SELF OSCILLATING PROPORTIONAL DRIVE ZERO VOLTAGE SWITCHING POWER SUPPLY The invention relates to a switch mode power supply (SMPS).
A SMPS produces a regulated DC output by varying the timing at which a switch such as a transistor turns on and off for coupling a raw supply voltage to the primary winding of a transformer. The regulated output is obtained by rectifying the voltage produced on a secondary winding of the transformer, and is fed back to a drive or control circuit that triggers switching.
In a resonant or tuned zero voltage switching type switch mode power supply, a chopper transistor switch on the primary winding of a chopper transformer is coupled in parallel with a clamping diode and in series with the primary winding. A capacitor is coupled to the primary winding to form a resonant circuit with the inductance of the primary winding. A substantially sinusoidal resonant voltage is generated across the inductance, during a portion of each period. At the end of a half cycle of oscillation, the diode conducts and clamps the collector of the transistor switch at zero volts. Switching the transistor on occurs when the collector voltage is at zero voltage for minimizing switching losses. The resonant circuit reduces the voltage across the transistor when the transistor is switched off, as compared to a comparable non-resonant switch mode supply.
In a forward converter, power from the unregulated input voltage is coupled to the output during conduction of the transistor switch. In a typical half-forward converter configuration of this type, the secondary winding of the transformer is coupled in series with a rectifying diode, an inductor or choke and a filter capacitor. The aforementioned series diode is coupled between the secondary winding and a catch diode. The cathodes of the two diodes are coupled to the series inductor or choke and then to the filter or output capacitor. An output supply voltage is developed in the filter capacitor. A buck effect is produced such that during forward conduction on
the primary winding, the series diode on the secondary winding conducts. The choke limits the rate of change of the current in the series diode.
When the transistor switch on the primary winding is turned off, the inductor on the primary-side resonant circuit reverses the voltage on the secondary winding of the transformer, causing the series diode to turn off. Instead, the catch diode provides a current path for conducting the current in the choke. A feedback signal is obtained from the voltage on the capacitor, for controlling the switching timings of the transistor switch.
A high power output switch mode power supply operating as, for example, a forward converter, requires a switching device that has a high current rating and a high breakdown voltage. A relatively inexpensive switching device for this application would be a single bipolar switching transistor. Bipolar transistors with high current capacity normally do not have high current gain, and require a high level of base current drive when conducting. Moreover, an even higher level of current drive is required when switching off, due to the need to remove the substantial base- emitter charge in such transistors, to cease conduction and reverse bias the base-emitter junction quickly.
A power MOSFET can be used as the switching device instead of a bipolar transistor, but at power outputs of 200 watts or more, the cost of power MOSFETs is substantially higher than that of bipolar transistors of comparable current capacity, and conduction losses are higher. It would be advantageous to configure a zero voltage switching SMPS optimally for operation using a bipolar switching transistor, preferably without the need for an unduly complex drive circuit. Causing the switching transistor in a switch mode power supply to switch at zero voltage minimizes switching losses and reduces the radio frequency radiation that would be produced by rapid changes of voltage in the switching transistor and rectifier diodes. Zero voltage switching does not require snubber circuits, reducing power consumption and RF noise, as well as power consumption and heat generation.
According to an inventive aspect, in a zero voltage switching SMPS, operating in a forward converter configuration, proportional drive base current is produced from the collector current. Advantageously, the proportional drive arrangement improves switching times and minimizes overdriving of the base of the switching transistor so that the transistor can be switched off quickly. The base of the switching transistor is driven at a current level that is proportional to the collector current of the transistor. The switching transistor is driven to saturation. According to another inventive aspect, a current transformer having a heavily loaded secondary winding is used to couple the switching transistor collector current to the base drive circuit. Advantageously, operating the transformer as current transformer enables utilizing a small core in the transformer. The base drive current thus can be kept proportional to the collector current to avoid unnecessarily charging the base-emitter junction during conduction, and to reduce the current drive needed to reverse bias the junction upon turn off.
In a SMPS, embodying an inventive feature, the switching transistor is turned off upon the collector current reaching a threshold level established by an error signal. Thereby, the peak current in the primary winding of the chopper transformer coupled to the switching transistor is controlled. In this way, the control circuit provides regulation in a current mode manner. Thus, the advantages of current mode regulation and the advantages of a tuned or resonant zero voltage switching switch mode power supply are obtained together.
As explained before, a resonant voltage is generated across the primary winding of the chopper transformer of a zero voltage switching SMPS, when the chopper transistor switch is turned off.
It may be desirable to apply the resonant voltage to the base of the chopper transistor for controlling the instant when the chopper transistor is turned on. The chopper transistor is turned on in zero voltage switching SMPS when the voltage across the chopper transistor is zero.
According to a further inventive feature, the resonant voltage is coupled to the base of the chopper transistor via the current transformer, for controlling the turn-on instant of the chopper transistor. Prior to turning on the chopper transistor, the transformer-coupled resonant voltage maintains the chopper transistor non-conductive.
According to yet another inventive feature, the current transformer provides self oscillation. Self oscillation may be desirable because no external oscillator is necessary for controlling the switching in the chopper transistor.
A switch mode power supply with zero voltage switching, embodying an inventive feature, includes a supply inductance. A source of an input supply voltage is provided. A transistor periodically applies the input supply voltage to the supply inductance to generate in the supply inductance a first ramping signal coupled to the transistor in a positive feedback manner for providing proportional drive, during a first portion of a given period when the transistor is switched on. A control circuit generates a control signal coupled to the first transistor for controlling when the transistor is switched off. A capacitance is coupled to the supply inductance to form a resonant circuit for generating a resonant signal in the supply inductance when the transistor is switched off. The resonant signal is coupled to the transistor for switching on the transistor at an end of a second portion of the given period, to provide for zero voltage switching of the transistor. A rectifier is coupled to the supply inductance for generating an output supply voltage applied to a load circuit.
FIGURE 1 is a schematic diagram showing an exemplary embodiment of the circuit of the invention; and FIGURES 2a through 2d are timing diagrams showing voltages and currents at several points identified in the schematic diagram of FIGURE 1, through two switching cycles.
Referring to FIGURE 1 , an exemplary zero voltage switching forward converter or power supply 300, embodying an inventive feature, is shown. Power, for example, 200 watts, is supplied to loads 303 and 302 coupled to secondary windings T1W2 and
T1W3 of a chopper transformer Tl , respectively, during the "on" or conducting time of a switching transistor Ql . Switching or chopper NPN transistor Ql operates as a switch in series with a primary winding TlWl of chopper transformer Tl for conducting current from an input supply, direct current (DC) voltage RAW B+. Transformer Tl can be considered a supply or coupling transformer. A current transformer T2, which can be considered a control transformer, supplies base current drive to the switching transistor Ql and its control circuit. Supply or coupling transformer Tl can serve, for example, as an isolation transformer separating the hot and cold grounds in a consumer electronic apparatus. Voltage RAW B+ can be derived, in that case, from a bridge rectifier that rectifies a mains supply voltage, and is coupled to a filter capacitor (not shown). The input voltage can also be provided from some other direct current source.
Also in series with transistor Ql is a current sensing resistor R7. A damper diode D8 clamps the collector of transistor Ql relative to ground, as explained later on. Capacitor C8 is coupled in parallel with diode D8 and also to primary winding TlWl . A resonant circuit 301 is formed, comprising capacitor C8, a reflected capacitance CSEC, an inductor Lres, primary winding TlWl and a primary winding T2W1 of transformer T2. Primary winding TlWl is coupled in series with the primary winding T2W 1 of current transformer T2 which provides base current drive for transistor Ql , as explained below.
The resonant circuit produces a resonant voltage VQl when transistor switch Ql is turned off, and in particular causes the voltage VQl across transistor Ql (and on capacitor C8) to rise to a peak and then fall to zero in a substantially sinusoidal half wave.
After resonant voltage VQl becomes zero, diode D8 clamps voltage VQl to ground potential. Transistor Ql is then switched on again at zero volts to provide for zero voltage switching.
One secondary winding T1W3 of transformer Tl is coupled to an anode of a rectifier diode DOUT3, the cathode of which is coupled to a filter capacitor CFILTER3. Winding T1W3 is coupled
via a low impedance current path, during forward conduction operation, to filter capacitor CFILTER3 and to load 302. Unlike some prior art circuits, no choke is provided in series with secondary winding T1W3, whereby the impedance in the current path between secondary winding T1 W3 and filter capacitor CFILTER3 is, advantageously, kept low.
Similarly, a second secondary winding T1W2 is coupled through rectifier diode DOUT2 to filter capacitor CFILTER 2 to provide output voltage REG B+. Secondary winding T1W2 is also coupled via a low impedance current path to filter capacitor CFILTER2. Likewise, the current path has a low impedance because no choke is used.
Capacitor CSEC can be included in one or both of the secondary winding circuits T1W2 and T1W3 in parallel with the winding on the anode of the respective rectifier. Capacitor CSEC is transformer coupled to winding TlWl forming a part of the resonant capacitance, as indicated before, of resonant circuit 301. Control of the duty cycle of transistor switch Ql is based on, for example, sensing output voltage REG B+ directly, rather than output voltage U. An error amplifier A is responsive to the voltage REG B+, and can include, for example, a comparator having inputs coupled to output voltage REG B+ and to a voltage divider providing a predetermined threshold. Error amplifier A is optically coupled through an opto-coupler μl to control a triggering level or threshold of a comparator transistor Q3.
Advantageously, each of winding T1W2 and T1W3 is tightly coupled to primary winding TlWl in transformer Tl in a manner to reduce leakage inductance. The leakage inductance LL is approximately 1.5 microHenry. Whereas, each of the secondary windings is coupled to its respective load via a corresponding low in impedance current path. Consequently, the voltages developed in secondary windings T1W2 and T1W3 tend to track each other. This is possible due to the absence of a conventional choke in series with each of the secondary windings. Advantageously, the inductance Lres on the primary side of transformer Tl is transformer coupled to limit the rate of change
of each of currents IDOUT3 and IDOUT2 in the current paths that include diodes DOUT3 and DOUT2, respectively, during forward conduction. Thus, advantageously, no choke is required to be coupled in series with any of windings T1W2 and T1W3. Advantageously, inductance Lres is shared in common with each of windings T1W2 and T1W3. Maintaining each of winding T1W2 and T1W3 tightly coupled to primary winding T lWl simplifies the design of transformer Tl and reduces losses in transformer Tl . In addition to the optically coupled signal from opto-coupler μl , the base drive circuits are coupled to current sensing resistor R7 in series with switching transistor Ql . When transistor Ql is turned on, as explained later on, the voltage across resistor R7, which is proportional to the current level in transistor Q l , is coupled to the base of comparator transistor Q3. Transistor Q3 forms a regenerative latch with another transistor Q2, which is coupled back to the base of switching transistor Ql and to the secondary winding T2W2 of current transformer T2.
Advantageously, the current provided in secondary winding T2W2 is proportional to the current in the primary winding T2W1 of transformer T2, which is in series with winding TlWl of transformer Tl and switching transistor Ql . Therefore, the base current drive signal iB varies approximately linearly with the collector current iQl . Advantageously, over-driving of the base of transistor Ql is prevented by a proportional drive technique. The same current transformer functions to provide the advantages of proportional drive, self-oscillation and zero voltage switching in forward type voltage converter 300, as explained later on. Transistor Q3 of the regenerative latch, comprising transistors Q2 and Q3, functions as a comparator. The current- representative voltage on resistor R7 is coupled to charge capacitor C7 through resistor R8, and the voltage on capacitor C7 is coupled to the base of transistor Q3 through a small resistor R9. When the voltage at the base of transistor Q3 exceeds the voltage at its emitter sufficiently to forward bias the base-emitter junction, transistor Q3 conducts and the latch formed by
transistors Q2 and Q3 draws current away from the base of switching transistor Ql . The voltage at the emitter of transistor Q3 is developed from the charge in capacitor C6. The emitter voltage in capacitor C6 is limited to a forward diode drop by diode D7, coupled to ground. The charge in capacitor C6 is replenished while transistor Q3 is conducting and is drained by opto-coupler μl when it conducts in response to an output signal of error amplifier A.
The collector of NPN transistor Q3 is coupled to the base of PNP transistor Q2 and the collector of transistor Q2 is coupled to the base of transistor Q3, forming a regenerative switch. A control voltage coupled to the control terminal (i.e., the base) of switching transistor Ql is developed at the emitter of transistor Q2, which forms an output of the regenerative switch arrangement and is coupled to the base of transistor Ql via a resistor R5.
Secondary winding T2W2 of current transformer T2 provides a drive current supply for switching transistor Ql . The voltage across winding T2W2 is an alternating current (AC) voltage, produced when switching transistor alternately conducts and is turned off.
In accordance with an inventive feature, when transistor Ql is turned on, transformer T2 provides proportional drive to transistor Q l for maintaining transistor Q l in saturation without over-driving transistor Ql . On the other hand, when transistor Ql is non-conductive, resonant voltage VQl at the collector of transistor Ql is coupled to the base of transistor Ql to maintain transistor Ql nonconductive.
FIGURES 2a through 2d illustrate certain voltage and current signals identified in FIGURE 1 , through two oscillation cycles. Power on, start-up of the oscillation cycles occurs due to current flowing through resistor R4. Resistor R4 coupled in series with resistor R2 couples the RAW B+ supply to the base of switching transistor Ql . Resistor R4 is large, and provides a small amount of start-up base current drive to transistor Ql . As transistor Ql conducts, however, current transformer T2 causes a current to flow in secondary winding T2W2 which is proportional to the
current in primary winding T2W1 , as a function of their turns ratio, for example 20% for a turns ratio of 2: 10. Diode Dl in series with secondary winding T2W2 couples this current via resistor R2 to the base of transistor Ql . The added base drive current sustains saturation for the added collector current in a regenerative manner, causing the base current to increase in proportion to the increase in collector current. Transistor Ql saturates and the collector current continues to flow until base drive current is removed by action of transistors Q2 and Q3. When the voltage across current sensing resistor R7 is sufficient to cause transistor Q3 to conduct, triggering current is provided at the base of transistor Q2, which also conducts and adds to the voltage at the base of transistor Q3, producing additional drive current in transistor Q3 and also operating in a regenerative manner to latch on. Resistor R3 and capacitor C4 provide proper biasing for transistor Q2. The low impedance of latched drive transistor Q2 quickly removes the base charge from the base of switching transistor Ql . The result is that transistor Ql is turned off. During the time that transistor Ql is conducting, positive current flows into the base through resistor R2 and capacitor C2, which causes capacitor C2 to charge to several volts, more positive on the terminal coupled to resistors R4 and R5 and less positive at the base of transistor Ql . When transistors Q2 and Q3 latch, they provide a low impedance path to ground, causing the voltage on capacitor C2 to apply a negative bias to the base of transistor Ql . This improves the speed at which transistor Ql switches off by quickly removing the base charge in transistor Ql .
Diodes D4 and D5 are coupled in series with one another and the emitter of switching transistor Ql . When transistor Ql is conducting, there is a forward biased voltage drop across diodes D4, D5, namely about two volts. Capacitor C5, in parallel with series diodes D4, D5, charges to this voltage. The charge on capacitor C5 provides additional negative bias during turn off of transistor Ql , especially during start-up, when capacitor C2 may not be fully charged. In this way, sufficient negative bias is
applied to the base of transistor Ql to ensure quick turn off. Diode D6 and resistor R6, which are coupled between the collector of transistor Q2 and current sensing resistor R7, shunt some of the reverse base current to resistor R7, which is low in impedance, for example a fraction of an ohm. This shunting reduces the tendency to overdrive the base of transistor Q3, which would otherwise cause excessive storage time and poor switching performance.
In accordance with a further inventive feature, after transistor Ql is turned off, transformer T2 winding T2W2 produces a negative voltage across series coupled diode D2 and resistor Rl . Drive transistors Q2 and Q3 remain latched until the current flowing through them drops below a threshold needed to keep them regeneratively latched. Thereafter, the voltage across series coupled diode D2 and resistor Rl keeps transistor Q l from conducting.
Eventually, the resonant action of resonant circuit 301 causes the base-emitter voltage to reverse polarity. When the voltage at the base of switching transistor Ql increases to a sufficient magnitude, current begins to flow in the base of transistor Ql , producing collector current that grows regeneratively as discussed, beginning the next cycle. Collector current iQl in transistor Ql begins flowing when collector voltage VQl is at zero volts. Thereby, zero voltage switching is obtained. In accordance with yet another inventive feature, current transformer T2 provides for self-oscillations. In the circuit coupled to secondary winding T2W2 of transformer T2, diode D2 and resistor Rl limit the negative voltage developed during the off time of transistor Ql . Because diode D2, resistor Rl and capacitor Cl form a low impedance, transformer T2 operates as a current transformer during the turn off interval. Diode Dl provides a current path for the forward drive current and also limits the charging of capacitor Cl , in parallel with diode Dl, to the forward voltage developed across diode Dl when conducting. Diode Dl , resistor R2 and the base-emitter junction of transistor Ql form a low impedance during the turn on interval of transistor Ql . Thus, transformer T2 operates as a current transformer.
Advantageously, by operating as a current transformer, transformer T2 need not have to store large magnetic energy and can have a small core.
Negative base current, which is blocked by diode Dl , flows through capacitor Cl during the turn off interval of transistor Ql . Diode D3 and capacitor C3 are coupled to rectify and filter the negative voltage produced by transformer T2, and provide a negative supply voltage coupled to the emitter of the phototransistor in opto-coupler μl . FIGURES 2a-2d illustrate waveforms useful for explaining the operation of the tuned switch mode power supply circuit shown in FIGURE 1. Similar symbols are used to identify points or paths in the circuit of FIGURE 1 with their voltage and current signals in FIGURES 2a-2d. FIGURE 2a shows the voltage VQl (in a solid line) and the current iQl (dashed line) in the collector of transistor Ql . FIGURE 2b shows the voltage VB (dashed line) and current iB (solid line) at the base of transistor Ql . When positive base voltage VB becomes available, the base current iB and collector current iQl rise gradually until transistor Ql current iQl reaches a peak at about 8A. The rectifiers in the secondary windings conduct, during the forward conduction times of transistor Ql , shown by current iDOUT in FIGURE 2c.
Upon turn off, the base current drive is driven to reverse abruptly to a negative absolute value greater than its positive value, for example by a factor of two. During turn off of transistor Ql, the resonant voltage VQl at the collector of transistor Ql , which is also the voltage on capacitor C8, rises and then falls, resonantly. During the resonant cycle, after the voltage VQl on capacitor
C8 falls to zero, diode D8 clamps the voltage to near ground potential, conducting for a time as shown in FIGURE 2d until base and collector currents iB and iQl begin to increase.
The inventive tuned switch mode power supply as shown operates in a current mode control, on a current-pulse by current-pulse control basis. The current pulses iQl an iB in the
collector and base of transistor Ql , respectively, terminate when the collector current reaches the threshold level of transistor Q3 in FIGURE 1 , namely the level of current sensed by resistor R7 sufficient to raise the voltage at the base of transistor Q3 by more than the base-emitter forward bias voltage level over the voltage on capacitor C6. The charge on capacitor C6 is adjusted by conduction of the phototransistor of opto-coupler μl , responsive to signals from error amplifier A. In this manner the voltage is closely regulated on a current pulse basis. The inventive circuit responds to current and can instantaneously correct in a feed forward manner for input voltage variations on RAW B+, without the need to use the dynamic range of error amplifier A and without the delay of waiting for input voltage variations to appear at the output. In this way, both the advantages of current mode regulation and of a tuned switch mode power supply are obtained.
The secondary windings T1W2 and T1W3 are tightly coupled in transformer Tl to primary winding TlWl . The low impedance in the current path of each of conductive diodes DOUT2 and DOUT3 is interposed between the corresponding winding
T1W2 or T1W3, and the corresponding filter capacitor CFILTER2 or CFILTER3. Advantageously, because of the low impedance in each current paths, unsensed voltage U is regulated to a significant extent even though only voltage REG B+ is sensed in error amplifier A.