CZ308969B6 - Charger for bidirectional energy flow and controlling it - Google Patents

Abstract

Charger for bidirectional energy flow, which has on the way from the mains voltage source (ug) to the battery: the first bidirectional converter (A) consisting of switches (S1 to S4) operating on the mains frequency, the second bidirectional converter (B) consisting of switches ( S5 to S8) operating at a frequency in the order of hundreds of kHz, a transformer (TR) and a third bidirectional converter (C) consisting of switches (S9 to S12) operating at the same frequency as the second converter. A coil (L) is connected in one of the buses (+, -) connecting the first and second converters (B, C), behind which the buses (+, -) are connected via a separate switch (S13) and a capacitor (C13). The method of controlling the charger by the control unit for the mains battery charging phase consists of a signal from the control unit being switched by a switch (S13) with period T / 2 and the second and third converters (B, C) are switched with period T. For safety reasons, the control unit leaves the switch (S13) open for an additional time during the safety delays tdead and tdelay before the interval to, or after this interval t0.

Description

Charger for bidirectional energy flow and method of its control

Field of technology

The invention relates to a charger which allows both the charging of an electric vehicle battery from a low-voltage electrical network and the reverse flow of energy, in which energy is pumped from the battery into the low-voltage electrical network. This opposite flow of energy can also be used to power the island network (eg an energy autonomous house).

State of the art

Bidirectional chargers of a given type usually have three bidirectional converters fitted with four switches with anti-parallel diodes in series on the way from the mains voltage source to the battery. The first bidirectional converter, which operates at the mains frequency, supplies direct current to the second bidirectional converter. The second converter, which converts direct current to alternating current, operates at a frequency in the order of kHz. The subsequent transformer supplies a high-frequency alternating current to the third bidirectional converter, which then converts the third converter operating at the same frequency as the second converter into direct current, which charges the batteries.

This type includes, for example, the chargers described in Chinese patent applications CN 106972604 A and CN 106849294 A. The chargers mentioned here are connected to a three-phase network. The connection has a voltage intermediate circuit and on the primary side of the transformer it contains a step-up voltage converter, which increases the voltage in the mode of energy recovery into the network. Due to the connection of several switching elements in series through which current flows, higher energy losses can be expected here.

The charger described in CN 109808522 A is intended for connection to a single-phase network. Its third converter allows to charge both the vehicle's traction battery with a voltage of 100 V and the auxiliary battery with a voltage of 12 V. This charger also has a voltage intermediate circuit between the first and second converter, which also contains a storage choke. This is short-circuited by the transistor at the appropriate times in the charging mode, which creates an increasing voltage. This allows harmonic current to be drawn from the network. At the same time, there are no dangerous voltage peaks. The storage choke is connected in one bus in series with one switch, while the buses are connected between the choke and this switch via another switch. The current in this circuit passes through more switching elements in series, which causes greater energy losses, ie lower efficiency.

The object of the invention is to provide a two-way charger intended in particular for automobiles, the circuit diagram of which ensures the smallest weight and dimensions and thus a high power and volume density.

The essence of the invention

The charger for bidirectional power flow, which has in order on the way from the mains voltage source to the battery: the first bidirectional converter consisting of four switches operating on the mains frequency, the second bidirectional converter consisting of four switches operating on the frequency in the hundreds of kHz, transformer and a third bidirectional converter consisting of four switches operating at the same frequency as the second converter. The essence of the charger lies in the fact that in one of the buses connecting the first and second converter a coil is connected, behind which the buses are connected via a separate switch and capacitor and wherein the switches of all three converters are individually controlled from the control unit.

The essence of the innovative way of controlling the charger via the control unit for the battery charging phase (BAT) from the mains is that the signals from the control unit are switched by the switch (S13) with period 772 and the second and third converters (B, C) are switched with period T , while closing the switches

-1 CZ 308969 B6 (S5 to Ss) of the second converter (B) energy is accumulated in the coil (L) as a current source for charging the battery (BAT) after its rectification _{during the time interval t 0, the control unit leaving the switch (} S13) additionally open during safety delays / dead and ídeiay before the interval t ₀ resp. after this interval t ₀ .

Clarification of drawings

The invention will be further elucidated with the aid of the drawings, in which FIG. 1 is a diagram of a preferred circuit of a charger according to the invention, FIG. 2 are switching diagrams of switches S5 to S13 for charging mode and FIG. Saw waveforms in the first three rows of the diagrams of FIGS. 2 and 3 indicate the function of the hardware timers (counts up) that are inside the microprocessor, wherein the marking Τι, T _2, T 3, T ₄ and T ₈ are numbered in their actual manual microprocessor. E.g. Ti is the designation of timer 1 and Chi is the number of the comparison channel that belongs to this timer. Typically, the processor that is used has four comparison channels for one timer. Four channels can be controlled by one timer. When the value of the timer it counts is the same as the value of the comparison register, the logic level of the output changes. Either from the log. 1 to log. 0 or vice versa, depending on the setting.

Examples of embodiments of the invention

The bidirectional charger, the diagram of which is shown in Fig. 1, ensures the transfer of electric current from the Mg source of mains voltage to the BAT battery. From the Mg source, the following are arranged in sequence: the first bidirectional converter A consisting of switches Si to S ₄ , operating at the mains frequency, and the second bidirectional converter B consisting of switches S5 to S ₈ . operating at a frequency in the hundreds of kHz. The first and second inverters A, B are connected by +, - buses. A transformer TR is connected to the second converter B and a third bidirectional converter C consisting of switches S9 to S12 operating on the same frequency as the second converter B is connected to it. which the buses +, - are connected via a separate switch S13 and a capacitor C13. Switches S5 to S13 are individually controlled from a control unit, which is not shown in the diagram.

The first inverter A on the network side switches in both operating modes (charging and discharging) on the low frequency of the LV network. Switches S5 to Si ₂ function as a bidirectional DC / DC converter with an isolating transformer TR. In order to achieve the lowest possible weight and dimensions of the entire device, this converter operates at a high frequency (hundreds of kHz). Only in this way can the dimensions and weight of the transformer TR and the coil L be significantly smaller. In both charging and discharging modes, all switching elements are actively controlled to achieve minimum power losses.

In charging mode, the first inverter A rectifies the mains voltage. Capacitor Ci serves as a high-frequency filter preventing the transmission of the switching frequency to the supply network. Coil L accumulates energy as long as switches S5 to Sg are closed. In terms of switching switches S5 to S ₈ and S9 to Si _{2 of} converters B, C, coil L is the current source and forms the so-called current intermediate circuit. At the same time, it acts as a filter against the penetration of unwanted high-frequency interference into the network. The energy transfer from the coil L to the transformer TR takes place during the closing of one of the diagonals of the second converter B, ie switches S5, Sg (positive voltage half-wave) resp. Sg, S7 (negative voltage half-wave). The secondary voltage of the transformer, then rectified TRje switches S9 to SI _second

In the discharging mode, the third inverter C operates with switches S9 to Si ₂ in the inverter function, the energy transferred by the transformer TR is directed by switches Sg to Sg. In this mode, coil L together with capacitor Ci serves to filter the pulse voltage from the second inverter B. The first inverter A with switches Si to S ₄ serves as an inverter for the subsequent transfer of energy to the grid.

The charger in charging mode (transfer of energy from the mains to the battery) works as follows - see the switching diagram in Fig. 2:

- 2 CZ 308969 B6

1.1 Switching switches Si to S4.

Switches S1 to S4 act as a rectifier bridge. In principle, it could also be an uncontrolled diode bridge. These switching elements are controlled to achieve higher efficiency (so that there are smaller voltage drops at switches S1 to S4). Switches S1 to S4 always switch in pairs S1, S4 and Sg, S3. They switch at the mains frequency, the switching moments are guided by the mains, by the passage of the mains voltage to zero.

1.2 Switching switches S3 to Sg and S13.

The switching cycle period is denoted by T, which are of the order of ps, which corresponds to a switching frequency in the order of hundreds of kHz. The switching period of switch S13 is halved (double frequency). Switches S3, Sx switch on and off simultaneously, as does the pair of switches S3, S7. In the following description, it is assumed that the time t = 0 is not the moment of the start of the circuit, but only the beginning of a new period. The circuit was already in a dynamic steady state from the previous period, the coils and capacitors have stored energy in them.

1.2.1 Time = 0.

At time t = 0, switches S3, Sg open, while S3, S7 are closed from the previous period. Switch S13 is open.

Current z'l flows through coil L, while the current through the primary winding of the transformer TR ž _p is zero at this time. Ideally, the current of coil L should start flowing immediately to the primary winding of the transformer TR. However, this is not possible due to the stray inductance of the real transformer. The current from coil L continues through the anti-parallel diode of switch S13 to capacitor C13. C13 is charging, energy is accumulating in it. At the same time, the current from _p . If C13 were not there, a dangerous voltage peak would be created between node 1 and node 0, which would destroy the switching elements. This is exactly the task of the serial connection S13, C13 ·

1.2.2 Time h.

In time, switch S13 will close.

Your time is short, on the order of a percentage of the period. It is inserted there as a safety delay between opening S3, Sg and closing S13. Their joint switching would mean a short circuit with condensate C13 ·

The current from _{p is} still increasing, as started in point 1.2.1. After a while, it reaches the same value as z'l. At that moment, the current flowing into the condensate C13 ceases, the direction of the current in C13 reverses. The capacitor C13 begins to discharge into the transformer TR, the current ž _p is now by the current through the condenser C13 higher than the current z'l.

In this part of the period, energy is transferred from the primary to the secondary side of the TR transformer.

1.2.3 Time ri.

In time, switch S13 will switch off. Capacitor C13 is thus disconnected from the circuit. The transfer of energy from the primary to the secondary side of the TR transformer still takes place.

The opening of S13 at time ti occurs due to the fact that at time Ϊ3 the pairs of switches S3, Sg and S3, S7 are closed at the same time (the accumulation of energy into the choke L begins). Therefore, S13 must turn off before all four switches S3 to Sg are closed to prevent a short circuit charged by condensate C13. The safety delay between ti and h is approximately one percent of the time of period T.

-3GB 308969 B6

1.2.4 Time h.

At time 7, switches Sg, Sg _{are closed, switches Sg, S 7} remain closed.

The primary winding of the transformer TR is short-circuited, the current from _p decreases to zero.

The instantaneous value of the rectified mains voltage Uá is now on coil L, the current through coil L increases and energy accumulates in it.

This condition lasts until time 1/27. We refer to its duration as the overlap time t ₀ (overlap).

Increasing t ₀ , i.e. decreasing the value 7, increases the energy accumulated in the coil L, the current flowing through the coil L and, ultimately, the total energy transmitted to the secondary side of the transformer.

Conversely, by shortening t ₀ (increasing the value of 7), the total amount of energy decreases.

Simultaneously with the shift 7, the slide 7 must also be moved in order to maintain the safety delay described in section 1.2.3.

1.2.5 Time 1/27.

Time 1/27 is half of the period; the plot begins to repeat with the difference that the pairs Sg, Sg and Sg, S ₇ exchanged roles.

At time 1/27, the switches Sg, S ₇ open, while Sg, Sg are closed from the previous period. Switching element Sb is open.

Current z'l flows through coil L, while current z _{p through the} primary winding of transformer TR is zero at this time. Ideally, the current of coil L should start flowing immediately to the primary winding of the transformer TR. However, this is not possible due to the stray inductance of the real transformer. The current from coil L continues through the anti-parallel diode of switch Sb to capacitor Cb. Capacitor Cb charges, energy accumulates in it. At the same time, the current from _p .

If Cb was not there, a dangerous voltage peak would be created between node 1 and node 0, which would destroy the switching elements. To prevent this, it is the task of the serial connection Sb, Cb.

1.2.6 Time 1/27 + t.

The switch Sb is closed. The situation is analogous to that described in 1.2.2.

1.2.7 Time 1/27 + 7.

The switch Sb is switched off. The situation is analogous to that described in 1.2.3.

1.2.8 Time 1/27 + 7.

Se, Sg. The overlap time 7 occurs, as described in 1.2.4.

1.2.9 Time 7.

At time 7, which is one period of the switching cycle, the whole process is repeated exactly as it started at time 7 = 0.

1.3 Switching of switches Sg to Sb.

-4CZ 308969 B6

Switching elements S9 to S12 function as a rectifier which rectifies an alternating voltage with a frequency of hundreds of kHz at the output of the transformer TR.

In the simplest case, a diode bridge would suffice for this purpose. However, the diodes used must be fast enough. Because the switching elements used (MOSFETs) have a smaller voltage drop in the closed state than the diode in the forward direction, the rectifier is switched (controlled) to reduce power losses. At the same time, this jumper must be equipped with active switching elements due to the possibility of the reverse function of the charger (energy transfer from the battery to the mains).

1.3.1 Time t = 0.

At time t = 0, the Sn switch opened. Switches S9 to S12 are now all open, rectifier converter C only functions as an uncontrolled, diode. At time t = 0, the secondary current of the transformer z _{s is} zero, the primary current z _p is also zero. As the current from _p gradually begins to increase, see 1.2.1, so does the current from _s , which closes through the anti-parallel diodes of switches S10 and Sn.

1.3.2 Time h.

In time, switches S10 and Sn close and current flows through them, instead of their anti-parallel diodes, thus reducing power losses. At this stage, most of the energy is transferred to the BAT battery.

1.3.3 Time 7.

Switch S10 opens, while Sn remains closed.

Opening S10 occurs because in a moment, at time 7, the primary winding of the transformer TR will be short-circuited by simultaneously closing switches S5 to S ₈ (see 1.2.4.). If the primary winding were short-circuited at the same time and both switches S10 and Sn were closed, the energy flow would be reversed, energy would flow from the BAT battery to the transformer TR. which must not happen.

1.3.4 Time h.

At time h, the state of switches S9 to S12 does not change.

1.3.5 Time 1/27.

The UTT time is half of the period, the process begins to repeat, with the difference that the pairs of switches S9, Sn and S10, Sn have exchanged roles. The Sn switch opens. Rectifier converter C functions as an uncontrolled current from _the secondary winding of the transformer TR. it gradually grows from zero and closes via the anti-parallel diodes of switches S9 and Sn.

1.3.6 Time 1/27 + / 1.

At time / 1, switches S9, Sn are closed and current begins to flow through them, instead of their anti-parallel diodes. This reduces power losses. At this stage, most of the energy is transferred to the battery.

1.3.7 Time 1/27 + 7.

Switch S1 is opened while Sn remains closed. This is an analogous situation to 1.3.3.

1.3.8 Time 1/27 + 7.

At time 7, the closing of switches S9 to Sn does not change.

-5CZ 308969 B6

1.3.9 Time T.

At time T, which is one period of the switching cycle, the whole process is repeated exactly as it started at time T = 0.

The charger in the discharge mode (transfer of energy from the battery to the mains) works as described below, see the switching diagram in Fig. 3.

2.1 Switching switches Si to S4.

Switches S1 to S4 function as a jumper in inverter mode. They switch the polarity of the DC voltage ua, depending on the polarity in the network. The voltage itself is not a constant value, but by changing the widths of the individual times 0 to C, it is modulated so that its shape is a repeating approximately sinusoidal half-wave. Inverter A with switches S1 to S4 inverts these half-waves as required, thus supplying an AC voltage of approximately sinusoidal to the network.

2.2 Switching switches S9 to S12.

The switching cycle period is denoted by T, which are of the order of ps, which corresponds to a switching frequency of hundreds of kHz.

In the following description, it is assumed that the time t = 0 is not the moment of the start of the circuit, but only the beginning of a new period. Capacitors and coils are in dynamic steady state from the previous period.

Switches S9 to S12 function as an inverter bridge, they alternate the DC voltage of the BAT battery to AC, which is further transformed by the transformer TR.

The designation of the individual windings of the transformer is left in this part of the explanation as well as in the explanation of the charging mode. Thus, the winding on the battery side of the BAT remains secondary and the winding on the mains side primary, even if in this mode their function is reversed.

2.2.1 Time t = 0.

At time t = 0, the switches S1, S12 close, and the battery voltage BAT is applied to the secondary winding of the transformer TR. Current begins to flow through the secondary and subsequently also the primary winding of the transformer TR. This transfers energy from the BAT battery to the grid side.

2.2.2 Time h.

In time, the state of switches S9 to S12 does not change.

2.2.3 Time Í2.

In time, the state of switches S9 to S12 does not change.

2.2.4 Time 6.

At time h, switch Sg opens. The transfer of energy from the BAT battery to the network side ends. By changing the time h, the mean value of the current and thus the value of the transmitted energy are regulated. The time can vary from ti to ti. For more details on time 6 in point 2.3.3. and to time 6 in paragraph 2.2.6.

The following is a description of what happens in the circuit in this situation, although this is not of immediate relevance to the explanation of the circuit diagram:

-6CZ 308969 B6

The voltage between nodes 0 and 1 drops to zero. Switch S12 remains closed. The voltages at ₉ and at _s decrease to zero, but the currents in the primary and secondary windings of the transformer TR are not zero. Because energy is stored in the stray inductance of the TR transformer, both currents circulate "idle" in closed loops.

The current z _s flows through a loop formed by the secondary winding of the transformer TR, closed by the switch S12 and the anti-parallel diode of the switch Sn.

The current ip flows through the loop formed by the primary winding of the transformer TR. switch S1, coil L, inverter A with switches S1 to S4 in parallel with capacitor C and switch Ss. Since the voltage between nodes 0 and 1 is currently zero, it is necessary to create an equally large voltage of opposite polarity in the loop to overcome the voltage Ua. This voltage is created by coil L.

2.2.5 Time here.

At time 7, the state of switches S9 to S12 does not change.

2.2.6. Time for you.

In time, switch S12 opens for you. This moment is slightly ahead (by the order of% of the time of period T), before the middle of period 772. In this way, the so-called dead time 7ead before the closing of Sm.

2.2.7 TIME.

Time 1/27 is halfway through the period. From this moment on, the process repeats, but with the opposite polarity of the current flowing through the transformer, with the difference that the pair of switches S9, S12 exchange the corner with the pair S10, Sn.

The switches S10, Sn close, the battery voltage BAT is applied to the secondary winding of the transformer TR. but in the opposite polarity than in 2.2.1. Current flows in the opposite polarity than in 2.2.1 through the secondary and primary windings of the transformer TR. The energy from the BAT battery is transferred to the mains side.

2.2.8 Time 1/27 + h.

At time 1/27 + ti, the state of switches S9 to S12 does not change.

2.2.9 Time 7.

At time 1/27 + ti, the state of switches S9 to S12 does not change.

2.2.10 Time 1/27 + 7.

At time 1/27 + 7, switch S10 will open. The transfer of energy from the BAT battery to the network side ends. For this time, what was written for time 7 in point 2.2.4 applies by analogy.

2.2.11 Time 1/27 + 7.

At time 1/27 + 7, the state of switches S9 to S12 does not change.

2.2.12 Time 1/27 + 7.

At time 1/27 + 7, the Sn switch is switched off. This moment is slightly ahead (by the order of 1% of the period 7), before the period T. In this way, the so-called dead time 7ead occurs before the switching on S ^.

2.2.13 Time 7.

-7 CZ 308969 B6

At time T, which is one period of the switching cycle, the whole process is repeated exactly as it started at time T = 0.

2.3 Switching switches Ss to S ₈ and S13.

Switches Ss to S ₈ essentially form an active rectifier converter B. In principle, a diode bridge would suffice instead (switching elements contain anti-parallel diodes, which can perform the function of a rectifier). However, to increase efficiency, the switching elements are controlled; a closed MOS-FET transistor has a lower voltage drop than a diode.

Pair of switches Sg, S ₈ and Sg, S? they switch at the same time. Simply put, at the moment when the current would naturally flow through a certain anti-parallel diode, the corresponding switch is also closed.

Capacitor C13 with switch S13 is used to limit the voltage peaks that occur in the circuit between nodes 0 and 1 when switching the voltage polarity on the transformer TR at time t = 0 and t = H2T. The resulting voltage peaks and their energies are smaller than when the circuit is operating in the charger mode, however, they are there and would threaten the circuit elements by overvoltage, so it is necessary to limit them. When a voltage peak occurs, it is absorbed by capacitor C13, spontaneously via the anti-parallel diode of switch Si3. At a certain point in the operating cycle, the accumulated energy is discharged back into the circuit by closing switch S13.

2.3.1 Time t = 0.

At t = 0 the switch Sg to state S _8, or S13 unchanged.

2.3.2 Time h.

In time, a pair of switches Sg, Sg and at the same time S13 switches on you. The time ti is of the order of 1 to 2% of the time of the period T.

2.3.3 Time Í2.

In time / 2, switch S13 switches off.

Time 7 is delayed by time ti of the order of 1 to 2% of the time of period T.

2.3.4 Time 6.

At time h, the state of switches Sg to Sg and S13 does not change.

2.3.5 Time ri.

At time ri, the pair of switches Sg, Sg is switched off. Time 7 is delayed by about 1% of the time of period T after time C.

2.3.6 Time Í5.

Over time, the state of switches Sg to Sg and S13 does not change.

2.3.7 Time 1/27.

Time 1/27 is half of the period. From this moment on, the process repeats as it started at time t = 0, with the difference that the pair of switches Sg, Sg exchanges a role with the pair Sg, Sg.

At time 1/27, the state of switches Sg to Sg and S13 does not change.

-8CZ 308969 B6

2.3.8 Time 1/27 + 7.

At time 1/27 + 7, the pair of switches Se, 87a switches at the same time S13. Time 7 is of the order of 1 to 2% of period 7.

2.3.9 Time 1/27 + 7.

At time 1/27 + ti, switch S13 switches off. Time 1/27 + ti is delayed by time 1/27 + / 1 of the order of 1 to 2% of the period T.

2.3.10 Time 1/27 + 7.

At time 1/27 + 7, the state of switches S5 to Sg and S13 does not change.

2.3.11 Time 1 / 2T + 7.

At time 1/27 + 7, the pair of switches Se, Sj is switched off. Time 1/27 + 7 is delayed by about 1% of the period 7 after time 1 / 2T + 7.

2.3.12 Time 1 / 2T + 7.

At time 1/27 + 7, the state of the Sg and S13 switches does not change.

2.3.13 Time 7.

At time 7, which is one period of the switching cycle, the whole process is repeated exactly as it started at time 7 = 0.

Claims (4)

PATENT CLAIMS 1. Charger for bidirectional energy flow, which has in _{series on the way from the source (w g} ) mains voltage to the battery (BAT): the first bidirectional converter (A) consisting of switches (Si to S ₄ ) operating on the mains frequency, the second bidirectional a converter (B) consisting of switches (S5 to Ss) operating at a frequency in the hundreds of kHz, a transformer (TR) and a third bidirectional converter (C) consisting of switches (S9 to S12) operating at the same frequency as the second converter (B), characterized by in that a coil (L) is connected in one of the buses (+, -) connecting the first and second converters (A, B), behind which the buses (+, -) are connected via a separate switch (S13) and a capacitor (C13) and wherein the switches (S1 to S13) are individually controlled from the control unit. A method of controlling a bidirectional power flow charger according to claim 1 via a mains charging phase (BAT) control unit, characterized in that the signals from the control unit are switched by a switch (S13) with a period of 772 and a second and a third converter (B). , C) are switched with period T, while when the switches (S5 to Ss) of the second converter (B) are closed at the same time, _{energy is accumulated in the coil (L) during the time interval t 0} as a current source for charging the battery (BAT) after its rectification, wherein, for safety reasons, the control unit keeps the switch (S13) open for the duration of the safety delays / dead and / delays before the interval t ₀ resp. after this interval t ₀ . Charger control method according to claim 2, characterized in that the signals from the control unit are switched by switches (S5 to S13) on the time axis as follows: in time Ti V Case Ti T / delay In Time 71 + T / 2 -t ₀ - / dead at time Ti + T / 2 -t ₀ at time Ti + T / 2 In Time Ti + T / 2 + / delay In Case Ti + T - t ₀ - / dead at time E + T -1 ₀ at time E + T and the cycle is repeated. switches (S5, Ss) open, switches (Se, S7) are closed and other switches are open; switches (S10, Su, S13) close; switches (S10, S13) open; switches (S5, Ss) close; switches (Se, S7, Sn) open; switches (S9, S12, S13) close; switches (S9, S13) open; switches (Se, S7) close; switches (S5, Ss, S12) open, switches (Se, S7) are closed, other switches open; A method of controlling a charger for bidirectional power flow according to claim 1 by means of a control unit for the phase of energy transfer from the battery (BAT) to the mains, characterized in that the signal from the control unit is a switch (S13) switched with period T / 2 and the second. and the third converter (B, C) are switched with period T, the next time the switches (S5 to S13) are switched by the control unit in the time axis, the coil (L) with capacitor (C13) filters the pulse voltage generated by the converter (B, C) from the voltage batteries (BAT) to transfer energy from the battery (BAT) to the mains: at time E V Case Ti T / delay At time Ti + / delay + ÍS13 at time Ti + ^ on at time Ti + ton 4 “/ delay At time Ti + T / 2 - / dead at time Ti + T / 2 At time Ti + T / 2 + / delay at time Ti + T / 2 + / si ₃ at time Ti + T / 2 + / _on at time Ti + T / 2 + / on> / delay at time Ti + T - / _dea d switches (S9, S12) close and the others are open; switches (S5, Ss, S13) close; the switch (Si ₃ ) opens; switch (S9) opens; switches (S5, Ss) open; switch (S12) opens; switches (S10, Sn) close; switches (Se, S7, Si ₃ ) close; switch (S _i3 ) opens; switch (S10) opens; switches (Se, S7) open; switch (Sn) opens; -10GB 308969 B6 at time 7 + 7 switches (S9, S12) close and the others are open and the cycle is repeated, while Gead is the safety time delay before TU, / delay is a delay of 2.5% T, / s is closed time 5 (S13 ) 1.25 to 2.5% 7a / _on is a regulated battery current transfer time (BAT) of 2.5% to 48.75% T.