Fig. 7.

The power supply unit in Fig. 7 comprises four functional blocks: a power control block SI, a high

voltage generating block S2, a filament current control block S3 and a processing/control block S4. Furthermore, the power supply unit is arranged to power a magnetron VI. In Fig. 7, the different blocks are schematically separated by dashed lines.

The power control block SI will now be described in detail .

The power control block SI of the power supply unit has inputs 10 and 20 connected to an AC voltage power line, which is preferable obtained from the public elec¬ tricity supply network which provides an alternating voltage of 50 or 60 Hz. The inputs 10 and 20 are connected to a full wave rectifying diode bridge Brl.

The positive output of the diode bridge Brl is con- nected to a first output 30 of the block SI via an induc¬ tor LI provided in series with a diode Dl . The negative output of the diode bridge Brl is connected to a second output 40 of the block SI via a resistor Rs. A transistor Tl has its collector/drain connected to a point between the inductor LI and the diode Dl and has its emitter/- source connected to the second output 40 of the block SI. Furthermore, a capacitor Cl is connected between the outputs of the block SI. Together, the inductor LI, the diode Dl, the transistor Tl, and the capacitor Cl form a high frequency switched power controller.

The on-off-periods of the switch transistor Tl are controlled by a control circuit R which in this example consists of a power factor control circuit, more specifi¬ cally an integrated circuit of the type UC3854 from Unitrode Integrated Circuits, its incorporation having been modified for obtaining the object of the invention. A more detailed description of the functions of this circuit is to be found in "Unitrode Integrated Circuits Application Note", page 303-310. The control circuit R comprises a squared signal provider X, a multiplier/divider M/D, a comparator C and a control unit PWC1. A control signal Pset (Power Set)

from the processing/control block S4 is connected to an input A of the multiplier/divider M/D. Furthermore, the positive output of the diode bridge Brl is connected to an input B of the multiplier/divider M/D and via a low- pass filter LF and the squared signal provider X to an input C of the multiplier/divider M/D. The output of the multiplier/divider M/D is connected to a first input of the comparator C and to the negative output of the diode bridge Brl via a resistance Rl. The second input of the comparator C is connected to the second output of the power control block SI. Furthermore, the output C of the comparator is connected to an input of the control unit PWC1, which output is connected to the gate of the transistor Tl . The function of the power control block S3, will now be described.

According to the invention, the power control block SI is arranged to ensure that a constant power, the mag¬ nitude of which is determined by the signal Pset from the processing/control block S4, is provided to the following blocks/steps independent of variations in the mains vol¬ tage or variations in the load or in the inherent charac¬ teristics of the magnetron. The power control block SI also makes sure that the current supply from the mains is sinusoidal and in phase with the voltage.

The incoming mains AC voltage is full-wave rectified in a conventional manner in the diode bridge Brl so that a pulsating direct voltage V in is obtained between the positive and negative output of the diode bridge Brl. Examples of voltage curves for the network voltage AC and the rectified voltage V ιn from the diode bridge Brl are shown in Fig 8a and 8b, respectively.

The current from the rectifying bridge Brl is then switched at a high frequency by the power controller LI, Dl, Tl, Cl so that it flows alternately through the tran¬ sistor Tl and the diode Dl . The high frequency switched power controller "chops" the low frequency (typically 50

or 60 Hz) pulsating current at a frequency of the order 100 kHz or higher.

In this specific embodiment, the high frequency switched power controller provides a rectified output voltage which is always larger than the input voltage. In this case, the following block must therefore be designed so that it may always be fed with a voltage which is higher than the highest peak value of the mains voltage. The operation of the high frequency switched voltage converter will now be described in detail. When the tran¬ sistor Tl is turned on, the current passing through the inductor LI will grow, thus making the inductance store energy. When the transistor Tl is turned off, the current through the inductor LI will charge the capacitor Cl via the diode Dl, whereby the current through the inductor decreases. Before the current has reached zero, the transistor is once again turned on, which means that the converter will operates at a continuos current.

The larger the part of the on-off-period of the transistor Tl that the transistor Tl is turned on, i.e. the larger the pulse width of the signal being provided to the gate of the transistor, the lager the difference between the voltage between the outputs of the capacitor Cl, i.e. between the outputs of the power control block SI, and the voltage from the diode bridge Brl, i.e. the transistor Tl controls the voltage which is obtained between the outputs 30, 40 of the power control block.

The curves of Fig, 8c, 8d, and 8e show the currents passing through the energy storing inductor LI at three different pulse width relationships but at the same total time period. In Fig. 8c, the active time is 20% of the total period, in Fig. 8d, the active time is 50% of the total period, and in Fig 8e, the active time is 80% of the total period. The derivative of the increasing cur- rent in the curves in Fig 8c to 8e is proportional to the voltage from the rectifying bridge Brl, and the deriva¬ tive of the decreasing current is proportional to the

difference between the voltage over the capacitor Cl and the voltage from the diode bridge Brl.

The relationship between the time that the transis¬ tor Tl is turned on and the time that the transistor Tl is turned off thus controls the power that is provided to the following blocks. According to the embodiment of Fig. 7, this relationship is controlled by the control circuit R.

The interaction between the transistor Tl and the control circuit R thus makes sure that the power provided by the power control block SI is constant and independent of load or variations in the mains voltage, considering the desired power level set by the signal Pset which is provided by the processing/control block. The selected power level, which is represented by the control voltage Pset (Power Set), is provided to the input A of the multiplier/divider M/D. The voltage V ιn from the rectifying bridge Brl is filtered in the low- pass filter LF, squared in the squared signal provider X and then fed to the input C of the multiplier/divider M/D. The voltage V ιn is also provided directly to the input B of the multiplier/divider M/D. The voltage from the rectifying bridge Brl is used as a sample or standard for the wave form of the current. In the equations below, for the sake of simplicity, it is assumed that the network power line voltage is sinusoidal. The mean value V m of the rectified network voltage V ιn is obtained from the low-pass filter LF and is fed via the squared signal provider X to the input C. The relationship between the voltage V ιn and its mean value V m is given by:

A n « Λ/2 V.. sinωt (1)

The function of the multiplier/divider M/D may schematically be described as follows. By dividing the desired power Pset with the mean value V m of the actual voltage, a value is obtained corresponding to the mean

value that the current has to show in order to give rise to the desired power Pset. By dividing the instantaneous value of the voltage V ιn with the value V m of the actual voltage V in , a factor representing the shape of the voltage wave form is obtained. The desired mean value of the current (which ensures the correct power) multiplied by the wave form of the voltage (which ensures that the current is in phase with the voltage) thus gives a measure of the instantaneous value that the current must have in order to accommodate both these requirements.

Therefore, a factor K is formed in the multiplier/- divider M/D, wherein:

The output K from the multiplier/divider M/D is hence generated as the currently desired current. This is compared by the comparator C with the actual current Is. Using the resistance Rl, the output signal K is represented as a voltage K x Rl, which is provided to one input of the comparator and which has its point of refer¬ ence at the negative output of the rectifying bridge Brl. Correspondingly, the current Is is represented using the resistance Rs as a voltage Is x Rs, which is provided to the other input of the comparator and which also has its point of reference at the negative output of the rectify¬ ing bridge Brl.

The difference between the two voltages K x Rl and Is x Rs, which are provided to the comparator C, forms an "error signal" which the power factor control unit is arranged to eliminate by having the comparator control the transistor Tl, as described above, via the pulse width control unit PWC1. As long as K is larger than Is, the transistor is kept turned on, and when Is becomes larger then K, the transistor is turned off. This way, the voltage drop over Rs is brought to be equal to the

voltage drop over Rl, i.e. if neglecting the high frequency ripple:

Rs = K Rl (3!

which together with the equation (2) gives

Rl Rl P.. et r- s = — K = — —- *1 sJ2 sin ωt Rs Rs V„

As is seen, the current Is follows the wave form of Vj .n . Furthermore, the power is obtained as the mean value of the voltage V ιn multiplied by the current Is:

Rl Rl

V,. Is = — P, = , 2 sin " ωt =

Rs sel Rs

which shows that the power is kept equal to Pset except for the factor Rl/Rs.

According to the embodiment of Fig. 7, the power control block SI thus accomplishes several preferable functions: the power from the diode bridge Brl is cont¬ rolled to be constant, based upon Pset, and the current Is supplied from the mains is controlled to have the same wave form as the voltage and to be in phase therewith. The power supply unit thereby shows a high power factor. (The power control block SI further comprises a mains filter (not shown) which makes sure that no high freq¬ uency components in the current is submitted to the mains . )

The description of the embodiment in Fig. 7 will now be directed to the high voltage converting block S2. Generally, the high voltage converting block Ξ2 comprises a high frequency switched half bridge having capacitive voltage division, a transformer and a voltage doubling rectifier. Furthermore, a pulse width control circuit is provided for reasons to be describe below.

More specifically, the block S2 comprises a pulse width control circuit PWC2 having an input, which recei¬ ves a control signal from the processing/control block S4, and two outputs which are connected to the primary coil of a first transformer Trl.

Two transistors T2 and T3 are connected in series between the outputs 30 and 40 of the block SI. Further¬ more, two capacitors C2 and C3 are connected in series parallel to the transistors T2 and T3. A first secondary coil of the transformer Trl is in one of its ends connec¬ ted to the gate of the transistor T2 and is in its other end connected to the emitter/source of the transistor T2. A second secondary coil of the transformer T2 is in one of its ends connected to the gate of the transistor T3 and is m its other end connected to the emitter/source of the transistor T3.

Furthermore, the high voltage converting block S2 comprises a second transformer Tr2 which has a primary coil having one end connected to a point between the two transistors Tl and T2 and having the other end connected to a point between the two capacitors C2 and C3 via an inductor L2.

The high voltage converting block S2 further has a first 50 and a second 60 output connected to the cathode and anode, respectively, of the magnetron VI. In Fig. 7, the anode is connected to ground. An inductor L3 and two capacitors C4 and C5 are connected in series between the outputs 50 and 60. Furthermore, two diodes D2 and D3 are connected in series parallel to the capacitors C4 and C5. The secondary coil of the transformer Tr2 is in one of its ends connected to a point between the two diodes D2 and D3 and is in its other end connected to a point between the two capacitors C4 and C5.

The block S2 further comprises a third transformer Tr3, having a primary coil in series with the primary coil of the transformer Tr2 and having a secondary coil which in one of its ends is connected to the second

output 40 of the block SI and in the other end is connected via a diode D4 to an input of the pulse width control circuit PWC2.

The function of the high voltage converting block S2 will now be described.

The output signal from the pulse width control cir¬ cuit PWC2 controls, via the transformer Trl, the transis¬ tors T2 and T3 to pass current alternately. The switching frequency of the transistors is for example 50 kHz. The choice of frequency depends among other on the power level at which the magnetron is designed to operate.

During normal operation, the high voltage converting block S2 is operated at a 100% duty cycle, i.e. the ratio between the time that any one of the transistors T2, T3 of the high voltage generating block S2 is passing cur¬ rent and the total period is equal to 1. The signal from the pulse width control circuit PWC2 makes each of the transistors T2, T3 pass current 50% of the period time. At start up, before a desired resonance has had time to be established in this "quasi-resonant" converter, the transistors T2 and T3 are activated a shorter time, i.e. at a duty cycle lower than 1. This kind of converter start up is commonly known as slow-start. It is under¬ stood that this initial control of the duty cycle of the high voltage converter block S2 is provided but for achieving a soft start of the converter and i s hence not intended to be used for continuos control of the power provided to the magnetron during normal operation, as is however the case in prior art. During normal operation, when the converter S2 operates at a full duty-cycle, half the output voltage from the power control block SI, which is kept at a relatively constant level, is applied with alternating polarity to the primary coil of the transformer Tr2, which up-transforms the voltage considerably, generally to a few kilovolts.

The leakage inductance of the transformer Tr2, in this case supplemented with a separate inductor L2, is approximately tuned to the switching frequency using the capacitors C2 and C3. The transistors T2 and T3 will thus be turned on and off when the current passing through them is very close to zero. The switching losses will hence be very small.

This provides a very advantageous aspect of the invention. Since the converter S2 operates at a 100% duty cycle, this tuning of the oscillation to the switching frequency is made possible, which in turn makes it possible to switch the transistors T2 and T3 when the current is zero.

If the converter were to operate at a lower changing duty cycle, this kind of tuning would not be possible, i.e. switching of the transistors T2 and T3 would not take place at zero crossings and would hence lead to unnecessary switching losses.

The use of a tuned oscillation with a 100% duty cycle consequently results in an increase in the efficiency of the high voltage converting block S2.

The function of the quasi-resonant converter may partly be illustrated by the schematic block diagram in Fig. 8f. In Fig. 8f, the power to the resonance circuit, which schematically comprises an inductor L and a capacitor C, is switched between plus and minus half the supplied voltage E using the transistors Ta and Tb, and the load, which is schematically illustrated by the resistor r, is connected to ground. In Fig. 8g and 9h, examples of the wave forms of the voltage V and current I in Fig. 8f are shown, said figures clearly illustrating that the current at the time of the switching is very close to zero.

Referring once again to Fig. 7, the serially connec- ted diodes D2 and D3 and capacitors C4 and C5 form a voltage doubling rectifier which puts a lower demand on the ration of the transformer Tr2 and contributes to the

protection against overload at flash-overs in the magnetron, which will be discussed below. The diodes and the capacitors are connected at one arm each of a bridge, wherein the secondary coil of the transformer Tr2 is con- nected at that diagonal which is formed by the connection points for the two diodes and the two capacitors, respec¬ tively, and wherein the output voltage is collected at the other diagonal, where the diodes are connected to the capacitors. The inductor L3 ensures that the magnetron is being provided with an inductive generator impedance, which is generally required for stable operation.

When the magnetron is subjected to very difficult load conditions, flash-overs or short-circuits may occur in the microwave circuits. In order to prevent this from damaging the magnetron or nearby located circuits, the current is monitored using the transformer Tr3 which provides a signal via the diode D4 to the pulse width control circuit PWC2 to temporarily deactivate the switching of the transistors T2 and T3 when a flash-over or a transient occurs in the magnetron. The energy which at that moment may be supplied to the magnetron during a half period is limited to the energy which is stored in the capacitor C3 or C4 since the previous half period, depending on during which half period of the switching that the break is made.

The filament current control block S3 in Fig. 7 com¬ prises a control unit FCC. The control unit FCC conven¬ tionally provides a filament current supplied through the cathode of the magnetron VI. The control unit FCC recei- ves its power supply from the power control block SI, but may be designed to use a control function of its own, independent of the operation of the power controller, for controlling the magnitude of the filament current. The operation of the control, unit FCC may also be based upon different parameters received from the processing/control block S4.

The processing/control block S4 comprises a micro¬ processor CPU which provides control and monitoring of the operation of the power supply unit. The microproces¬ sor sends control signals to, and monitors (not shown) , the other function blocks. More specifically, the micro¬ processor is connected to the input A of the multiplier/- divider M/D in the control circuit R for providing the signal Pset thereto, said signal indicating the desired power level. The microprocessor CPU is also connected to the pulse width control circuit PWC2 in the high voltage converting block S2 for controlling the above mentioned slow-start function which is used during slow start or build up of a steady oscillation in the high voltage converting step S2. Furthermore, the microprocessor CPU is connected to the control unit FCC of the filament current control block S3 for providing control signals thereto.

Moreover, the microprocessor may communicate exter¬ nally using an analogue control voltage (for controlling the power of the magnetron) or using a digital, prefer¬ ably serial, interface (not shown) which may be bi-direc¬ tional. According to an embodiment, the interface may use the power supply as a communication link.

Even though the invention has been described with reference to specific exemplifying embodiments, it is understood by those skilled in the art that different modifications, alterations and combinations of the diffe¬ rent described embodiments may be made within the scope of the invention, as defined in the accompanying claims. For example, it is understood that control of the operation of the power supply unit may be realized in many different ways, and hence that the invention is not limited to the embodiment described above.

Although the invention has been described in rela- tion to the supply of power to magnetrons, it is contem¬ plated that certain aspects of the inventions could also be applicable to other types of power supply situations.