Electronic feedback
Aug 1, 2004 12:00 PM, By John Battison, P.E., technical editor, RF
Most of the radio transmitters in use today depend on feedback, negative or positive, for their successful operation. Positive feedback can be found in the RF side of transmitters and negative feedback in the audio portion. However, as equipment designers become more esoteric in their circuitry we sometimes find feedback voltages of unexpected sign in unusual places and doing unusual things.
Feedback is simply described as the transfer of controlled amounts of energy or data, from the output of a stage back into its input circuit so that the output and performance of the device are affected by the feedback signals. These signals may be in phase or out of phase, i.e. positive or negative, depending on the purpose of the feedback system.
The good and the bad
If the feedback signal is in phase it is known as regenerative because it adds to the signal passing through the stage. Unless properly controlled, this can lead to instability and oscillation, and is normally undesirable except in particular cases such as stages that are required to oscillate and produce RF or AF voltages. RF amplifiers are particularly susceptible to undesired feedback through internal input and output capacitances in the electron device. This is counteracted by a neutralizing circuit, which generally requires readjustment or re-neutralizing, when a tube is changed.
An example of a basic inverse feedback circuit.
A second cause of undesired feedback is magnetic (mutual coupling) feedback between the input and output circuits in an amplifier. This occurs when plate voltages that are 180� out of phase get into the grid or input circuit.
If the stage amplification factor is greater than the attenuation of the out-of-phase signal, the stage will oscillate at a frequency other than the desired operating frequency. This will be unstable and lower than the tuned tank-circuit frequency. This kind of undesired feedback is eliminated in the circuit design by careful shielding and inductance layout.
Inverse feedback, often called degenerative, has the opposite effect and is useful in transmitter design. In this case, the out-of-phase voltage tends to oppose or reduce the desired signal. However, not only is the desired signal reduced, but undesired voltages such as noise and hum are reduced. A secondary benefit is often an improvement in the frequency response of the stage because low levels of feedback at certain frequencies result in low gain reduction. This increases basic gain at these frequencies and tends to improve the overall response. At higher audio frequencies up to 30dB of feedback can be used. As audio frequencies increase the output and gain tend to decrease. Sometimes a sudden, quick rotation in the phase shift occurs due to resonances in transformers and stray capacities. This may result in phase lag, which can slow the feedback signal and cause oscillation at supersonic frequency, which may be as high as 50kHz. This, of course, can result in distortion, overload and possible damage to the transmitter caused by the spurious signal.
As would be expected the stage gain is reduced. This means that the input voltage has to be increased to compensate for the stage gain loss. The degree of gain reduction is noted in decibels. For instance, raising the input by 14dB to maintain the same output indicates the presence of 14dB of inverse feedback.
Feedback in transmitters
The method and component values required to produce this desired feedback are determined by the transmitter design engineer. If the component values of the feedback ladder are changed by the user there will be a discrepancy from the manufacturer’s specifications. This important point can be easily overlooked. The consequence may be complaints from the program director or general manager that the signal is noisy and has hum. Sometimes component values change with continued operation and slow deterioration in signal quality is not noticed. Check a few component values and run a resistance and capacitor check at maintenance time to ensure that no substantial changes have occurred over the passage of time.
The application of inverse feedback in a transmitter design offers several advantages to the circuit engineer. Apart from minimizing distortion it can also be used to minimize hum. Power tube filaments can produce a surprising amount of hum if proper precautions are not taken, especially in old transmitters.
The heavy current flow in the filament of a large power tube produces a strong magnetic field, which can slightly deflect the electron path in the tube. Because this magnetic field varies at the power line frequency, the election flow is modulated at this frequency and an ac hum develops in the carrier. Many years ago, before inverse feedback was developed, it was standard practice to: use a dc source for the filaments to eliminate PA hum.
Fortunately, inverse feedback can be applied to most of the transmitters currently in use today. This includes all transmitters in which the modulated RF output envelope is strictly in phase with the modulating signal at the point of feedback injection. It is essential that the phase of the feedback signal agree with the phase of the audio signal, otherwise the benefits of inverse feedback will not only be lost but audio quality will be affected.
To produce the necessary in-phase copy of the modulating audio signal a small sample of the modulated RF output is rectified and an audio voltage replicating the RF envelope is produced. This is applied at the input of the modulator stage.
The major characteristic required for successful inverse feedback application is a modulation envelope, which closely replicates the phase of the audio signal at the point of feedback injection. A problem can arise in plate-modulated transmitters because the inevitable phase shift that occurs in modulation and driver transformers makes it difficult to apply inverse feedback over the whole stage. This excessive phase shift occurs at higher audio frequencies.
Because of this excessive phase shift caused by the various transformers and other iron-cored devices, feedback is normally only used in the modulator stage and excluding the modulation transformer. This will correct distortion and hum produced in this stage, but not any produced in the RF stage.
Inverse feedback can be applied in most transmitters in use today, provided that audio phase remains constant over the whole audio frequency range in use. It can be used with the Doherty system, grid modulation and class B linear amplifiers and circuits where phase relationships can be precisely maintained so that gain and noise reduction benefits will result.
Feedback is not a panacea that will correct an improperly tuned or misadjusted transmitter. It will not cure distortion resulting from negative overmodulation. Instead, it merely causes heavier negative overmodulation.
E-mail Battison atbatcom@bright.net.