Automatic level control

原文地址：http://www.robkalmeijer.nl/techniek/electronica/radiotechniek/hambladen/radcom/1995/07/page61/index.html

How does ALC work?

ALC stands for Automatic Level Control, and is a way of preventing excessive RF drive to the PA stages of your transmitter. By 'excessive' I mean a level of drive which will cause either distortion of SSB signals, increased levels of harmonics or physical damage to the final stages. Fig 5 shows the basic concept. It's a feedback control system very much like the AGC in your receiver. A DC control signal is derived from either the input or the output of the final amplifier, and fed back into the exciter to reduce the gain somewhere in the low-level stages. In the case of SSB or keyed CW the ALC signal is varying, so the time-constants in the ALC 'clop are designed to react quickly to the highesioutput signal and then hold the required ALC level for some time afterwards. Usually ALC is threshold-activated, which means that no ALC is applied until the output signals exceed a certain threshold level in the ALC detector. The result of all this is that no matter how loud you shout into the microphone, the RF output cannot exceed the level set by the ALC loop. Nonetheless, there are limits to the effectiveness of ALC, as outlined later.

Fig 5: Typical configuration of ALC system in an HF transceiver used with an external PA. ALC detection is at the output for most solid-state PAs, and at the input for most valve PAs.

Another use for the ALC loop is to reduce the drive level when a detector at the output of the transmitter senses an unacceptably high VSWR. Although this feature is not shown in Fig 5, it is very commonly used to avoid damage to solid-state final amplifiers.

All modern HF and VHF multimode transceivers have an internal ALC system, often linked to the RF POWER control. This frontpanel control introduces a constant DC bias into the ALC system to permanently reduce the level of drive. However, many transceivers also have an input for an ALC signal from an external power amplifier.

To understand what kind of external ALC signal is required, we need to dip into history. The first ALC circuits date back to HF exciters using valves, and it was very convenient to control the gain of an amplifier stage in the exciter by applying a few volts of negative bias to its control grid. Linear amplifiers following these exciters were designed to produce a suitable ALC signal, so 'a few volts negative' became an industry standard. Today, the valve exciters have been replaced by solid-state transceivers, but people tend to keep the same HF linear amplifiers. As a result, the EXT ALC inputs of modern HF solid-state transceivers still have to accept a negative-going voltage, and that standard seems here to stay. Unfortunately there are no comparable industry standards for VHF/ UHF transceivers and their add-on linear amplifiers. Although some transceivers may have an EXT ALC input, ALC outputs from either solid-state 'bricks' or valve PAs are quite rare. To ensure that the ALC signal from the final amplifier will always control the ALC loop, the transceiver's own ALC circuit is usually designed to react more readily to an external input than to its own internal ALC detector.

Fig 6: Typical ALC detector circuits. (a) Simplest circuit without threshold. (b) R1-R2 provides a threshold below which there is no ALC output. Unmarked components are mainly for RF bypassing.

Fig 6 shows some typical ALC detectors for HF linear amplifiers using valves. The detector is normally located at the input rather than the output of a valve PA to provide some protection to the control grid. Fig 6a is a very basic diode rectifier fed from the RF input via a capacitive divider. C1 and C2 are proportioned to provide the correct level of negative rectified DC. This circuit has the major drawback that there is some ALC action at all but the very lowest drive levels - in other words, ALC is acting all the time, which isn't really what we want. Fig 6b is an improved circuit that is very common (give or take the usual minor variations on diode RF detector circuits). The potential divider R1-R2 is fed from a convenient positive supply, and is wired to reverse-bias the RF detector diode Dl. There is no ALC output at all until the peak RF signal is large enough to overcome this reverse bias, after which the ALC signal goes increasingly negative. The component values are chosen to provide a healthy ALC voltage well before there is any risk of damage to the valve.

Misuse of ALC

ALC is not intended as a form of speech processing, or to correct grossly excessive levels of drive. Remember that it's a feedback system, and it can only begin to react after the final is being overdriven, Each time this happens, there will always be a brief moment before the ALC loop 'catches hold' and reduces the drive. These spikes of overdrive can cause severe sideband splatter and may even damage the external PA.

When the transmitter is working properly, the ALC should be operating only on speech peaks - not all the time. The output power of the transceiver needs to be broadly compatible with the drive power requirements of the external linear. For example, a typical '100W' HF transceiver driving a linear requiring 5070W for full output will be nicely controlled by an industry-standard ALC feedback signal from the linear. When your MIC GAIN control is set correctly, the ALC meter on the transceiver should just pop-up occasionally on speech peaks. However, if you are using speech processing in the transceiver, the drive level spends a lot of time quite close to its peak value, so there may be only a narrow range between no ALC action and almost fulltime ALC. If so, you must turn down the output of the speech processor until the ALC hardly ever operates. Used that way, it's the speech processor that primarily determines your maximum RF output level, with the ALC acting only as a 'long-stop' to catch any transient peaks.