Audio amplifiers built and repaired by radio amateurs often become a source of headaches due to the subsequent hum of alternating current with a frequency of 50 Hz, noticeable audibly in loudspeakers or telephones (headphones).
If this happens, you should check whether the microphone is correctly connected to the PU (pre-amplifier) - the common wire of the device must be connected to the braided screen of the cord - and also whether the output of the PU and the input of the power amplifier (PA) are correctly connected. The fact is that sometimes two amplifiers (preliminary and PA) are used in one device, having different polarities of the common wire. In amplification circuitry, such inclusion is not a problem; the main thing for a high-quality amplifier is the compatibility of the input impedance and the amplifier’s own noise level. However, incorrect (incorrect) connection of the amplifiers to each other and the pre-amplifier to the sound source (for example, to a microphone) is often the cause of the hum with a frequency of 50 Hz.
Practical hum elimination in amplifiers 34
To localize this problem, there is a simple way to connect sound sources to the pre-amplifier (this can be not only a microphone, but also another source with a low signal level of up to 10 mV). Let's analyze this method based on an example with connecting a microphone.
The central conductor in the braided microphone cord is connected to the input of the control unit, as a rule, to an isolation capacitor, limiting resistor or voltage divider. The braid (screen) is not connected directly to the common wire, but in series with the RC circuit (a parallel connected resistor with a resistance of 2 kOhm (±20%) and an oxide capacitor with a capacity of YumkF with the same tolerance for possible deviation from the nominal value). Here, the resistance of the resistor and capacitor is calculated for devices with a power supply voltage in the range of 6-20 V.
The positive plate of the oxide capacitor in this case is turned on in accordance with the polarity of the power source so that if the common wire is connected to the “minus” of the power source, then the oxide capacitor is connected to the common wire with the negative plate, and vice versa
This method eliminates hum in most amplifiers with different common power supply wires, including older tube amplifiers where filtering of the rectified voltage leaves much to be desired. In most cases, in this way it was possible to solve the problem of background with a frequency of 50 Hz in dynamic heads, which arises after replacing a standard microphone with another (with similar electrical characteristics), as well as in the case of replacing a high-impedance microphone (for example, MD-47, equipped with a matching transformer and having a resistance of 1600 Ohms) to a low-impedance microphone type MD-201 with a coil resistance of 200 Ohms or similar in electrical characteristics.
AC background
Reasons leading to the appearance of AC background:
- Contact with AC power circuits into low frequency stages.
- Influence of electric and magnetic fields to low-frequency circuits due to poor placement of individual wires and parts.
- Background overlay on high frequency circuits or a modulating background, audible only when the receiver is tuned to a radio station.
The presence of a constantly audible background indicates that it is superimposed in one way or another on the low-frequency circuit of the receiver. Therefore, first of all, you should check whether the DC ripple is sufficiently smoothed out by the rectifier filter. For this purpose, a calibrated high-voltage capacitor with a capacity 40-100 µF connected in parallel first to the second and then to the first capacitors of the smoothing filter of the receiver or amplifier being repaired.
If this gives the desired effect, then you need to replace one or both of the anti-aliasing filter capacitors or increase the capacitance of the capacitors in the anode or grid decoupling filters. If such an event does not cause a noticeable weakening of the background, then most likely there is a second reason.
To quickly detect in which low-frequency cascade the background is superimposed, remove all the lamps one by one, starting from the input and up to the pre-terminal one, and monitor which of them stops the background when removed.
The final stage lamps cannot be removed when the power is on., since the resulting sharp decrease in the rectifier load leads to a significant increase in the anode voltage, which in turn can cause breakdown of the smoothing filter capacitors.
Frequent causes of hum due to interference are breaks in shielding shells and the appearance of a leak between the filament and the cathode at the input lamp of the low-frequency amplifier. The cause of the modulating background can also be poor pulsation smoothing voltages supplying high-frequency lamps. The input stages of receivers (RF amplifier and converter), as well as the local oscillator, are especially sensitive to this, and therefore an additional smoothing filter cell is sometimes installed to power these stages.
The modulating background of alternating current, audible only when receiving local stations, is easily eliminated by blocking the anode of the kenotron to its cathode or ground ( pic. 1 ), as well as blocking the shoulders of the step-up winding of the transformer with capacitors with a capacity 0.005-0.01 µF; the operating voltage of these capacitors must be no less than triple the voltage of the arm of the step-up winding of the power transformer ( 1000-1500 V).
Before eliminating the background that appears when receiving radio stations, you need to make sure that the background modulation occurs in the receiver and not in the transmitter. To do this, it is best to check the reception of the same radio station using another receiver.
Rice. 1. Elimination of modulating background
Particular attention should be paid to methods for eliminating the background in equipment with direct incandescent lamps when their filaments are powered with alternating current. It is necessary here precise balancing of the filament circuit, which is not always ensured by the device for tapping the midpoint of the filament winding. A more effective measure is to include a low-resistance potentiometer between the terminals of the filament, the slider of which should be considered as a terminal from the cathode of the lamp. Precise balancing of the thread is carried out when the power is turned on by ear by setting the potentiometer slider to a position in which the alternating current background is least audible.
A similar measure can significantly reduce the background coming from the filament circuits in low-frequency amplifiers with a high gain (in tape recorders, microphone amplifiers). If the device is installed again, the background noise may be due to poor placement of individual circuits and transformers.
It is important to identify not only which circuit is affected by the unwanted influence, but also which circuit produces this influence. To do this, we apply a method of changing the reactivity of subsequent circuits, which consists in connecting a capacitor of larger or smaller capacity to the anode load resistances of the lamps, starting from the output of the receiver, and so gradually approaching the source of self-excitation or its complete cessation.
Let's assume that connecting a capacitor to the output transformer only reduced the volume without changing the nature of self-excitation. This means that the final stage is not covered by self-excitation and the circuit that creates an undesirable effect on the amplifier input must be sought before it. But, if, for example, when a capacitor is connected parallel to the primary winding of the output transformer, self-excitation is removed or its character changes, then either this circuit or the subsequent one (the circuit of the secondary winding of the output transformer) affects the input circuit of the amplifier.
Having determined between which two circuits a harmful interaction occurs, it is not difficult to carefully examine their installation to find the location of the relationship and to eliminate self-excitation by shielding or partially changing the installation of these circuits.
Rice. 2. Electronic light indicator
- Short probe
- Power hose
- No hesitation
- There are fluctuations.
Self-excitation via HF does not always manifest itself in the form of an extraneous sound constantly heard in the loudspeaker; more often it can be judged by the presence of loud whistles when tuning into a station or by characteristic distortions, a sharp decrease in volume and other specific features. Such self-excitation can be detected using a lamp voltmeter or an electronic light indicator, which are connected in series to all oscillatory circuits of the cascades under study ( pic. 2 ).
Audio frequency amplifiers (AF), created and repaired by radio amateurs, often become a source of “headaches” due to the subsequent hum of alternating current with a frequency of 50 Hz, noticeable audibly in loudspeakers and telephones.
If this happens, you should check whether the microphone is correctly connected to the pre-amplifier - hereinafter referred to as the PU (the common wire of the device must be connected to the braided screen of the cord), and also whether the output of the PU and the input of the power amplifier (PA) are correctly connected. The fact is that sometimes two amplifiers (preliminary and PA) are used in one device, having different polarities of the common wire. As you know, in amplification circuitry such inclusion is not a problem—the main thing for a high-quality amplifier is the compatibility of the input impedance and the noise level. However, incorrect (incorrect) connection of the amplifiers to each other and the pre-amplifier to the sound source (including a microphone) is often the cause of the hum with a frequency of 50 Hz.
In order to localize this problem, I propose a simple method involving the inclusion of sound sources to the pre-amplifier (this can be not only a microphone, but also another source with a low signal level of up to 10 mV). Let's analyze this method based on an example with connecting a microphone.
The central conductor in the braided microphone cord is connected to the input of the amplifier (PA) according to the circuit, usually to a coupling capacitor, limiting resistor or voltage divider.
The braid (screen) is not connected directly to the common wire, but in series with an RC circuit, which represents a parallel connected resistor with a resistance of 2 kOhm ± 20% and an oxide capacitor with a capacity of 10 μF with the same tolerance for possible deviation from the nominal value.
Here, the resistance of the resistor and capacitor is calculated for devices with a power supply voltage of 6 to 20 V.
The positive plate of the oxide capacitor in this case is connected in accordance with the poles of the power source (PS) so that if the common wire is connected to the “minus” of the PS, then the oxide capacitor is connected to the common wire with the negative plate, and vice versa.
This method eliminates hum in most amplifiers with different common power supply wires, including older tube amplifiers where filtering of the rectified voltage leaves much to be desired.
In most cases, in this way it was possible to solve the “problem” of background with a frequency of 50 Hz in dynamic heads, which arises after replacing a standard microphone with another (with similar electrical characteristics), as well as in the case of replacing a high-impedance microphone (for example, MD-47, equipped with a matching transformer and having a resistance of 1600 Ohms) to low-resistance (type MD-201).
Literature: Andrey Kashkarov - Electronic homemade products
One of the main problems that we have to contend with when designing and creating high-quality tube ULFs is the AC background. In this case, the AC background is understood as the voltage existing at the output of the amplifier, in addition to the useful signal, which has
frequency equal to or a multiple of the power supply frequency. The presence of the AC background in question in any sound reproducing device is a very serious drawback, since such a background narrows the dynamic range of the amplifier and sharply worsens the subjective impression of the reproduced signal.
The main reasons causing the appearance of background noise in low-frequency tube amplifiers can be conditionally divided into several groups, two of which are the main ones: ripple of supply voltages and alternating current pickup into various circuits in the amplifier. Therefore, background elimination should be carried out in two directions, namely by improving the filtering of supply voltages and reducing the influence of interference.
One of the main reasons for the appearance of background in tube ULFs is pulsation of the rectified voltage that supplies the circuits of the anodes and screen grids of the lamps. In this case, the higher the internal resistance of the lamp, the less the influence of ripples. As is known, the internal resistance of pentodes is greater than that of triodes, therefore, from this point of view, it is better to use pentodes in the first stages of a tube amplifier. In addition, the background noise arising from voltage ripple can be reduced by improving the circuit and improving the parameters of the rectifier.
When using a choke in a power supply filter, this element largely determines the background level. The inductance of the inductor is usually on the order of 520 H and should depend little on the load current. To improve filtering, it is useful to bypass the inductor with a capacitor, the capacitance value of which is selected so as to form a circuit tuned to the ripple frequency (100 Hz with full-wave rectification). The schematic diagram of a filter with a circuit of this type is shown in Fig. 3.34.
The reasons for the occurrence of an alternating current background may lie in the fact that either the screen grids of the lamps are powered by an insufficiently smoothed voltage, or the anode current unnecessarily loads the elements of the smoothing filter.
often powered by voltage with the same ripple. However, the permissible screen voltage ripple for most terminal pentodes and beam tetrodes is 2030 times less than the anode voltage ripple. Therefore, the screen grid circuits must be fed through an additional smoothing circuit.
In order to reduce the influence of leakage between the cathode and the filament, it is sometimes recommended for the first stages of the amplifier, instead of automatic bias circuits, to use a separate rectifier with a filter, with the help of which a constant bias voltage is generated, supplied to the lamp grid. Schematic diagrams of possible variants of such rectifiers are shown in Fig. 3.35.
AC background
Reasons leading to the appearance of AC background:
- Contact with AC power circuits into low frequency stages.
- Influence of electric and magnetic fields to low-frequency circuits due to poor placement of individual wires and parts.
- Background overlay on high frequency circuits or a modulating background, audible only when the receiver is tuned to a radio station.
The presence of a constantly audible background indicates that it is superimposed in one way or another on the low-frequency circuit of the receiver. Therefore, first of all, you should check whether the DC ripple is sufficiently smoothed out by the rectifier filter. For this purpose, a calibrated high-voltage capacitor with a capacity 40-100 µF connected in parallel first to the second and then to the first capacitors of the smoothing filter of the receiver or amplifier being repaired. If this gives the desired effect, then you need to replace one or both of the anti-aliasing filter capacitors or increase the capacitance of the capacitors in the anode or grid decoupling filters. If such an event does not cause a noticeable weakening of the background, then most likely there is a second reason.
To quickly detect in which low-frequency cascade the background is superimposed, remove all the lamps one by one, starting from the input and up to the pre-terminal one, and monitor which of them stops the background when removed.
The final stage lamps cannot be removed when the power is on., since the resulting sharp decrease in the rectifier load leads to a significant increase in the anode voltage, which in turn can cause breakdown of the smoothing filter capacitors.
Frequent causes of hum due to interference are breaks in shielding shells and the appearance of a leak between the filament and the cathode at the input lamp of the low-frequency amplifier. The cause of the modulating background can also be poor pulsation smoothing voltages supplying high-frequency lamps. The input stages of receivers (RF amplifier and converter), as well as the local oscillator, are especially sensitive to this, and therefore an additional smoothing filter cell is sometimes installed to power these stages.
The modulating background of alternating current, audible only when receiving local stations, is easily eliminated by blocking the anode of the kenotron to its cathode or ground ( Fig.1 ), as well as blocking the shoulders of the step-up winding of the transformer with capacitors with a capacity 0.005-0.01 µF; the operating voltage of these capacitors must be no less than triple the voltage of the arm of the step-up winding of the power transformer ( 1000-1500 V).
Before eliminating the background that appears when receiving radio stations, you need to make sure that the background modulation occurs in the receiver and not in the transmitter. To do this, it is best to check the reception of the same radio station using another receiver.
Particular attention should be paid to methods for eliminating the background in equipment with direct incandescent lamps when their filaments are powered with alternating current. It is necessary here precise balancing of the filament circuit, which is not always ensured by the device for tapping the midpoint of the filament winding.
A more effective measure is to include a low-resistance potentiometer between the terminals of the filament, the slider of which should be considered as a terminal from the cathode of the lamp. Precise balancing of the thread is carried out when the power is turned on by ear by setting the potentiometer slider to a position in which the alternating current background is least audible.
A similar measure can significantly reduce the background coming from the filament circuits in low-frequency amplifiers with a high gain (in tape recorders, microphone amplifiers). If the device is installed again, the background noise may be due to poor placement of individual circuits and transformers.
It is important to identify not only which circuit is affected by the unwanted influence, but also which circuit produces this influence. To do this, we apply a method of changing the reactivity of subsequent circuits, which consists in connecting a capacitor of larger or smaller capacity to the anode load resistances of the lamps, starting from the output of the receiver, and so gradually approaching the source of self-excitation or its complete cessation.
Let's assume that connecting a capacitor to the output transformer only reduced the volume without changing the nature of self-excitation. This means that the final stage is not covered by self-excitation and the circuit that creates an undesirable effect on the amplifier input must be sought before it. But, if, for example, when a capacitor is connected parallel to the primary winding of the output transformer, self-excitation is removed or its character changes, then either this circuit or the subsequent one (the circuit of the secondary winding of the output transformer) affects the input circuit of the amplifier.
Having determined between which two circuits a harmful interaction occurs, it is not difficult to carefully examine their installation to find the location of the relationship and to eliminate self-excitation by shielding or partially changing the installation of these circuits.
Self-excitation via HF does not always manifest itself in the form of an extraneous sound constantly heard in the loudspeaker; more often it can be judged by the presence of loud whistles when tuning into a station or by characteristic distortions, a sharp decrease in volume and other specific features. Such self-excitation can be detected using a lamp voltmeter or an electronic light indicator, which are connected in series to all oscillatory circuits of the cascades under study ( Fig.2 ).