Solar Thermal Cogeneration[Overview] [Program] [System] [Collector] [Turbine] [Generator] [Controller] [Battery]
Charge Controller
Introduction
The battery charge controller converts the multi-phase sinusoidal power output from the STC's electric generator into regulated DC power for charging the batteries. The generator's output levels vary during the day and the year. The optimum charging rate at any interval during the charging process depends on the battery state-of-charge and the end-use loads during the interval. The load on the turbine/generator must be regulated to achieve maximum power when needed, and maximum economy otherwise. The charge controller design should minimize embedded costs, maximize energy transfer, and minimize losses. The basic design should also work in reverse, driving the generator from the batteries to aid startup of the low-torque turbine. It should also be reusable in other system components including variable-frequency motor drives, DC steppers and AC inverters for end-use appliances, and even in fluorescent light ballasts and audio speaker amplifiers, maximizing energy efficiency.
Battery Charging Cycle
The charge controller is wired to the battery and the battery is wired to the end-user load. The controller measures controller voltage, battery voltage and load voltage and calculates the currents using cable resistance constants that have been measured and stored in the controller. When the controller is started each morning it begins monitoring the battery voltage and the end-user load current. If the battery is discharged (its voltage is lower than that at max-rated charge) the controller adjusts the charge voltage to supply up to the battery's max-rated charge current, plus the steady end-user load current. When the battery's max-rated charge voltage is reached, the controller holds it until the battery reaches its max-rated charge (the charge current decreases to a value known to indicate max-rated charge at the max-rated charge voltage). Then the controller continues supplying the steady end-user load current and maintains the battery max-rated charge, until the end of the day's charging interval.
The battery's max-rated charge voltage, max-rated charge current, voltage at max/min-rated charge, and current at max-rated voltage and charge are constants measured or chosen and stored in the controller. If the battery room temperature fluctuates more than about 10°F across the year then temperature compensation for the stored constants are necessary. A half hour rest period is needed to accurately measure the battery voltage at max/min-rated charge.
Commercial battery max-rated charge current is typically one eighth of the battery's amp-hour rating. Too high a charge current creates excessive heat which wastes power and weakens the battery plates, and excessive gassing which causes plate shedding, all of which reduce battery life. Too high a charge voltage creates the same problems. A battery's max-rated charge voltage is typically 2.3 to 2.4 V per cell, depending on materials and temperature. Too low a charge voltage results in less than full charge which, over time, allows the plate sulfate to harden, reducing battery life. The battery's max-rated charge is reached when the charge current levels off to a quiescent value while the charge voltage is held at max-rated value. An occasional overcharge may be beneficial to the battery by dissolving hardened sulfate.
See also: Lead-Acid Battery: Charging Strategy
Switch-Mode Voltage Conversion
The charge controller utilizes switch-mode voltage conversion which can convert variable and wide-ranging input voltages into variable and wide-ranging output voltages, with up to 95% efficiency. A switch-mode converter operates a high-speed, high-power solid-state switch at a high frequency and uses a relatively small inductor and capacitor to store energy during the voltage conversion. The switch is operated by a high-frequency rectangular wave signal with its pulse width determining the input/output voltage ratio. The inductor and capacitor values and switch frequency determine the output ripple. The switch pulse width and frequency may be changed to compensate for changes in the source or load parameters.
Like a transformer, the switch-mode converter's input/output voltage ratio is the inverse of its input/output current ratio, preserving power minus losses. Both can be designed to transfer a specific power to the load by supplying a specific voltage. But unlike a transformer, a converter is capable of dynamically changing the voltage ratio so the converter can thereby dynamically change the power transfer. Like a linear voltage regulator, a converter can supply DC output and can maintain a specific output voltage during changes in source and load parameters. Linear regulators are very inefficient because they do not preserve power but dissipate the input/output power difference as waste heat. Given the switch-mode converter's ability to preserve power at low cost, it is widely used to supply DC voltages in computer equipment, replacing both the transformer and the linear regulator. The switch-mode method is flexible enough to efficiently convert most any shape input power waveform to most any shape output power waveform. Switch-mode converters can dynamically change the load as seen by the source, a useful capability for stabilizing a solar-thermal system during intermittent cloud cover. There are two basic configurations:
Boost Configuration:
A
lower voltage is converted to a higher
voltage. An
inductor is connected in series between the voltage source and the
resistive load, and a switch is connected in parallel with the
load. When the switch is closed, current flows through the
inductor and
charges it. When the switch is opened, the inductor
maintains continuity of current flow through the load and the load
voltage becomes the current times the load resistance. A
capacitor
in parallel with the load maintains this voltage after the
switch closes to start a new cycle. A diode between the switch
and the capacitor prevents the capacitor's discharge. The
input-to-output voltage ratio equals the
ratio of the switching period to the switch off-time.
Buck Configuration: A
higher voltage is converted to
a lower voltage. A
switch and
inductor are connected in series
between the voltage source and the resistive load. When the
switch is
closed, a current charges the inductor, limited by the load
resistance. When
the switch is opened, the inductor maintains
continuity of current flow through a diode
connected
to ground, and the load voltage becomes the load
resistance times the current. A
capacitor in parallel with the load smooths out the
load
voltage. The input-to-output voltage
ratio equals the
ratio of switch on-time to the switching period.A third configuration, known as buck-boost, shares the same components but arrange them in yet another way and allows either voltage step-up or step down. There are limitations for each configuration which should be accounted for in finalizing a design.
Integration Into The Controller
The STC's generator supplies multi-phase power sinusoids and the charge controller actively rectifies them into DC and then converts/regulates the DC to voltage/current levels needed by the battery. The sun's energy varies widely over the course of the day, all of which is needed in the winter, resulting in a wide voltage range at the generator output. Use of a single converter, either buck or boost, requires that this voltage range be held completely above or completely below the battery voltage, which hovers in a relatively small range.
The generator coils may be connected in series to generate a range of high voltage while the battery cells are connected in parallel to accept a low voltage, and a buck converter employed. Or the generator coils may be connected in parallel to generate a range of low voltage while the battery cells are connected in series to accept a high voltage, and a boost converter employed. Efficient DC power distribution favors a high battery voltage. But the turbine/generator are most suited to high speed and voltage, so a careful analysis should determine this configuration.
The turbine/generator delivers maximum power along a certain current profile as a function of voltage. The converter may locate the input current that gives maximum power transfer by periodically adjusting the voltage conversion ratio while measuring the power. This periodic adjustment is required as the generator voltage and resistance, and battery voltage and resistance, change. This adjustment is subject to the battery's charge voltage/current requirements and the generator's and battery's rated limits. If the turbine/generator is producing more power than the battery can safely sink, then the system can intervene to throttle the turbine.
Rectifier and Power Factor Correction
The generator's multi-phase AC output is fed into a bridge rectifier, traditionally composed of a diode pair per phase with similar outputs connected together to form a pair of terminals. The result is a DC voltage plus a ripple that is traditionally filtered with a large capacitor. The capacitor charges only when the ripple peaks above the capacitor voltage so the current from the generator is pulsed at the ripple frequency. This high amplitude pulsed-current waveform has many high-frequency harmonics that reflect back to the generator and dissipate power by a.) increased skin effect resistance in wires, b.) increased eddy currents in coil wires and cores and c.) increased leakage across insulation. To eliminate these current harmonics the current waveform must be conformed to match the shape and phase of the voltage waveform. This creates a purely resistive load (zero reactance) in which all generator power is sunk, and is referred to as power factor correction (PFC) with a power factor of unity.
A conventional means of achieving PFC for single-phase AC is to employ a boost converter between the single-phase bridge rectifier diodes and the rectifier capacitor. The converter monitors the input voltage waveform and adjusts the conversion ratio to match the input current waveform shape/phase to the input voltage waveform shape/phase. The converter raises its output voltage to a point greater than the capacitor's voltage such that for each switching period the capacitor plus load draw a level of power at the converter output that presents that period's target current to the converter input.
The converter also monitors the output voltage and/or current and adjusts the input current waveform's envelope (profile of waveform peaks) to deliver the average power demand of the load, which in the case of battery charging corresponds to specific voltage and current profiles maintained within rated bounds across the charging period.
The switching frequency is set to be much higher than the input voltage frequency to minimize quantization error in the input current waveform. The switching signal is derived from the product of the various monitor signals. If a given application requires a lower output voltage than the input voltage peak, a second switch-mode stage in a buck configuration may be needed to lower the output voltage or it may be possible to employ a buck-boost configuration in the PFC stage.
In the single-phase
case, it might be more economical to
integrate the rectifier and converter, i.e. the rectifier diode
bridge and capacitor replacing the converter's diode/capacitor.
The converter's inductor
and switch are placed in front of the bridge and the rectifier
bridge/capacitor serve as the
converter's diode/capacitor. In the
multi-phase case, integrating the rectifier and converter is required
because the
phase currents become discontinuous, or pulsed, as
the phase voltages reverse-bias each-other's diodes. In the figure, it might be possible to remove the switches (S) and replace the rectifier diodes (D) with switches. This reduces the parts count and also enables the design to be used in motor, generator or motor/generator configurations. It's also possible that the inductors (L) may be eliminated and the generator coils used in their place by the PFC circuit.
Switch Signal Generation
Electronic circuits are needed to generate the power switch signals. These are fixed-frequency square-wave signals with a variable duty cycle. It may be necessary to control both the pulse start and the pulse width of each signal to optimize the power switch performance. With control of both parameters by the control/monitor subsystem, timing optimization may be performed in software, reducing components and increasing reliability. Discrete transistors can minimize power draw.
A master sawtooth oscillator drives a set of replicated power switch drive circuits (see figure). This oscillator is an astable multivibrator. The time constant of the RC network on its output side is much longer than the other side's time constant so its ramp will occupy the majority of the signal period. Each drive circuit is a series of stages: The first stage, a differential amplifier, or comparator, changes the input sawtooth into a rectangle with a DC-controlled duty cycle. This is a classic pulse width modulator but in this context its function is to produce a unique DC-controlled start offset for its power switch drive signal relative to the other drive signals. The start offset is manifest in the rising edge of the stage's rectangle output. The rectangle is then converted to a pulse, then to a sawtooth which drives a second pulse-width modulator which changes the sawtooth into a rectangle of width controlled by a second DC voltage. This start-controlled and width-controlled rectangle drives the power switch.
The oscillator should run at a fixed frequency that minimize switching losses and maximizes the power factor. A lower frequency minimizes switching losses by reducing the number of switching transitions. A higher frequency enables better power factor correction by reducing the quantization error. Switching losses might be further reduced with snubbers across the switch sources/drains.
Small Signal Transistors
Monitor / Control Circuits
The converter input and output voltages/currents must be measured and the power switches controlled by the control/monitor subsystem through electronic interfaces. The control/monitor subsystem has an analog-digital converter (ADC) and digital-analog converter (DAC) with several input and output signal channels. Voltages to be measured are stepped down appropriately with resistor voltage dividers. Currents are measured using a differential amplifier wired to the two ends of a power cable. Resulting voltages are sampled by the ADC. The DAC provides DC voltage levels that control the relative phases and pulse widths of the switch signals. The charge controller function is integrated into the control/monitor subsystem, and the ADC/DAC perform system-wide signal conversion. A single multi-tasking control/monitor subsystem reduces costs and provides opportunities to increase overall system performance and reduce development costs.
Components
Power component characteristics are critical to the design strategy, the circuit layout, and the reliability, efficiency and economy of the charge controller.
Capacitors have a working voltage rating which should be significantly higher than the peak voltage applied. The same is true of the ripple current rating. These must be derated at higher ambient temperatures. Strict observance of temperature, voltage and current ratings is required. Manufacturers should be asked directly about the lifespan of their capacitors. Capacitors have small series and parallel resistances and inductances that may affect the efficiency of the system, especially at higher frequency and higher current operation. More on Capacitors
Inductors have DC winding loss and AC winding and core losses which should be minimized. Core saturation and stray flux should also be considered. DC winding loss is a function of current and winding resistance. AC winding loss is like DC winding loss but increases with frequency due to the skin effect. AC core losses (eddy currents and hysteresis) also increase with frequency and also depend on core characteristics. Core saturation depends on current and core characteristics. Stray flux depends on the inductor geometry. More on Inductors
MOSFET switches should have as low on-resistance as possible. The voltage drop is current times on-resistance and the power loss is voltage drop times current. Since on-resistance increases with temperature, MOSFET efficiency improves with heatsinking and cooler ambient temperatures. The gate levels that set the on and off states are important for minimizing on-resistance and maximizing off-resistance. A shorter switching transition improves efficiency by reducing lossy conduction in the linear range. A higher switching rate decreases efficiency by increasing the percentage of time spent in transitions. Switching transients (spikes, ringing) may add to these power losses and stress the MOSFETs.
IGBT switches are an alternative to MOSFET switches. IGBTs have lower on-resistance than MOSFETs and may be more economical. The IGBT's on-resistance changes little with current and temperature. IGBTs have a small tail current when switched off which may affect efficiency at higher speeds. Generally, MOSFETs are better for higher speeds (>100kHz) and IGBTs are better for higher power (>1kW) .
Power diodes should generally have low forward voltage drop, low reverse leakage current, high reverse breakdown voltage, fast switching time, and low switching transients. Diode switching transients may exceed the breakdown voltage of connected devices or damage the diode itself and create other losses and stresses, especially when amplified by inductor resonances. Forward voltage drop results in a conduction loss. Reverse leakage current usually creates inefficiencies in the circuit. For converter output voltages greater than 12 volts, ultra-fast recovery diodes are usually specified, and schottky otherwise. Switched MOSFETs might perform the diode function more efficiently by providing a lower forward voltage drop.
Diodes Basics
More on Power Semiconductors
Why Opt for IGBTs in SMPS Applications?
Selection of MOSFETs in Switch Mode DC-DC Converters (pdf)
More on Capacitors
There are three basic types of capacitors in mass production. Film capacitors are usually rolled cylinders of plastic dielectric film and aluminum foil. Or the film is metalized, eliminating the foil. Often the film is polypropylene. Film/foil capacitors are particularly reliable and stable in capacitance and electrical properties over time, temperature and voltage/current ranges but are bulky due to low dielectric constants and are sensitive to soldering heat. Ceramic capacitors have much higher dielectric constants and are popular for their compactness but have poor electrical properties with higher capacitance and lose capacitance with time. Electrolytic capacitors can have very high dielectric constants and serve high capacitance applications with low cost and compact size. The aluminum types are much lower cost than the tantalum types. They are typically used in relatively low-voltage DC applications and require protection from voltage reversal. Electrolytic capacitors sometimes have marginal electrical properties or poor reliability/longevity. Some manufacturing problems have been documented.
The polypropylene film + aluminum foil rolled cylinder appears to be the most robust and reliable capacitor for high voltage, high current, high temperature applications where physical size isn't an issue. Heat primarily determines a capacitor's lifespan. Heat is generated by electrical resistance in the foil, and resistance to voltage change in the film. Voltage-handling capability is related to film thickness. Power-handling capability is related to voltage squared and foil area, and inversely with film thickness.
Film/foil capacitors are relatively easy to construct. At least two layers of polypropylene film dielectric should be layered to prevent problems with pinholes. In a rolled cylinder, several terminal connections to the foil are necessary to minimize resistance and inductance. Narrow aluminum strips (2 or 3 x foil thickness) should be soldered to the foil with lead-free, flux-free solder after removing the oxide using a suitable abrasive. The strips should be soldered to the foil side that makes the solder berm curvature consistent with the cylinder's after rolling up.
Capacitance (farad = coulomb/volt, ampere = coulomb/second) equals the dielectric constant (2.0 for polypropylene) * 8.854 x 10-12 F/m * area (m) / thickness (m) of the dielectric. For a given capacitance, a capacitor constructed with a larger container surface area dissipates more heat, thus a parallel set of small capacitors dissipate more heat than one large capacitor. But a larger number of smaller capacitors has a greater risk of fabrication defects. (see Thermal / Layout)
To prevent moisture/dust from being rolled up in the capacitor, it should be constructed in a dry, dustless, moisture-controlled environment. Dust and particulate control is achieved with an air purifier. Moisture control is achieved by warming the table and materials to keep them warmer than the air. The materials should be kept as clean as possible and should not be touched. A foil, film, foil, film stack is formed for rolling. The film should extend past the foil edge on all four sides of the foil to prevent edge arcs. Extra strips of film might be placed over the terminal strips to alleviate the extra mechanical stress on nearby materials after the capacitor is rolled up. The terminal strips for the two foil layers should extend at opposing ends of the cylinder, making what is called an axial-lead capacitor. This makes the terminals easier to manage. Rolling tension should be consistent. After the capacitor is rolled up the ends should get some epoxy to mechanically fix the terminals to prevent their bending at the foil/film. Then the terminals are soldered together using the absolute minimum heat to form the capacitor leads. Then the capacitor is fully sealed with epoxy to prevent moisture from entering.
For energy storage, a rough calculation based on a commercial polypropylene film capacitor of capacitance 2.2uF, operating voltage 75V, and volume 7e-6 m^3, revealed its volumetric energy density to be about 75000 wh/m^3, roughly three times that of lead-acid batteries. Foil capacitors (as opposed to film) may be roughly equal to the lead-acid battery's. The capacitor has some very significant advantages in material toxicity, weight, maintenance and charge/discharge rates. Its disadvantages probably include fabrication costs for the plate/dielectric sheets, and leakage rate.
More still on Capacitors
Capacitor Construction
Capacitor ESR Ratings
Why Capacitors Fail
CapSite 2007
More on Inductors
The inductor's purpose in a switch-mode converter is to transfer energy during the switching period by storage and release. In the STC, this involves levels of current and voltage that traverse a large range of values during the generator rotational period so both the peaks and the averages should be considered. Smaller value components may be ganged to share the voltage and current loads.
The energy stored in an inductor's magnetic field is proportional to the component's inductance times the current squared. Inductance equals roughly the number of coil turns squared times the loop area times the core permeability divided by the coil length. In switch-mode converters, a higher switching frequency lowers the voltage across the inductor enabling the use of a smaller component for a given energy transfer, reducing losses, space and cost. Higher core permeability further reduces these.
The core is the volume bounded by the loop area and coil length. The simplest inductor has an air core but a ferrous (soft magnetic, easily magnetized) core has greatly increased permeability and inductance. The ferrous core is magnetized by the coil and its magnetization adds to the coil's field strength in proportion to the core's permeability (see magnetic domains). The added inductance peaks and tapers as the ferrous core nears full magnetization (saturation) beyond a certain coil current.
Laminated iron cores are suitable for low-frequency applications, e.g. line frequencies. For higher frequencies powdered iron cores are more effective. Dielectric between the iron granules provides gaps that block eddy current loops.
Iron permeability/inductance is up to ninety times that of air. But switch-mode power supplies create rectangular pulses with much high-frequency energy and the eddy current loss in iron is substantial, even for the finest granularity powder. A hundred times higher permeability/inductance and a hundred times lower resistive loss to eddy currents are achieved with ferrite cores. One ferrite that might be easily obtained is a nickel-zinc alloy.
Many common core shapes leave gaps in the magnetic circuit for easier fabrication and winding. But if the core makes a closed magnetic circuit, e.g. a toroid, then maximum permeability/inductance and also minimum emissions are achieved. Ferrite toroid performance is so high that a switch-mode design may require only a small number of coil turns allowing wide spacing between them. This spacing reduces wire losses, reduces wire capacitance, and increases reliability.
But core gaps are needed in nearly all applications to linearize the magnetization such that the energy stored is proportional to the field applied, below the saturation current. The closed toroid relies on the distributed gaps of the powder and ferrous-type core materials for linearity. The optimum distributed gap density matches the core's saturation current with the application's maximum operating current.
A lower permeability core material is effective when the switching frequency is lower and the current requirement is higher. Energy storage increases with a larger volume core which provides more flux capacity for the same permeability and enables more loop area and turns.
The core cross-section should be circular instead of square to minimize wire stress. To minimize skin effect loss, the wire size may be reduced, and multiple strands used in parallel to achieve the necessary current handling. To minimize proximity effect loss, the wire should be separated by approximately one wire diameter. This may be accomplished by coating bare wire with a suitable insulation. More on magnet wire.
Ferrite Cores
Practical Construction Tips For Coils Using Iron Powder Cores (pdf)
Magnetic Cores for Switching Power Supplies (pdf)
Thermal / Layout
In high frequency, high power application, residual capacitance, inductance and resistance become factors to consider and component layout becomes important. High-current wires should be short and fat. Current loop areas should be minimized. Ground wires should not be shared but connected separately to a single point. Replicated components intended to share loads should have symmetric layout. Mechanical stress on component leads and thermal stress from soldering should be minimized.
High temperatures degrade component efficiency and function and reduce longevity through thermal stress. The first step is to minimize resistive losses in the design. Next is proper heatsinking of semiconductors, and load sharing, with capacitors in parallel and inductors in series. And third, the controller should be mounted in a moderate temperature environment, e.g., in the battery room with proper ventilation.
Charging Ripple
Ripple in the charging voltage/current can create a periodic over-voltage/current, causing excess heating and gassing, and/or a periodic under-voltage and discharging. These effects are likely to be diminished at higher ripple frequencies due to electrochemical response inertia. A certain amount of ripple may increase battery performance by mixing electrolyte components thereby improving the electrochemical process rate.
Battery Capacity Monitoring
The control/monitor subsystem can be programmed to monitor the battery capacity and performance. Recording the battery's voltages and currents allows monitoring its internal DC resistance over time to develop a profile that can help diagnose battery problems and predict battery lifespan. Electrochemical Impedance Spectroscopy (EIS) is a method of analyzing the battery by injecting small amplitude sinusoidal charging voltages over a range of frequencies and recording the charging current to obtain internal battery impedance versus frequency. This may provide additional information over DC resistance.
Cabling / Breakers
Circuit breakers are inserted between the batteries and the end-user loads to prevent short circuits and resulting fire hazard and battery damage. But the breakers should also be tripped when the battery discharges below its rated minimum-charge limit. A single set of breakers may be tripped by the controller on either low charge or short circuit, because it is capable of detecting either condition. If the condition was low charge, the breakers would be reset by the controller after the battery is re-charged. If the condition was a short circuit, the breakers would be manually reset through the control/monitor subsystem's user-interface. These breakers should be mechanically sprung such that no quiescent power is needed to maintain them either on or off.
DC Residential Circuits
To economically transmit power over very long distances it is necessary to minimize both the wire thickness and resistance losses which requires using very high voltages that have to be stepped down for end-use. In the past it wasn't feasible to step DC voltage up and down, while transformers were used to step AC voltage. Besides, even moderate voltage DC arcs across mechanical switches, inflicting severe contact wear while AC does not create this problem. However, transformers are inefficient and expensive and now that solid-state electronic steppers are cost-effective, high voltage DC can be more effectively distributed with step-down power electronics at points of use. Step-down also provides a variety of DC voltages needed by various appliances.
Variable Frequency Drive / Inverter
A solid-state variable frequency drive (VFD) provides a sinusoidal AC power signal to a synchronous AC motor, enabling dynamic control over the motor's speed/torque, and allowing smooth starts/stops that can greatly extend the life of the motor. In variable load applications, such as centrifugal pumps and fans, a VFD may be connected in a control loop to maintain the max-efficiency point on the load/speed curve and sometimes greatly reduce the motor's overall power consumption compared with fixed frequency drive. The VFDs ability to respond quickly in control loops is utilized in driving the STC's boiler feedpump. VFDs also eliminate motor soft starters and pump throttle valves.
The simplest type of VFD is a linear amplifier driven by an oscillator, but linear amplifiers are inherently inefficient. A more efficient type of VFD is a switch-mode voltage converter configured to convert DC power input into AC power output with power factor correction. The VFD should maintain a constant ratio of voltage to frequency, as is produced when the motor is driven as a generator. The VFD must measure the back-EMF of the motor to maintain synchronization of the motor with the drive signal.
The full costs of the VFD, which can be substantial, have to be weighed against the benefits. Typically the benefits are greater for heavier duty motors/applications. Extended periods of low operating speed may allow heat buildup in the motor but this is less of a problem when driving centrifugal devices because of their variable torque character.
An AC inverter converts DC supplied by batteries to conventional AC power for end-use AC appliances. It shares the basic design of the VFD, except that its frequency and amplitude are fixed at standard values allowing for optimizations. A related device is electronic ballast for fluorescent lighting. A high quality AC inverter design with good power factor correction may be adapted to function more efficiently than commercial electronic ballasts. Power factor correction in the inverter design protects sensitive digital end-use appliances.
More on VFDs
Variable Frequency Drives Introduction
More on/References:
Switch Mode Power Supplies
SWITCHMODE™ Reference manual (pdf)
Power Supply- Switching Mode (SMPS) Circuit Design, Tutorials, Software
Circuits For Power Factor Correction With Regards To Mains Filtering (pdf)
Analysis and design of direct power control (DPC) for a three phase synchronous rectifier (pdf)
Switch Signal Generation Circuit
Updated: FILEDATE
Copyright (c) 2005-2009 Robert Drury
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.2
or any later version published by the Free Software Foundation;
with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts.
See "GNU Free Documentation License".
Disclaimer: This information may contain inaccuracies and is provided
without warranty. Safety first when working with high temperatures,
pressures, potentials, speeds, energies, various tools and materials.