Take as an example a simple H-bridge amplifier for a brush motor. If the upper left and lower right transistors are fully on, you get the bus voltage at the left side of the motor and the bus return at the right, trying to push current left to right. If the upper right and lower left transistor are fully on, you get the bus voltage at the right and return at the left, trying to push current to the left.
In sign/magnitude PWM, the sign signal selects which of these two pairs of transistors can be turned on. The magnitude signal varies the on time from 0 to 100% for the selected pair. To command a net zero voltage, the PWM signal is always off. To command a +/-10% voltage, the PWM signal has a 10% duty cycle, with the sign controlling the direction.
In 50/50 PWM, there are two separate PWM signals, one for each of these pairs. These signals are complementary to each other -- when one is on the other is off (there must be a slight "deadtime" between turning off one and turning on the other, but that is a detail). To command a net zero voltage, each signal is on 50%. To command a +/-10% voltage, you use a 55%/45% or a 45%/55% arrangent. The difference in the on-time percentage controls the net voltage.
50/50 PWM creates more current ripple in the motor, because it exposes the motor to the full plus and minus bus voltage each cycle, instead of going between the bus voltage and zero voltage. However, it provides a better transition through the zero-voltage state, reducing the resulting deadband, because there is nothing special about this state.
Both are DC type PWM. Sign-magnitude PWM is also called in the literature unipolar, or half-bridge, or non-regenerating, or circulating PWM technique.
Compare with 50/50 PWM a.k.a. bipolar, or full-bridge, or regenerating, or non-circulating PWM.
Advanced unipolar PWM combines the advantages of both:
- low switching power electronics loss, low filter capacitor and additional motor PWM caused iron losses, better EMI / RFI of unipolar PWM;
- high servo performance of bipolar PWM.
Contact me for further details on advanced unipolar PWM, motor PWM iron loss etc.
The PWM wave form is derived from a reference sine waveform a given required output frequency, a triangular wave is then applied to the sine wave to determine the length of the firing pulses for IGBT's .
From my understanding this is what gives us the sign magnitude of the PWM waveform because the resultant firing pulses will vary as the reference sine wave is varied.
PWM 50/50 is more related to older drives and some DC trapezoidal servo drives where there are only 4 transistors and they are pulsed to give a trapezoidal output to the motor which gives poorer low speed performance compared to a true PWM or sinusoidal drive which can give very good low speed performance.
