This proposed high-fidelity amplitude detector design resulted from another use of the class B push-pull amplifier. The extremely low harmonic distortion produced by such an amplitude detector is the result of the high linearity achieved through negative feedback.
Introduction
The simplified Class B amplifier schematic composed of two complementary symmetry transistors shown in Fig.1. Each transistor will work on one half cycle of the input waveform. The NPN transistor VT1 works as an emitter-follower and drives a positive current into a load according to the positive half-wave of the input signal (red). On the contrary, the PNP transistor VT2, that also works as an emitter-follower, drives a negative current into a load according to the negative half-wave of the input signal (blue). The sum of the positive half-wave and the negative half-wave of the current in load produces the output signal to maintain the same shape as the input except the Crossover Distortion (violet).
In the amplifier circuit in Fig.1, the direction and shape of the emitter current of the transistors attract attention. It is obvious that each transistor works as an envelope detector of the input signal: transistor VT1 of its positive half-waves, and transistor VT2 of its negative half-waves.
The idea of a high-fidelity amplitude detector is to use the emitter current in one of two transistors of a class B amplifier output stage as the output signal of the detector. To improve linearity and reduce distortion of the detector the recovered on the RL-load full signal feeds into the deep negative feedback circuit.
Schematic of High-fidelity Amplitude Detector with Negative Feedback
Fig.2 shows the schematic diagram of an envelope detector with negative feedback:
The class B push-pull amplifier is easily recognizable in the diagram. Each “arm” of this push-pull amplifier is the Sziklai pair of the bipolar transistors: one N-P-N and one P-N-P transistor. The loading of the amplifier is R11 resistor. This circuit differs from the classic class B push-pull amplifier circuits by resistors R9 and R10, the resistance of which is chosen equal to the resistance of R11 resistor. The positive and negative half-waves of the alternating current in the loading resistor R11 also flow in R9 and R10 resistors respectively. This half-waves create the output signal voltage of the envelope detector on R9 and R10 resistors (XT3 and XT4 pins).
The input signal is fed to the point XT1 and then through the capacitors C2 and C3 to the bases of the right transistors of the VT1 and VT2 transistor sets. The left transistors work like diodes to make some little bias current of the right transistors. The bias current of the right transistors is practically equal to the current through the left transistors because the transistor pairs work in “current mirror” mode. The level of the current through the left transistors is setting by the voltage of the reference adjustable shunt regulator D1 and resistors R5 and R6. The bias current of the right transistors of the VT1 and VT2 transistor sets can be adjusted with resistor R2.
When an input signal is absence the bias current of the right transistors of the VT1 and VT2 transistor sets have to be so that the transistors VT3 and VT4 are close to opening, but still remained closed throughout the operating temperature range. That is, the current in transistors VT3 and VT4 either have not be more than a few microamperes, or have to be absent altogether. The value of this current can be checked by measuring the voltage on the resistor R10 (or R9). In this mode, the class B push-pull amplifier is working as a low-distortion highly sensitive amplitude detector.
Testing such an amplitude detector on a real signal
To verify the high-fidelity and its repeatability of such an amplitude detector four modules were made according to the schematic diagram shown in Fig.2. These modules were made on double-sided printed circuit boards with surface-mounted components soldered on them as shown in Photo 1:
In order to measure the real characteristics of the amplitude detector module the measuring instruments were connected to it according to the schematic shown in Fig.3:
Every the amplitude detector module was tested with an amplitude-modulated signal with a carrier frequency of 465 kHz, a modulating signal frequency (envelope) of 1 kHz, and a modulation index of 30%. The amplitude detector operation is illustrated by the oscillograms in Photo 2 and Photo 3.
Photo 3 shows the oscillogram of the output signal when an amplitude-modulated signal with a carrier level of 0.5 V (RMS) and a modulation index of 30% is applied to the input of the amplitude detector (point XT1):
The product of filtering the signal at the amplitude detector output by the RLC Low-Pass Filter consists of a certain DC component and an audio frequency signal. The magnitude of the DC component is proportional to the amplitude of the carrier of the input amplitude-modulated signal, and the audio frequency signal according to the envelope of this signal. An amplifier connected to the output of the low-pass filter amplifies both of these components.
Specifications of the amplitude detector
As mentioned above, every the amplitude detector module was tested with an amplitude-modulated signal with a carrier frequency of 465 kHz, a modulating signal frequency (envelope) of 1 kHz, and a modulation index of 30%. The total harmonic distortion of the demodulated signal was measured at different levels of the AM signal at the input of the amplitude detector. The results of measurements for all of four AM-detector modules were as follows:
Table 1
Parameter | Voltage of Input AM-signal (RMS) | |||
10 mV | 30 mV | 100 mV | 500 mV | |
Voltage of Output Audio Frequency Signal (RMS) | 0,47 mV | 2,16 mV | 8,24 mV | 45,0 mV |
Total Harmonic Distortion, less than | 1,60% | 2,00 % | 0,70 % | 0,43 % |
The voltage of the demodulated audio frequency signal is specified taking into account the gain of the final amplifier of the measuring circuit shown in Fig. 3. The gain of this amplifier was set equal to 51 for a small signal (less than 150 mV) and equal to 6 for a large one (over 150 mV).
The dependence of the total harmonic distortion on the level of the input signal at the same modulation index is shown in Fig.4. This dependence has a pronounced maximum at an input signal level of about 20 mV.
In order to explain this behavior of the amplitude detector, its transfer characteristic was built as a dependence of the DC voltage at the output of the passive RLC low-pass filter (see the diagram in Fig.3) on the amplitude of the unmodulated signal at the detector input. This dependence in a simplified form, that is, without reference to specific values, is shown in Fig.5.
The transfer characteristic of the AM-detector is represented by a blue line in the diagram. The voltage range of the input HF-signal, in which the transfer characteristic is almost linear, starts from 20 mV (RMS) and extends up to 1V (RMS). In this voltage range of the input HF-signal, the linear approximation of the transfer characteristic of the AM-detector represented by the red line in the diagram. To approximate the transfer characteristic of the AM-detector the least-squares approach was applied. The standard deviation of the transfer characteristic from its linear approximation is less than 0,55%. It is quite consistent with the total harmonic distortion measurement results given in Table 1. Such a small value of the standard deviation in such a wide voltage range of the input HF-signal attests the high linearity of the amplitude detector with deep negative feedback.
Where You can find out the negative feedback in the circuit
The feeding a signal from the collectors of VT3 and VT4 transistors into the emitters of the right transistors of the VT1 and VT2 transistor sets is the negative feedback in this amplitude detector. Since the entire signal recovered at the load (resistor R11) is fed into the emitters of the input stage transistors, the negative feedback is fully 100 percent and the class B push-pull amplifier works as an input signal voltage follower.
Negative feedback makes sense only when the loop gain of the amplifier covered by negative feedback is much greater than unity. To determine the loop gain in our circuit, let’s break the negative feedback circuit as shown in Fig. 6:
In this case, the negative feedback is opened in AC (capacitor C blocks the HF-signal from entering the emitters of the input transistors), but closed in DC (resistor R) so as not to disturb the operation mode of the transistors. In order to determine the loop gain of the class B push-pull amplifier the very small signal was applied to its input and the signal at its output (on resistor R11) was measured. The gain for every AM-detector module was greater than 100, which is the depth of negative feedback in the presented amplitude detector.
Copyright © Sergii Zadorozhnyi, 2019
Please give us your valuable comment
You must be logged in to post a comment.