MB Quart Power Supply White Paper Power Converter The MB Quart Q-Series of amplifiers are provided with fully regulated voltage rails from which draw power. The rail voltages are preset for optimum performance of the amplifier into either 2 or 4 ohm speaker impedances. The load impedance must be selected via the READ board to match the power supply voltage rails to the amplifier output impedance as described below. The power supply is designed to be unconditionally stable under all specified operating conditions. The power supplies used on the MB Quart Q-Series amplifiers utilize current mode control. The controller chip is a modern fixed frequency Pulse Width Modulator (PWM) controller. It includes an oscillator, a temperature compensated reference, an internal wide bandwidth error amplifier, a high speed current sensing comparator, a steering flip-flop, and dual high current totem pole outputs. It also has protective features including input and reference undervoltage lockouts with hysteresis, cycle-by-cycle current limiting, and a latch for pulse metering. This controller can be configured for either current or voltage mode control. In this application it is used in a current mode control topology. Each power supply output rail is monitored and the voltage scaled to the reference voltage provided by the PWM. The scaled outputs of each rail are then compared (diode or-ed) and the lowest valued voltage is fed to the input of the system error amplifier. The system error amplifier provides negative feedback for forcing the controlled secondary to tightly track the precision reference voltage. This is the PWM error amplifier. The error amplifier in conjunction with the pulse current then adjusts the pulse width drive to the PWM outputs. These then drive the transformer primary switching Mosfets to compensate and keep the rail voltage within specified design limits.. There are either two (for the four channel amplifier Power supply) or four (for the mono and two channel amplifiers) TO-247AC Mosfets connected in each side of a push-pull connection to drive the power transformer primary. Current mode control is accomplished by providing a scaled representation of each transformer primary current pulse to the PWM. This scaled current is then compared to the output of the system error amplifier. As the current rises, the comparator input voltages cross causing the comparator to trip, terminating that pulse. This information is passed to the PWM latch and steering circuits which assure that each output is turned on in sequence and no double pulsing occurs. This pulse by pulse current programming provides feedforward for faster response to input voltage changes. The feedforward results because as the input voltage rises, the primary current rises more rapidly causing the comparator to trip more quickly thus reducing the pulse width. The secondaries are each full wave rectified and averaged using an L-C filter. The power supplies utilize a coupled inductor approach toward providing quasi-regulation for all non-controlled secondaries. Each secondary is magnetically coupled to the control winding and will exhibit a crude regulation that tracks changes in the control winding. Because each rail voltage does have a voltage monitor and scaling circuit (all diode or-ed), the control winding may be any one of these at any time. Load regulation in this coupled inductor structure is basically non-existent except for the controlled winding, and regulation of all other secondaries is controlled by drift in the controlled winding combined with load changes on all windings. The net regulation is therefore always linked to the controlled windings voltage and the distribution of load currents between the windings not being controlled and the controlled winding at any given time. One aspect of using regulated supplies for audio amplifiers is that for high power amplifiers whose input is basically in the 12.6 to 14.4 volt range, the power supply circuit topology is increasingly limited to doubleended forward converters as the power level increases. There are several reasons for this. One is that as the output power level increases, the primary currents increase and can become very large (into the multiple hundred ampere range). This can make necessary many large power supply transistors and large, expensive magnetics. The voltage requirement of these transistors in this configuration is double the input voltage plus ringing spikes. In this range, 50-60 volt high current transistors are readily available. The large ferrite power transformer cores are also readily available at moderate prices. The output filter inductor cores are less readily available in the larger sizes and can be somewhat expensive. So the power supply is designed to provide high power to the amplifier. However, the audio amplifier output currents can also all go to zero at once causing the inductor to ‘run dry’ for that time. In a regulated supply designed for operation in the inductor continuous conduction mode, this means that the minimum current flowing in the inductor must be sufficient to maintain flux in the core. If it is not, the output filter will cease to function like an averaging filter and will act as if the inductor were shorted. If this lasts long enough or recurs frequently enough, the output can increase to the peak of the secondary voltage, ie battery voltage times the transformer turns ratio. At approximately 70,000 pulses per second, this can take only a small fraction of a second. There are methods of assuring that the inductor does not ‘run dry’. Neither is inexpensive or small, both are effective. One commonly used method is to use a ‘swinging choke’ inductor configuration. This method uses stacked cores of differing permeabilities. A high permeability core is used to maintain inductance to very low values of current. As the current increases, the high permeability inductor saturates and the low permeability core provides reduced inductance to considerably higher current levels. There are composite cores available which can be used to medium car audio power levels. Another approach is to use a coupled inductor using low permeability core materials such as mollypermalloy powder (MPP). With this method, all the output windings are magnetically coupled through the core. This method requires that the turns ratio of the windings on the inductor be identical to the turns ratio of the transformer secondary windings. With this approach, any and all currents flowing in any secondary contribute to maintaining flux in the core and therefore inductance and averaging results. This is the method being utilized on the MB Quart Q-Series amplifier power supplies. The amplifier bias currents are sufficient to maintain regulation. A high frequency ferrite core transformer provides isolation between the power supply input power and the output secondaries. For the amplifier configuration being used, the amplifier rail voltages float on the output voltage. Housekeeping voltages for ground referenced circuits in the power supply are generated on separate windings. These windings are only quasi-regulated due to the magnetic coupling between windings and does not have monitor and scaling circuits associated with them. They are, however, referenced to ground potential and linear regulators on each provide solid rail voltages for all the control and logic circuits used in the power supply. 2/4 OHM OPTION This key feature of the power supply allows the user to select the speaker impedance being used with the amplifier. This circuitry then matches the power supply output rails to the required amplifier voltage rails for optimum performance at max power output for that speaker impedance. A logic command from the READ board determines how the power supply is configured. PROTECTION CIRCUITS The power supply is fully protected against over and under voltage operation. The PWM itself has an under voltage protection function. Both battery voltage and the precision reference voltage generated by the PWM are monitored. A sag of either voltage below a preset minimum results in the power supply being turned off. Backup over and under voltage protection is also provided by a separate monitor circuit. This circuit monitors the AP line input. It is designed to stop the converter when that voltage falls outside of the normal operating limits. Two speed fan control is provided on the power supply. The READ circuits monitor temperatures on the amplifier mosfet bars. When the temperature exceeds a preset level the fan is turned on at a low speed under READ control. If the temperature continues to increase, at another preset level the fans are switched to high speed operation. Independent of this, the temperature of each mosfet set the power supply is monitored. A surface mount thermistor mounted directly on each mosfet bar monitors the power supply mosfet case temperature. At a preset temperature, the PWM is stopped and hysteresis is provide to allow the unit to cool. This trip temperature is set to be higher than the amplifier fan control temperatures and will be the last resort thermal protection. This should occur only under the most extreme conditions since these amplifier units utilize two very quiet fans to move air across the heatsink fins for optimum control. Because the power supply utilizes current mode control, pulse by pulse current limiting is inherent to the design. So as not to interfere with normal operation at high amplifier output levels, this is set to occur at relatively high current levels.