|
|||
|
|
|
|
|
|
| About Feed-Forward Correction This new patented (US Patent 6,653,824) technique allows an AC regulator to deliver clean, pure, constant-voltage sine wave AC output even when the input AC is poorly regulated, noisy and distorted. AC power is commonly supplied worldwide at nominal frequencies of 50 Hz or 60 Hz and nominal voltages ranging from 100 volts to 250 volts rms. Deviations from nominal “line voltage,” although expected to some degree in any utility system, can become large enough to cause problems in the operation of certain electrical and electronic equipment. These problems tend to be more severe with electric power supplied in developing countries and with portable power generating systems, such as those used by “location” crews in audio, video, or film productions, for example. Voltage regulation of such power can be very beneficial to the operators of all kinds of AC operated equipment, which might otherwise malfunction. Devices used to eliminate such fluctuations are called AC voltage regulators. They fall into two broad categories, passive and active. To accomplish their task, active units use electronic amplifiers while passive units use transformers, sometimes teamed with inductors and capacitors. Of the existing passive AC regulators, the ferro-resonant type is by far the most widely used. Because of the inherent reliability of passive components, these units tend to be quite robust and dependable. But they are also heavy, expensive, inefficient and have response times that are quite slow. . Closed-Loop or Feed-Back Regulation Of the existing active AC voltage regulators, most use some form of “closed loop” or “negative feedback” to control their output - attempting to hold it constant. The closed loop control process requires measurement of the output's effective or “rms” value. Because the output is an AC voltage, its effective value can only be determined by a process requiring averaging or integration time. This limits the ability of the system to respond to rapid fluctuations of the input. Actual correction is generally accomplished by adding voltage to or subtracting voltage from the incoming line voltage. This type of regulator is often referred to as a “buck/boost” regulator and the most common designs work by switching taps on a transformer in response to the output voltage measurement and subsequent comparison to a stable reference. The switching is usually done by solid-state switches. Another version of this method is the motor-driven variable auto-transformer or Variac . This embodiment effectively provides a very large® number of transformer taps - selected via the motor-driven "wiper." Another method, which also uses negative feedback, can regulate its output in a smoother, more continuous fashion than the "stepped" regulation provided by tap- switching. This technique typically uses a power amplifier to drive the primary of a transformer whose secondary voltage can be made to either "buck" or "boost" the incoming line voltage. The "error signal" generated by comparing the measured output voltage to a stable reference is then used to modulate the amplitude and phase/polarity of a correction signal that drives the amplifier, which subsequently corrects the regulator’s output voltage. Regulators using feedback to stabilize their outputs are potentially unstable. They compare the regulator output to a reference to derive an error signal used to change the output. Because the error signal changes the output, the error signal also changes, which again changes the output. This process of feeding back on itself is potentially unstable, since corrections cannot be made instantaneously and even small time delays can cause over-corrections known as “overshoot.” In some cases these iterative over- correction and under-correction cycles will sustain themselves, resulting in continuous oscillation or “hunting.” Feedback systems must be very carefully designed to avoid these instabilities. Certain types of loads on such regulators will introduce additional time delay into the feedback loop, causing an otherwise stable system to become unstable. Similar problems afflict audio power amplifiers with certain speaker loads. Open-Loop or Feed-Forward Regulation On the other hand, a feed-forward system compares the regulator input to a reference to determine how much correction is necessary - in a sense anticipating corrective action before a disturbance can reach the output. Such systems are sometimes called “open-loop” systems. Since such a regulator doesn’t “see” its corrected output, the system is inherently stable under all load conditions and oscillation is simply impossible. The new, patented regulator uses an error feed-forward technique to completely avoid the time lag associated with an output measurement. It continuously and instantaneously compares the waveform of the incoming AC voltage to an internally- generated “reference” waveform that’s phase-locked with the incoming AC line. This comparison is done by a differential amplifier that subsequently drives a high-efficiency power amplifier. The power amplifier output acts to either "buck" or "boost" the incoming AC voltage seen at the output. By properly scaling of circuit gain, a drop in input voltage is instantaneously corrected by a corresponding boost voltage from the amplifier. Conversely, a surge at the input is corrected by a corresponding buck voltage from the amplifier. This method is so fast that it can correct the wave-shape of the incoming AC “on the fly,” delivering a clean sine-wave output. “Flat-top” distortion, which reduces the peak value of the AC, is the most common type. Removing such waveform distortion can significantly improve performance of most audio power amplifiers and many video displays because their internal power supplies depend on the peak value for proper output. The most useful benefits of the feed-forward technique are speed (to correct waveform distortion) and stability (to prevent over-correction). |
||||
|
|
|