High-precision power amplifier based on resonant auxiliary commutation technology and parameter design method

Abstract

The application relates to a high-precision power amplifier based on resonance auxiliary commutation technology and a parameter design method, and belongs to the technical field of power amplifiers. The application is proposed in view of the problems of current zero distortion and dead zone effect of the existing PWM modulated switching power amplifier. The power amplifier comprises a capacitor C1, a capacitor C2, four MOSFET switching tubes S1, S2, S3 and S4, auxiliary inductors L d1 , d2 , o and filter capacitors C o , the switching tubes S1, S2, S3 and S4 form two bridge arms in parallel with branches formed by the capacitor C1 and the capacitor C2; the auxiliary inductors L d1 and the auxiliary inductor L d2 are connected in series between the two bridge arms; a connection point between the auxiliary inductors L d1 and the auxiliary inductor L d2 is taken as a C point, filter inductors L o and filter capacitors C o are connected in series between the C point and the connection points of the capacitor C1 and the capacitor C2 in sequence; and a load R is connected in parallel with the filter capacitors C o . The application can realize soft switching of the power amplifier in a full load range.

Description

TECHNICAL FIELD

[0001] The application relates to a high-precision power amplifier based on a resonant auxiliary commutation technology and a parameter design method, and belongs to the technical field of power amplifiers. BACKGROUND

[0002] A high-precision motion platform in a semiconductor manufacturing device needs to realize positioning precision of mu m or even nm, and therefore a motor on the motion platform needs to be equipped with a high-precision power amplifier. In order to prevent precise components from swelling and deforming due to heat, the temperature rise of the power amplifier must be limited. A traditional linear power amplifier has high precision, but the loss reaches 40%, which leads to serious heat dissipation, and the linear power amplifier is not suitable for a high-precision motion platform. Although a hybrid power amplifier has a smaller loss than the linear power amplifier, the hybrid power amplifier still cannot meet the low heat dissipation requirement of the semiconductor manufacturing device. Moreover, the hybrid structure of the linear power amplifier and the switching power amplifier increases the complexity of the circuit, and the series or parallel structure of the two may also have a small signal stability problem, so the hybrid power amplifier is not recommended to be used. The loss of the switching power amplifier can be as low as within 5%, and the heat dissipation is much smaller than that of the linear power amplifier and the hybrid power amplifier, so the switching power amplifier is the only suitable choice.

[0003] For the switching power amplifier using PWM modulation, the disadvantages are that hard switching leads to large device loss, serious heat dissipation, current zero-crossing distortion and dead-time effect, and high output current distortion degree. However, the advantages are that the switching is at a fixed frequency, and the control system delay is not sensitive. For the switching power amplifier using hysteresis control, the disadvantages are that the switching is at a variable frequency, and the control system delay is sensitive. However, the advantages are that the switching is soft switching in a full load range, the device heat dissipation is small, there is no current zero-crossing distortion and dead-time effect, and the output current distortion degree is low. SUMMARY

[0004] In view of the current zero-crossing distortion and dead-time effect problems of the existing PWM-modulated switching power amplifier, the application provides a high-precision power amplifier based on a resonant auxiliary commutation technology and a parameter design method.

[0005] The high-precision power amplifier based on the resonant auxiliary commutation technology comprises a capacitor C1, a capacitor C2, four MOSFET switching tubes S1, S2, S3 and S4, an auxiliary inductor L d1 , an auxiliary inductor L d2 , a filter inductor L o and a filter capacitor C o .

[0006] The capacitor C1 and the capacitor C2 are connected in series between a positive electrode of a power supply and a negative electrode of the power supply.

[0007] The positive electrode of the power supply is connected to the drain of the switch tube S1, the source of the switch tube S1 is connected to the drain of the switch tube S2, and the source of the switch tube S2 is connected to the negative electrode of the power supply; the positive electrode of the power supply is also connected to the drain of the switch tube S3, the source of the switch tube S3 is connected to the drain of the switch tube S4, and the source of the switch tube S4 is connected to the negative electrode of the power supply;

[0008] The connection point between the source of switch tube S1 and the drain of switch tube S2 is set as point A, the connection point between the source of switch tube S3 and the drain of switch tube S4 is set as point B, and an auxiliary inductor L is connected in series between points A and B. d1 and auxiliary inductor L d2 ; Set the auxiliary inductor L d1 and auxiliary inductor L d2 The connection point between them is point C, and the filter inductor L is connected in series between point C and the connection point of capacitor C1 and capacitor C2. o and filter capacitor C o ; Load R and filter capacitor C o in parallel.

[0009] According to the high-precision power amplifier based on the resonance auxiliary commutation technology of the present invention, the auxiliary inductor L d1 and auxiliary inductor L d2 The inductance values ​​are equal.

[0010] According to the high-precision power amplifier based on the resonance-assisted commutation technology of the present invention, the auxiliary inductor L d1 and auxiliary inductor L d2 The two half-bridge arms formed by the switch tubes S1 to S4 are coupled so that the two half-bridge arms serve as resonant auxiliary commutation branches.

[0011] According to the high-precision power amplifier based on the resonant auxiliary commutation technology of the present invention, SPWM modulation is used for the two half-bridge arms, so that the two half-bridge arms are staggered and run in parallel to form a differential mode current, and the auxiliary inductor L d1 , auxiliary inductor L d2 It resonates with the junction capacitance of the four MOSFET switches S1 to S4 to achieve zero voltage switching in continuous current mode.

[0012] The present invention also provides a parameter design method for a high-precision power amplifier based on resonance-assisted commutation technology, and performs parameter design on the high-precision power amplifier based on resonance-assisted commutation technology;

[0013] For the filter inductor L o and filter capacitor C o The filter consists of a filter inductor L o and filter capacitor C o Need to meet:

[0014]

[0015] 2π·f o,max ·L o,eff ·I o ≤k v ·U o ,

[0016] 2π·f o,max ·C o ·U o ≤k i ·I o ,

[0017] In the formula, L o,eff is the equivalent value of the filter inductance L o , k f is the cutoff frequency coefficient, f o,max is the maximum value of the actual output frequency of the power amplifier; I o is the effective value of the output current, k v is the maximum inductance voltage drop coefficient, U o is the effective value of the output voltage, k i is the maximum capacitance shunt coefficient.

[0018] According to the parameter design method of the high-precision power amplifier based on the resonant auxiliary commutation technology of the application, the design values of the auxiliary inductance L d1 and the auxiliary inductance L d2 need to meet the following conditions:

[0019] The peak current flowing through the auxiliary inductance L d1 and the auxiliary inductance L d2 is not greater than the soft switching current threshold of the switch tubes S1 to S4; the switch tubes S1 to S4 keep soft switching at the maximum load.

[0020] According to the parameter design method of the high-precision power amplifier based on the resonant auxiliary commutation technology of the application, the inductance values of the auxiliary inductance L d1 and the auxiliary inductance L d2 are both represented as L d , and the peak current ΔI d of L dm needs to meet the following conditions:

[0021]

[0022] In the formula, the peak current ΔI dm is the peak-to-peak value of the differential mode current, t is time, U dc is the DC bus voltage between the positive electrode of the power supply and the negative electrode of the power supply, m is the modulation ratio, and T s is the switching period.

[0023] The parameter design method of the high-precision power amplifier based on the resonant auxiliary commutation technology according to the application keeps the switch tubes S1 to S4 in soft switching at the maximum load, L d The following conditions need to be met

[0024]

[0025] i s_dx The current flowing through the auxiliary inductor L d1 or the auxiliary inductor L d2 , C oss is the equivalent junction capacitance of the MOSFET switch tube, I dx,ZVS is the soft switching current threshold; Q oss is the charge amount stored by C oss when the switch tube bears the voltage U ds , U ds is the positive bus voltage.

[0026] The power amplifier according to the application can be used in the precise driving of voice coil motors. It is realized based on the auxiliary commutated resonant pole (ACRP) technology, can realize the soft switching of the power amplifier in the full load range, eliminates the current zero-crossing distortion and dead-time effect, and is not sensitive to the delay of the control system. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 is the circuit structure diagram of the high-precision power amplifier based on the resonant auxiliary commutation technology according to the application; in the diagram, C oss1 is the junction capacitance of the switch tube S1, C oss2 is the junction capacitance of the switch tube S2, C oss3 is the junction capacitance of the switch tube S3, C oss4 is the junction capacitance of the switch tube S4, i d1 is the current flowing through the auxiliary inductor L d1 , i d2 is the current flowing through the auxiliary inductor L d2 , i Lo is the current flowing through the filter inductor L o , i o is the load current.

[0028] Figure 2 is the typical waveform diagram of the resonant auxiliary commutation soft switching of the power amplifier in the parameter design method of the high-precision power amplifier based on the resonant auxiliary commutation technology according to the application; in the diagram, d S1_d ~ d S4_d are the drive signals of the switch tubes S1 to S4 containing dead-time, U A is the voltage at point A, UB is the voltage at point B;

[0029] Figure 3 The dead zone effect diagram of the power amplifier of the present invention adopts SPWM modulation; S1 to d S4 is the driving signal of the switches S1 to S4 without dead zone, u′ A and u′ B is the ideal value of the voltage at points A and B when there is no dead zone, u′ AR and u′ BR is the actual value of the voltage at points A and B when the dead zone is included, u' Ae and u' Be Represents the error between the ideal value and the actual value, that is, u' Ae =u' A -u' AR , u' Be =u' B -u' BR ;U' eq is the average error voltage at point C;

[0030] Figure 4 yes Figure 3 A1 is the positive voltage error, A2 is the negative voltage error, and T r1 is the negative resonance time, i.e. u' BR Fall time, T r2 is the forward resonance time, i.e. u' BR Rise time;

[0031] Figure 5 The filter inductor L o and filter capacitor C o Schematic diagram of the phase relationship of the filter parameters at the fundamental frequency; in the figure, u Lo is the filter inductor L o voltage;

[0032] Figure 6 yes Figure 5 Equivalent circuit diagram of

[0033] Figure 7 It is a schematic diagram of the design space of the filter; in the figure, f c is the filter cutoff frequency;

[0034] Figure 8 This is a schematic diagram of the fitting results of the junction capacitance when the switch tube model is C3M0060065K;

[0035] Figure 9 It corresponds to Figure 8 Schematic diagram of charge calculation results; Q ossThe output charge quantity of the junction capacitor of the switch tube;

[0036] Figure 10 is the fitting result diagram of the junction capacitor when the switch tube model is IDL08G65C5; in the diagram, C D is the junction capacitor of the diode, U R is the reverse breakdown voltage of the diode;

[0037] Figure 11 is the charge quantity calculation result diagram corresponding to Figure 10 ; in the diagram, Q D is the output charge quantity of the junction capacitor of the diode;

[0038] Figure 12 is the design space diagram of L d ; in the diagram, I neg is the negative soft switching current under the maximum load condition;

[0039] Figure 13 is the experimental waveform diagram of the load current and the currents of the three inductors in the embodiment; in the diagram, i Lo1 is the current flowing through the filter inductor Lo1 in the first module in Figure 23 ; in the diagram, i Lo2 is the current flowing through the filter inductor Lo2 in the second module in Figure 23 ; in the diagram, i Lo3 is the current flowing through the filter inductor Lo3 in the third module in Figure 23 ; in the diagram, i

[0040] Figure 14 is the experimental result diagram of the soft switching in the embodiment; in the diagram, u ds2 is the drain-source voltage of the switch tube S2, u gs2 is the gate-source voltage of the switch tube S2;

[0041] Figure 15 is the experimental waveform diagram of the voltage error in the embodiment; in the diagram, u gs1 is the gate-source voltage of the switch tube S1, u A is equivalent to U A ;

[0042] Figure 16 is the current self-balancing experimental result diagram of the auxiliary inductor L d1 and the auxiliary inductor L d2 in the embodiment;

[0043] Figure 17 is the average error voltage U e schematic diagram under different dead zones in the embodiment;

[0044] Figure 18 is the average error voltage schematic diagram under different switching frequencies in the embodiment; in the diagram, f sThe switching frequency is 250 kHz, the load current THD is 0.5% at 25 Hz, and the load current THD is 0.5% at 250 kHz.

[0045] Figure 19 is a schematic diagram of the average error voltage under different loads in the embodiment;

[0046] Figure 20 is a schematic diagram of the load current THD experiment results when the input voltage is 300 V, the load winding is 30 mH+4.8 Ω, f s =250 kHz, f o =25 Hz, T d =100 ns and 150 ns;

[0047] Figure 21 is a schematic diagram of the load current THD experiment results when the input voltage is 300 V, the load winding is 30 mH+4.8 Ω, f s =250 kHz, f o =25 Hz, T d =100 ns; the load winding is 4.8 mH+0.9 Ω, f s =250 kHz, f o =250 Hz, T d =100 ns;

[0048] Figure 22 is a schematic diagram of the load current THD experiment results when the input voltage is 100 V, the load winding is 30 mH+4.8 Ω, T d =100 ns, f o =25 Hz; when f s =250 kHz, L d =15 mH; when f s =62.5 kHz, L d =60 mH;

[0049] Figure 23 is a schematic diagram of the high-precision power amplifier described in the present application being staggered in parallel in multiple modules;

[0050] Figure 24 is a power amplifier operation mode diagram corresponding to each time point in Figure 2 ; wherein (a) is the operation mode diagram of stage 1; (b) is the operation mode diagram of stage 2; (c) is the operation mode diagram of stage 3; (d) is the operation mode diagram of stage 4; (e) is the operation mode diagram of stage 5; (f) is the operation mode diagram of stage 6; (g) is the operation mode diagram of stage 7; (h) is the operation mode diagram of stage 8; (i) is the operation mode diagram of stage 9; (j) is the operation mode diagram of stage 10; (k) is the operation mode diagram of stage 11; (l) is the operation mode diagram of stage 12

[0051] Figure 25 is an equivalent circuit diagram of the power amplifier in the resonant and diode clamping mode according to the present application;

[0052] Figure 26 is an auxiliary inductance L d1 auxiliary commutation soft switching power amplifier state plane trajectory diagram;

[0053] Figure 27 is an auxiliary inductance L d2 auxiliary commutation soft switching power amplifier state plane trajectory diagram. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0055] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0056] The present application will be further described below in combination with the drawings and specific embodiments, but is not limited to the present application.

[0057] DETAILED DESCRIPTION Figure 1 According to the present application, a high-precision power amplifier based on resonant auxiliary commutation technology is provided, which comprises a capacitor C1, a capacitor C2, four MOSFET switch tubes S1, S2, S3 and S4, an auxiliary inductance L d1 , an auxiliary inductance L d2 , a filter inductance L o and a filter capacitor C o .

[0058] The capacitor C1 and the capacitor C2 are connected in series between the positive electrode of the power supply and the negative electrode of the power supply.

[0059] The positive electrode of the power supply is connected to the drain of the switch tube S1, the source of the switch tube S1 is connected to the drain of the switch tube S2, and the source of the switch tube S2 is connected to the negative electrode of the power supply. The positive electrode of the power supply is also connected to the drain of the switch tube S3, the source of the switch tube S3 is connected to the drain of the switch tube S4, and the source of the switch tube S4 is connected to the negative electrode of the power supply.

[0060] The connection point between the source of the switch tube S1 and the drain of the switch tube S2 is taken as point A, and the connection point between the source of the switch tube S3 and the drain of the switch tube S4 is taken as point B. The auxiliary inductance L d1 and the auxiliary inductance L d2 are connected in series between point A and point B.d1 and the connection point between the auxiliary inductor L d2 and the auxiliary inductor L o and the filter capacitor C o ; the load R and the filter capacitor C o are connected in parallel.

[0061] In the embodiment, the auxiliary inductor L d1 and the auxiliary inductor L d2 have the same inductance value.

[0062] Further, the embodiment adopts the resonant auxiliary commutation technology, and the two half-bridge arms formed by the switching tubes S1-S4 are coupled by adding the auxiliary inductor L d1 and the auxiliary inductor L d2 , so that the two half-bridge arms are resonant auxiliary commutation branches, as shown in Figure 1 .

[0063] The SPWM modulation is used for the two half-bridge arms, so that the two half-bridge arms are operated in staggered parallel connection to form a differential mode current, and the auxiliary inductor L d1 , the auxiliary inductor L d2 and the junction capacitances of the four MOSFET switching tubes S1-S4 are resonated to realize zero voltage switching (ZVS-ON) in a continuous current mode (CCM).

[0064] In the embodiment, the filter auxiliary inductor L d1 and L d2 have passed zero before the switching tube (SiC MOSFET) is turned off, so that the current is not clamped to 0 by the anti-parallel diode of the switching tube, and the current zero-crossing distortion of the traditional half-bridge power amplifier is overcome. In addition, the resonant auxiliary commutation process eliminates the dead zone effect, and only a slight dead voltage error caused by the non-ideal rising edge and falling edge of the SiC MOSFET drain-source voltage exists, as shown in Figure 3 and Figure 4 .

[0065] The amplifier described in the embodiment is compatible with SPWM modulation, the switching frequency is fixed, and the control system hardware and software delay is not sensitive. Therefore, a relatively complex control algorithm can be used, and a delta-sigma ADC sampling chip with high precision but large delay can also be used.

[0066] Specific implementation two, as shown in Figures 1 to 12 , the application also provides a parameter design method of a high-precision power amplifier based on a resonant auxiliary commutation technology, which is used for parameter design of the high-precision power amplifier based on the resonant auxiliary commutation technology.

[0067] For the filter composed of filter inductance L o and filter capacitance C o , firstly, the filter design space needs to be established. For the filter, its design needs to consider two requirements, one is to meet the bandwidth requirement of the power amplifier; the second is that the values of filter inductance voltage drop u Lo and filter capacitance current i Co cannot be too large, that is, the reactive power needs to be limited.

[0068] Filter inductance L o and filter capacitance C o need to meet:

[0069] For the bandwidth, the following limiting conditions should be met:

[0070]

[0071] After equivalent processing on u sw,eff , the equivalent input voltage of the filter is u sum , and the total input current is i o . Figure 5 is the phase relationship diagram between load current i o , load voltage u sw,eff , u sum and i Co . Figure 6 is the equivalent circuit after multi-module interleaved parallel connection. When the output frequency increases, the reactive power absorbed by the filter capacitance and the filter inductance will also increase. If the parameter design is not reasonable, excessive reactive power may cause the switching device to burn out. Therefore, the filter parameter design needs to consider i Lo and u o,max under the maximum output frequency, that is, the following two limiting conditions need to be considered:

[0072] 2π·f o,eff ·L o ·I v ≤k o ·U o,max ,

[0074] 2π·f o ·C o ·U i ≤k o ·I o,eff ,

[0075] In the formula, L o is the equivalent value of filter inductance L f , k o,max is the cutoff frequency coefficient, generally taking an empirical value of 4; f o is the maximum value of the actual output frequency of the power amplifier; Io is the effective value of the output current, k v is the maximum inductance voltage drop coefficient, U o is the effective value of the output voltage, k i is the maximum capacitance shunt coefficient. k v and k i The values ​​can be 15% and 33% respectively.

[0075] In summary, we can get Figure 7 The filter design space shown in Figure 1. The area that meets the bandwidth constraint is the lower left of the blue hyperbola, which satisfies the maximum u Lo The area with the constraint condition should be on the left side of the yellow straight line, satisfying the maximum i Co The area of ​​constraints should be below the orange straight line, and the final filter design space is the intersection of the three-phase constraints.

[0076] The ideal filter parameter value point should be located in the upper right part of the filter design space. At the same time, the influence of the component standard value should also be considered. For example, the capacitance value is usually 220nF or 330nF. Take the following indicators as an example: ① DC bus voltage 2U dc =300V;②output current effective value I o The peak value reaches 10A (resistive load); ③ switching frequency 250kHz; ④ target bandwidth 20kHz. The final filter parameters are the point marked by the red star (L o =10μH, C o =220nF).

[0077] Furthermore, the auxiliary inductor parameters are designed:

[0078] Auxiliary inductor L d1 and auxiliary inductor L d2 The design value must meet the following requirements:

[0079] Under the premise of a given input voltage level, switching frequency and filter inductance, the current flowing through the auxiliary inductor L d1 and auxiliary inductor L d2 The peak current is not greater than the soft switching current threshold of the switch tubes S1 to S4, otherwise the switch tubes are easily damaged; at the maximum load, the switch tubes S1 to S4 still maintain soft switching.

[0080] For the first condition: set the auxiliary inductor L d1 and auxiliary inductor L d2 The inductance value is expressed as L d , then L d Peak current ΔI dm The following conditions must be met:

[0081]

[0082] where ΔI is the peak current dm is the peak-to-peak value of the differential mode current, which must be within the safe operating area (SOA) of the SiC MOSFET; t is time, U dc is the DC bus voltage between the positive and negative power supply, m is the modulation ratio, T s is the switching period.

[0083] For the second condition, the switching tubes S1-S4 remain soft-switched at maximum load, L d The following conditions need to be met:

[0084]

[0085] i s_dx is the current flowing through the auxiliary inductor L d1 or the auxiliary inductor L d2 , C oss is the equivalent junction capacitance of the MOSFET switching tube, I dx,ZVS is the soft-switching current threshold; Q oss is the charge stored by C oss when the switching tube withstands voltage U ds , U ds is the positive bus voltage. Where C oss is the sum of the equivalent junction capacitances of the SiC MOSFET and SiC SBD.

[0086] Below, taking C3M0060065K and IDL08G65C5 switching tube devices as examples, Figure 8 and Figure 10 the fitting results of the junction capacitances of the two are given, Figure 9 and Figure 11 the calculation results of the output charges of the two are given. The equivalent output capacitance values of C3M0060065K and IDL08G65C5 at 300V can be calculated to be 153.3nF and 46.7nF respectively according to the junction capacitance calculation formula.

[0087] The working principle of the power amplifier described in the present application is described below in combination with Figure 2 , Figures 24 to 27 The typical waveforms of the power amplifier described in the present application and 12 specific operating modes within one switching period are shown in

[0088] Figure 2 and Figure 24 The switching tubes S1-S4 are SiC MOSFETs, L d1 and L d2 are auxiliary inductors, and the values of the auxiliary inductors are equal, i.e. Ld1 = L d2 = L d . L o is the filter inductor. C o is the filter capacitor, C oss1 ~ C oss4 is the equivalent junction capacitance of SiC MOSFET, and C oss1 = C oss2 = C oss3 = C oss4 = C oss . The input voltage is ±U dc , and the output voltage is u o . i d1 , i d2 , i Lo , and i o are the two auxiliary inductor currents, the filter inductor current, and the output load current, respectively. S1, S2, and L d1 form a bridge arm unit, and S3, S4, and L d2 form another bridge arm unit. The two bridge arms are identical in parameters and are auxiliary to each other.

[0089] Among the 12 modes, there are 4 resonant auxiliary stages. The equivalent junction capacitance of SiC MOSFET participating in resonance can be calculated as follows:

[0090]

[0091] where Q oss is the amount of charge stored by the junction capacitance when the switch tube withstands voltage U ds , and C oss (U ds ) can be obtained from the data manual.

[0092] During the resonance process, the characteristic impedance Z r and the angular frequency ω r are:

[0093]

[0094] It is emphasized here that in practical applications, SiC MOSFETs are usually connected in parallel with SiC SBDs, so C oss in formula (2) does not include the junction capacitance of SiC SBD by default. The junction capacitance of SiC SBD is calculated by formula (1), only the variable symbol is different, which is not repeated here.

[0095] The common-mode voltage u C output by the two bridge arms is:

[0096]

[0097] Where u A is the voltage at point A, u B is the voltage at point B;

[0098] Current i d1 and i d2 It can be decomposed into differential mode current i dm and common mode current i cm :

[0099]

[0100] The voltage drop across the filter inductor is given by u C and output voltage u o Together, according to Kirchhoff's current law, the filter inductor current i Lo with i d1 、i d2 The sum is expressed as:

[0101]

[0102] According to formulas (4) and (5), we can get i d1 、i d2 The current expression is:

[0103]

[0104] Next, the operating modes of the power amplifier are analyzed.

[0105] Phase 1: [t0, t1): such as Figure 2 and Figure 24 As shown in (a). At t0, S2 achieves ZVS-ON, after which S2 and S3 remain open, and the voltages at points A and B are pulled down and up to -U respectively. dc and +U dc .i d1 in-U dc Under the incentive of d2 in U dc The incentive for continued growth. cm In SiC MOSFET, L o and L d The voltage generated at both ends is equal to U dc Together they form u o ,i dm In L d1 and L d2 The voltage generated on the dc The difference between the on-state voltage drop of SiC MOSFET and the on-state voltage drop of SiC MOSFET. In summary, the equation group is obtained:

[0106]

[0107] where R ds,on is the on-state resistance of the switch tube.

[0108] Solving the equations of (6) and (7) can get the expressions of i d1 , i d2 in the stage [t0, t1):

[0109]

[0110] where I cm,t0 and I dm,t0 are the initial values of i cm and i dm at t0.

[0111] Stage 2: [t1, t2): i d1 has reversed before t1, S2 is off at t1, i d1 begins to charge the equivalent junction capacitor of S1 and discharge the equivalent junction capacitor of S2. In this stage, 2C oss and L d + L d || L o occur in series resonance. Since the amplitude variation of i Lo is far less than i d1 and i d2 , in order to facilitate calculation, it is assumed that i Lo remains unchanged during the resonance, and its current value is always equal to the initial value at t1. The initial values of this stage circuit are u S1 (t1) = 2U dc , u S2 (t1) = 0, i d1 (t1) = I d1,t1 , i Lo (t1) = I Lo,t1 . The corresponding current path is shown in Figure 24 (b), and the equivalent circuit is Figure 25 (a). The circuit equation set is as follows:

[0112]

[0113] where u S1 is the drain-source voltage of the switch tube S1, and u S2 is the drain-source voltage of the switch tube S2.

[0114] Solving can get i d1 , i d2 , u S1 and u S2 :

[0115]

[0116] The trajectory diagram corresponding to the resonance process described by this mode is Figure 26 The circular arc with a radius of r1, and the expression of the radius is:

[0117] r1=Z r |i s_dx |(11),

[0118] According to the geometric relationship shown in the trajectory diagram, r1≥2U dc to ensure that u S2 increases from 0 to 2U dc , and then C oss1 discharges to 0V at t2, realizing ZVS-ON of S1. Otherwise, C oss1 will not be able to discharge completely before S1 is turned on, resulting in the loss of ZVS-ON. The minimum i s dx The following conditions should be met:

[0119]

[0120] I dx,ZVS is the soft switching current threshold.

[0121] Phase 3: [t2, t3): After t2, i d1 flows through diode D1 (the antiparallel diode of S1), u S1 and u S2 are clamped to 0 and 2U dc by D1 respectively, and the equivalent circuit is Figure 25 (b). This ensures that ZVS-ON of S1 is realized at t3. After t2, i Lo begins to rise.

[0122] Phase 4: [t3, t4): At t3, S1 realizes ZVS-ON, and both S1 and S3 are in the on state, with the voltage at points A and B both being U dc , i d1 and i d2 are in opposite directions. The circuit equation set is as follows:

[0123]

[0124] Solving the equation set of formulas (6) and (13) gives the expressions of i d1 and i d2 as follows:

[0125]

[0126] where I cm,t3 and I dm,t3is the initial value of common mode current and differential mode current at t3.

[0127] Phase 5: [t4, t5): S3 is off at t4, L d +L d ||L o and 2C oss occurs series resonance, resonance current charges C oss3 , discharges C oss4 . The initial value of circuit in this phase is u S3 (t4) = 0, u S4 (t4) = 2U dc , i d2 (t4) = I d2,t4 , i Lo (t4) = I Lo,t4 . The current path in this phase is shown in Fig. Figure 24 (e), and its corresponding equivalent circuit is shown in Fig. Figure 25 (c). The circuit equation set is as follows:

[0128]

[0129] The expressions of i d1 , i d2 , u S3 and u S4 are obtained:

[0130]

[0131] where I d2,t4 is the initial value of i d2 at t4, u S3 is the drain-source voltage of switch S3, and u S4 is the drain-source voltage of switch S4.

[0132] The trajectory diagram corresponding to this mode is a circular arc with radius r4 in Fig. Figure 27 The expression of this radius is:

[0133]

[0134] When the resonance process ends, C oss4 is discharged to 0V.

[0135] Phase 6: [t5, t6): After t5, i d2 flows through diode D4 (anti-parallel diode of S4), u S3 and u S4 are clamped at 2U dc and 0 respectively, and its equivalent circuit is Figure 25(d). This ensures that ZVS-ON of S4 is achieved at t6. After t5, i Lo begins to decrease.

[0136] Stage 7: [t6, t7): ZVS-ON of S4 is achieved at t6. Since the voltages at A and B are U dc and -U dc respectively, the voltage across the auxiliary inductor is 2U dc , which makes i d1 increase, i d2 decrease, and i d2 changes direction before t7, so that the charging and discharging of the equivalent junction capacitances of S1 and S2 are completed during the resonance after t7, achieving soft switching. The variation processes of i d1 , i d2 and i Lo can be described by the following equations:

[0137]

[0138] Solving the equations shown in equations (6) and (18) gives the expressions of i d1 and i d2 as follows:

[0139]

[0140] where I cm,t6 and I dm,t6 are the initial values of the common-mode current and the differential-mode current at t6.

[0141] Stage 8: [t7, t8): S4 is turned off at t7, L d + L d || L o and the equivalent junction capacitances of S3 and S4 2C qoss resonate in series, and the resonance current discharges C oss3 and charges C oss4 . The initial values of the circuit are u S3 (t7) = 2U dc , u S4 (t7) = 0, i d2 (t7) = I d2,t7 , and i Lo (t7) = I Lo,t7 . The current path in this stage is shown in (h), and the corresponding equivalent circuit is shown in (e). The circuit equations are as follows: Figure 24 Figure 25

[0142]

[0143] The solutions of i d1 ​​i d2 u S3 and u S4 The expression is as follows:

[0144]

[0145] The corresponding trajectory of this stage is Figure 27 a circular arc with radius r3, whose radius is calculated in the same way as in stage 2.

[0146] Stage 9: [t8, t9): After t8, i d2 flows through diode D3 (anti-parallel diode of S3), u S3 and u S4 are clamped at 0 and 2U dc respectively, and its equivalent circuit is Figure 25 (f). This ensures that S3 achieves ZVS-ON at t9. After t8, i Lo begins to rise.

[0147] Stage 10: [t9, t 10 ): S3 achieves ZVS-ON at t9, S1 and S3 are both in the on state, and the voltages at points A and B are both U dc . i d1 and i d2 are in opposite directions. The circuit equation is:

[0148]

[0149] Solve the equations shown in equations (6) and (22):

[0150]

[0151] where I cm,t9 and I dm,t9 are the initial values of the common-mode current and differential-mode current at t9.

[0152] Stage 11: [t 10 , t 11 ): S1 is turned off at t 10 , L d + L d || L o and the sum of the equivalent junction capacitances of S1 and S2, 2C oss , resonate in series, and the resonance current charges C oss1 and discharges C oss2 . The initial values of this stage circuit are u S1 (t 10 ) = 0, u S2 (t 10 ) = 2U dc , id1 (t 10 )=I d1,t10 ,i Lo (t 10 )=I Lo,t10 The current path at this stage is as follows Figure 24 (k) shows that the corresponding equivalent circuit is Figure 25 (g). The circuit equations are as follows:

[0153]

[0154] Solve for i d1 、i d2 、u S1 and u S2 The expression:

[0155]

[0156] The trajectory corresponding to this stage is Figure 26 The radius of the arc with radius r2 is calculated in the same way as in stage 5.

[0157] Phase 12: 11 , t 12 ):t 11 After the moment i d1 Flows through diode D2 (anti-parallel diode of S2), u S1 and u S2 Clamped in 2U dc and 0, its equivalent circuit is Figure 25 (h). This ensures that S2 achieves ZVS-ON at time t0 in the next cycle. 11 After the moment i Lo Start to descend.

[0158] The complete operating mode of the power amplifier is thus obtained. In this embodiment, the number of bridge arms connected in parallel can be two or more. The specific number of bridge arms connected in parallel is selected based on actual usage requirements.

[0159] Embodiment: The effect of the present invention is described below by way of embodiment:

[0160] In this embodiment, capacitor C1, capacitor C2, four MOSFET switches S1, S2, S3, S4, auxiliary inductor L d1 , auxiliary inductor L d2 and filter inductor L o The structure is a module. In actual use, multiple modules can be used in parallel according to the use requirements to meet the soft switching current threshold and safety current threshold requirements.

[0161] Take the following indicators as an example: ① DC bus voltage 2U dc =300V;②Output current I o Reach peak value 10A (resistive load); ③ switching frequency 250kHz; ④ target bandwidth 20kHz. The above conditions ① and ② correspond to Figure 12 The area to the right of the green vertical line and the area below the purple curve, where the green vertical line corresponds to 10μH. The three red lines with different gradients are the different L values ​​under 1 to 3 modules. d The maximum soft switching current at this time can be calculated by the following formula; Obviously, at least two modules need to be staggered in parallel to meet the soft switching current threshold and safety current threshold requirements. The final number of staggered parallel modules is 3, as shown in Figure 23 As shown. Finally, the auxiliary inductor L is selected d =15μH.

[0162]

[0163] Where i neg is the negative soft switching current, N is the total number of bridge arms, and R is the load.

[0164] Figure 13 The 20kHz load current and filter inductor current experimental waveforms show the actual load current i of the prototype. o and the three filter inductor currents i Lo1 ,i Lo2 and i Lo3 The prototype's load current peaks at 10A, with a bandwidth of 20kHz. At this bandwidth, there is no current attenuation at all, so 20kHz represents the usable bandwidth, not the bandwidth defined by -3dB. The switching frequency is 250kHz (three modules in an interleaved parallel configuration, equivalent to a switching frequency of 1.5MHz), and the sinusoidal current exhibits no noticeable distortion.

[0165] Figure 14 The soft switching waveform of S2 under maximum load condition is given. ds2 After dropping to zero and staying there for a short time, u gs2 That is, the drain-source voltage drop process of S2 ends before its drain current rises, and S2 achieves ZVS.

[0166] Figure 15 The experimental waveform shows the slight voltage error caused by the non-ideal voltage rising and falling edges. A with u gs1 and u gs2 For example, u A The falling edge of u gs1 The falling edge of u A The rising edge of u gs2the falling edge of the input voltage. Obviously, u A The edges of the gate signal and the voltage error are synchronized, and the dead time causes the voltage error to be eliminated, only a small voltage error caused by the non-ideal rising edge and falling edge of the drain-source voltage exists.

[0167] Figure 16 The experimental results of the self-current-sharing of the auxiliary inductor current are shown. The input voltage is 2U dc =100V, the switching frequency is 250kHz, and the load current frequency is 500Hz. The S1 off time is delayed by 10ns at the initial moment to simulate the fluctuation of the duty cycle d A (t). It can be seen that the 10ns off delay causes the peak value of i d1 to rise to 3.8A, and the peak value of i d2 to drop to 2.9A, thus causing a current deviation of about 15%. When the 10ns S1 off delay is cancelled, the current self-sharing is achieved in one modulation wave period, without any delay and oscillation.

[0168] Figure 17 、 Figure 18 and Figure 19 are the average error voltage Fourier series (taking the first 100 harmonics) of the conventional interleaved parallel half-bridge power amplifier and the power amplifier proposed in the present application respectively under different dead times, switching frequencies and load currents. Specifically, Figure 17 It is shown that the average error voltage of the power amplifier proposed in the present application is irrelevant to the dead time, and the harmonic distortion caused by the dead time is fundamentally avoided. The average error voltage of the proposed power amplifier is only about 3% of that of the conventional scheme.

[0169] Figure 18 It is shown that although the average error voltage of the power amplifier proposed in the present application increases with the increase of the switching frequency, its amplitude is still about 30 times lower than that of the conventional scheme. Therefore, the power amplifier of the present embodiment can withstand a higher switching frequency while maintaining a relatively low THD.

[0170] It can be known from Figure 19 that the average error voltage of the power amplifier proposed in the present application increases by about 12 times in the range of 3.3-10A load current, but is still much lower than that of the conventional scheme, showing great advantages.

[0171] Figure 20 The measured load current THD data under different dead times T d are shown, where I o,peak is the peak value of the load current. The THD increases with the increase of the load, but is overall at a low level of 0.3%-0.8%. It can be known from the experimental results that the load current THD of the power amplifier proposed in the present application is not affected by T d , and there is no need to pursue an excessively small T d, thereby reducing the complexity of hardware circuit design.

[0172] Figure 21 The measured data of load current THD at different output frequencies are shown. When the output frequency is 250Hz, the output current THD is about 0.2% to 0.5% higher than that at 25Hz. This is due to the decrease in open-loop gain when the output frequency is high. In addition, f o =250Hz, I o,peak =9A, the THD inflection point appears. This is because the modulation ratio decreases at light load, which causes THD to increase.

[0173] Figure 22 The THD measurement results are shown for 100V input voltage and switching frequencies of 250kHz and 62.5kHz. Since the modulation ratio remains unchanged, the output current peak value I o,peak To ensure that the auxiliary inductor current is the same at 250kHz and 62.5kHz switching frequencies, L d Take 15μH; when the switching frequency is 62.5kHz, L d Take 60μH. Experimental results show that switching frequency is positively correlated with THD. Figure 16 Another conclusion can be drawn that the THD at 100 V input voltage is much lower than that at 300 V input. The final THD is as low as 0.063%.

[0174] In practical applications, there is no theoretical limit to the number of bridge arms that can be staggered and operated in parallel, but the number of bridge arms must be kept to an even number to eliminate harmonics at multiples of the switching frequency. Figure 23 Theoretically, there is no limit to the number of modules connected in parallel. In practical applications, the number of modules connected in parallel can be determined based on the load size.

[0175] Although the present invention is described herein with reference to specific embodiments, it should be understood that these embodiments are merely illustrative of the principles and applications of the invention. It should be understood that many modifications may be made to the illustrative embodiments, and that other arrangements may be devised, without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that the various dependent claims and features described herein may be combined in ways other than those described in the original claims. It should also be understood that features described in conjunction with individual embodiments may be used in conjunction with other described embodiments.

Claims

1. A parameter design method of a high-precision power amplifier based on resonant auxiliary commutation technology, the high-precision power amplifier based on resonant auxiliary commutation technology comprising a capacitor C1, a capacitor C2, four MOSFET switch tubes S1, S2, S3, S4, an auxiliary inductor L d1 , an auxiliary inductor L d2 , a filter inductor L o , and a filter capacitor C o , A capacitor C1 and a capacitor C2 are connected in series between the positive pole of the power supply and the negative pole of the power supply; The positive pole of the power supply is connected to the drain of a switch tube S1, the source of the switch tube S1 is connected to the drain of a switch tube S2, the source of the switch tube S2 is connected to the negative pole of the power supply; the positive pole of the power supply is also connected to the drain of a switch tube S3, the source of the switch tube S3 is connected to the drain of a switch tube S4, the source of the switch tube S4 is connected to the negative pole of the power supply; The connection point between the source of switch tube S1 and the drain of switch tube S2 is set as point A, and the connection point between the source of switch tube S3 and the drain of switch tube S4 is set as point B. An auxiliary inductor L is connected in series between points A and B. d1 and auxiliary inductor L d2 ; Set the auxiliary inductor L d1 and auxiliary inductor L d2 The connection point between them is point C, and the filter inductor L is connected in series between point C and the connection point of capacitor C1 and capacitor C2. o and filter capacitor C o ; Load R and filter capacitor C o in parallel; SPWM modulation is used for the two half-bridge arms, so that the two half-bridge arms are operated in staggered parallel to form a differential mode current, and the auxiliary inductor L d1 , the auxiliary inductor L d2 resonates with the junction capacitances of the four MOSFET switch tubes S1 to S4, so as to realize zero voltage switching in continuous current mode; characterized in that The parameter design of the high-precision power amplifier based on the resonant auxiliary commutation technology includes: For a filter consisting of a filter inductance L o and a filter capacitance C o the filter inductance L o and the filter capacitance C o must satisfy: 2π · f o,max · L o,eff · I o ≤ k v · U o , 2π · f o,max · C o · U o ≤ k i · I o , where L o,eff is the equivalent value of the filter inductance L o k f is the cut-off frequency coefficient, f o,max is the maximum value of the actual output frequency of the power amplifier; I o is the effective value of the output current, k v is the maximum inductance voltage drop coefficient, U o is the effective value of the output voltage, k i is the maximum capacitance shunt coefficient.

2. The parameter design method of the high-precision power amplifier based on the resonant auxiliary commutation technology according to claim 1, characterized in that, auxiliary inductor L d1 and the design value of the auxiliary inductor L d2 must satisfy: The peak current of the auxiliary inductor L d1 and the auxiliary inductor L d2 is not greater than the soft switching current threshold of the switch tubes S1-S4; the switch tubes S1-S4 remain soft switching at maximum load.

3. The parameter design method of the high-precision power amplifier based on the resonant auxiliary commutation technology according to claim 2, characterized in that, The inductance value of the auxiliary inductor L d1 and the auxiliary inductor L d2 are both represented as L d , then the peak current ΔI d of L dm must satisfy the following condition: where the peak current ΔI dm is the peak-to-peak value of the differential mode current, t is time, U dc is the DC bus voltage between the positive supply and the negative supply, m is the modulation ratio, T s is the switching period.

4. The parameter design method of the high-precision power amplifier based on the resonant auxiliary commutation technology according to claim 3, characterized in that, The switching tubes S1-S4 remain soft-switched at maximum load, L d The following conditions must be met i s_dx for the current flowing through the auxiliary inductor L d1 or the auxiliary inductor L d2 , C oss is the equivalent junction capacitance of the MOSFET switch, I dx,ZVS is the soft switching current threshold; Q oss is the charge stored in C oss when the switch is subjected to a voltage U ds , U ds is the positive bus voltage.

5. The method of parametric design of high precision power amplifier based on resonant auxiliary commutation technique as claimed in claim 1, wherein, The auxiliary inductor L d1 The inductance value of the auxiliary inductor L d2 is equal to the inductance value of the auxiliary inductor L 6. The parameter design method of the high-precision power amplifier based on the resonant auxiliary commutation technology according to claim 5, characterized in that, The auxiliary inductor L d1 and the auxiliary inductor L d2 The two half-bridge bridge arms formed by the switching tubes S1 to S4 are coupled, so that the two half-bridge bridge arms are resonant auxiliary commutation branches.

Images