All diode-based rectifier circuits suffer from a signal-level-dependent frequency response. This response occurs because the dynamic impedance of the diodes varies as the signal level changes. The result is a slower frequency response and distorted rectified outputs at low signal levels compared with high levels.

An alternative approach to a half-wave rectifier is to use the input-signal polarity to control the signal channel gain. The input signal shuts down the gain for one polarity and passes the signal on for the other.

You need a high-speed comparator to detect zero crossings and an adjustable gain stage with broadband signal and gain-control channels to implement this approach (Fig 1a). A wideband, current-feedback amplifier with an output-clipping feature acts as the comparator to control a high-speed AGC amplifier's gain. The polarity of I_SIG controls the comparator operation. IC₁ connects the comparator operation and a flexible output voltage that matches the gain control range of IC₂. You can extend the comparator's dynamic range by connecting the optional compression network to the inverting node of IC₁. IC₂ is an AGC amplifier with a channel bandwidth greater than 200 MHz and can switch from full on to off in less than 2 nsec.

Fig 1b shows test results with a 10-MHz, 1V p-p input. If you carefully match the delays for the signal and gain-control channels, you can achieve higher speeds. You can produce a full-wave rectifier by summing the original input into the AGC's output with the correct gain. The circuit passes negative half-cycles around the gain-control path directly to the output. This path and the gain-controlled path act on positive half cycles. (DI #1645)