The task of undistorted transmission of a sound program from the performer to the listener is as old as the world. Like the world of electroacoustics ...
Raymond Skuruls is a radio engineer and sound engineer, founder and owner of Acoustic Power Lab. In 2005, after three years of work, he received a Latvian patent (LV1334213) for a new technology for correcting the frequency characteristics of loudspeakers. Pro Sound News Europe lists AJFL correction technology as one of the top three innovations in this area in Europe. Following the results of the AES exhibition in New York, the new development was awarded the 2007 Excellence prize. In 2010, the author develops a version of the technology for use in a car.
One of the prerequisites for this is the absence of linear distortion. From a cursory academic glance, everything seems very simple: they measured the frequency response, created an equalizing filter, and the job was done. Many such attempts have been made, but no result has been achieved. Of course, in the opinion of the authors of these attempts and their supportive marketing, there is a result. But the impassive world of professionals is of a different opinion.
The problem is that sound system evaluators accept and evaluate sound differently than human hearing. They “see” more “problems” than our auditory perception (no matter how paradoxical it may sound). These problems have their origins in the physical interference of sound waves at the location of the sound pressure measurement. But interference occurs only when, in the simplest case, two signals arrived - direct and reflected (steady-state case). But for a short moment there is only a direct signal and there is no interference. This short moment is enough for our hearing to make an assessment.
I will try to prove the temporal selectivity of hearing and its ability to ignore interference with two easy-to-repeat experiments. First experience. The test signal "chirp" (sinusoidal signal with a rapidly changing frequency), short, 150 - 300 ms, logarithmic, subjectively sounds completely different when played, starting from low frequencies to high and vice versa. When playing up, the signal appears dim, with highs lost. Playing down - it sounds beautiful, musical, with pronounced highs. And for a spectrum analyzer, both cases are the same and indistinguishable.
The second experience. Let's sit in front of a classic stereo system. Let's give a mono signal. If everything is in order in the system, we will hear a narrow imaginary sound source exactly in the middle between the speakers. Now we ourselves are moving from side to side. In this case, we will only hear that the imaginary source will slightly move in the same direction as us. Now let's put the microphone in our place. Let's listen to the signal from this microphone and move it. Hear a beautiful flanger effect created by a variable comb filter. Try it.
So. In my opinion (which I have been converting into real technology for almost ten years now), one should measure and evaluate the sound system in the same way as our hearing does. This turned out to be possible if, instead of trying to understand something from the results of measuring the sound pressure at one point, measure the frequency response of the radiated sound power of the loudspeaker. This is the basis of my work and decisions.
I would like to take the liberty of reconsidering the approach to undistorted broadcasting of a sound program. Here's a classic principle. In the room (studio, open area) a microphone is installed in front of the performer, which converts the sound pressure into a proportional electrical signal, regardless of frequency. Behind it is the transmission path (preamplifier, radio channel, time delay device, etc., etc.), ending with an amplifier and a loudspeaker in the listening room. The path must transmit the signal in the same way, regardless of the frequency, and the loudspeaker must proportionally convert the electrical signal into sound pressure. And again - regardless of the frequency. We have verified whether the loudspeaker meets this requirement in a damped chamber on its "acoustic axis" and now we are waiting for success. Often this expectation turns out to be in vain and naive.
The approach I am developing is different. In order to obtain an undistorted sound image, the loudspeaker at the listening position must emit the same or proportional in spectral composition and time characteristics sound power which the musician emits at the place of performance.
The correctness of this approach has already been repeatedly tested in practice and was demonstrated with great success at the AES exhibition in May 2007, when a recording of an accordion duo was played through a corrected path ending with the Radiotehnika S90 speakers well known to Russians, and was compared with a live performance of the same duo who agreed to participate in the experiment.
By the way: here's another episode from the life of the S90. A small company left over from the flagship of Soviet electroacoustics, the Riga Radio Plant, had the courage to take part in the test of the leading Russian audio magazine with its budget-class loudspeakers. The results were impressive, without a single reproach about the sound and with the comment "I can't understand why it sounds good", even though the frequency response curves did not indicate this in any way. The answer is simple: we used the AJFL program and measurement technique to tune this loudspeaker.
The accuracy of the method makes it possible to use it in studios with the highest quality monitors, at the same time, the possibilities of the correction depth are so great that even a bucket will sound. We also put on such an experience ...
How is the emitted acoustic power correction method implemented in practice? The acoustic pressure is measured at many (about 200) points in space located on an imaginary surface or its segment. Simply put: the meter draws an imaginary array of vertical lines with a microphone in the air, it takes about a minute. A specially developed program independently fixes the sound pressure value at individual points, and then calculates the frequency response of the acoustic power (AJFL), where the factors of interference and phase shifts are taken into account. Based on this characteristic, a correction curve is synthesized. It is created as a mirror image in relation to the frequency response curve of the radiated power, while it is possible to follow this curve with an accuracy inaccessible to traditional equalizers. The fact is that a finite impulse response filter - FIR is used as an equalizer in the AJFL technology. For radio engineering, it is not new, but it has been used extremely rarely in sound equipment until now. One might even say that it was not used at all (I know of only one device with a FIR filter, its creators themselves do not really know how to work with it). This happens for three reasons: high requirements for computing power, an insignificant practical benefit from the obtained accuracy and the complexity of control, hence the return to understandable and familiar parametric and graphic equalizers.
And one more thing: phase correction. In AJFL technology, it happens automatically. The fact is that if the problem (unevenness) was caused by the minimum phase system (and such is the majority electrical circuits and filters with one signal path from input to output), then, having created a minimum phase corrector, the problem is corrected ideally - both in amplitude and in phase. The corrective filter-equalizer used in the AJFL system is just such a minimum phase one.
In 2010, there was also a solution for the car. Here it was necessary to slightly modify both the measurement technique and the instrument block responsible for the subsequent correction. Taking into account that the acoustics is more complicated than in an ordinary room, the frequency response of the emitted power in the cabin is recorded in several stages and in three (not two) coordinates. The measurement results are interpreted by a special version of the software on a laptop and loaded into a block that remains on board between the signal source and the amplifiers. In the course of measurement and adjustment (this is important), it is possible, in addition to automatic correction according to the "mirror" curve, to make manual adjustments, for this a subsystem of a high-precision parametric equalizer is provided.
The dimensions of the block with analog and digital inputs / outputs are 18 x 15 x 5 cm, the supply voltage is from 7 to 16 V. There is a Remote input and a delayed Remote output for controlling the activation of amplifiers. Now a simplified modification of the device is in operation, half the size and only with analog inputs / outputs. And in a couple of months "fast" loading of filters via USB-interface will be ready. So I think we still have a reason to meet here. If you don’t want to wait, it’s not difficult to find me, the address is in this issue of the magazine.
According to the AJFL method, measurements are taken not at one, but at many points that form a surface segment
Demonstration of the method at the AES in Vienna in 2007
Based on the frequency response of the radiated power synthesized from a variety of point measurements, the program builds a "mirror" correction curve
The result of the correction: the frequency step in units of hertz is not available for traditional equalizers
One of the most difficult cases (in the car). The result is similar
The first car model of the correction unit
The links of broadcast channels introduce amplitude-frequency distortion. This means that their gain or attenuation is a function of frequency and the frequency response of the gain differs from the horizontal line.
In many broadcasting devices, the magnitude of the amplitude-frequency distortion, which manifests itself as a decrease in the transmission coefficient at extreme frequencies, is reduced to a normalized value by a rational construction electrical circuit, the choice of the values of its elements and the mode of operation, the use of negative feedback... But the amplitude-frequency characteristics of some links of the broadcast channel, connecting lines, sound recording and reproducing devices, long-distance lines, wire broadcasting lines do not have a horizontal section. In these cases, the amplitude-frequency distortion is reduced by including a special circuit in the broadcast channel - correction contour QC.
Correction principles
Frequency response QC should be such that the overall frequency response of the distorting link and. QC in a given frequency band from fmax before fmin was a horizontal line. So, the condition for frequency correction of the distorting link:
where and - respectively, the attenuation (transmission) coefficient of the distorting link and the correcting circuit.
The methods of frequency predistortion are close to the methods of correcting amplitude-frequency distortions according to technical techniques and methods of calculation. Frequency predistortion refers to the artificial distortion of the broadcast signal spectrum in order to improve SNR. Frequency predistortion is widely used in broadcasting channels such as connecting lines, sound recorders, and frequency modulation broadcasting.
Since the SLs are included in the broadcast channel in various arbitrary combinations, they are considered as independent channel links. It is undesirable to compensate for the amplitude-frequency distortions introduced by the SL in other links of the channel - LU or PU, since in that case it is impossible to maneuver the amplifiers and SLs and connect any SL to any amplifier. Each trunk must be adjusted independently of the other links in the channel. Identity of the frequency response of the corrected trunk lines facilitates their operation and mutual redundancy. The frequency response of the corrected SL should fit within the template:
In SLs, fundamentally different methods of correcting the frequency response are used than in wire broadcasting lines. Due to the large number of SLs that are sequentially connected to the broadcast channel, a high correction accuracy is required (see Table 1).
The connecting lines are loaded on an active resistance, the value of which is commensurate with the modulus of the SL wave impedance. Under these conditions, the SL attenuation monotonically increases with frequency. Physically, this phenomenon can be explained using an equivalent circuit.
It is valid if the line length does not exceed a quarter of the wavelength of the transmitted signal, i.e. with an electrically short line. The resistance of the line wires together with the resistance formed by the resistances of active and capacitive leakages between the line wires and the load resistance form a voltage divider. As the frequency increases, the modulus increases and the modulus decreases. Therefore, the transmission coefficient of this circuit decreases with increasing frequency, and the attenuation increases.
Additional amplitude-frequency distortion occurs due to a change in the input impedance of the connecting line over the frequency range. Since the SL is the LO load, changes in the input impedance of the SL lead to a change in the voltage drop across the internal resistance of the broadcast signal source - the LO. But with a small value of the internal resistance of the LA, these distortions are insignificant, and they are not taken into account.
To correct the frequency response of the SL, a special four-terminal device with lumped parameters is used - a correcting circuit (CC). Its attenuation in the operating frequency range should be changed so that the total attenuation of the SL and KK does not depend on the frequency. The assumption that the total attenuation of the SL and KK is equal to the sum of the attenuations and is valid only if the input resistance of the KK is constant in the operating frequency range and is equal to the load resistance. Otherwise, when the CC is connected to the SL, the load of the SL will change and its attenuation will change.
The greatest attenuation of the KK should be introduced at the lowest operating frequency. Up to frequencies of 500-700 Hz, the attenuation should remain approximately constant, and then smoothly drop to zero at the highest operating frequency. The physical properties of SL and KK are different; line - a four-terminal network with distributed parameters, KK, - a four-terminal network with lumped parameters. Therefore, it is impossible to achieve full compensation of amplitude-frequency distortions introduced by SLs with the help of QC.
The more points are taken on the frequency axis, for which the CC attenuation should coincide with the attenuation obtained from the idealized curve, the more complex the CC circuit is.
The QC should have a minimum number of customizable (selectable) elements. At the highest frequency, the CC attenuation should approach zero. Switching on the CC should not change the frequency response of the attenuation of the link associated with it, in this case, the SL, otherwise the frequency correction will turn into a complex and laborious process of empirical selection of CC elements. When you turn on the CC at the end of the SL, you should use the CC with a constant input resistance, and when you turn it on at the beginning of the SL, with a minimum output resistance. A decrease in the output resistance of the CC is also desirable when the CC is turned on at the end of the SL, since this reduces the voltages of external noise induced on the input circuit of the amplifier following the CC. The constancy of the input impedance is also useful in those cases when the CC is switched on before the SL, as this stabilizes the LU mode.
Therefore, the CC should have constant input impedance, minimum output impedance, minimum attenuation at the highest operating frequency, and the smallest number of tunable elements.
Basic QC schemes:
The simplest two-terminal network connected in series with the load or parallel to the load does not give a good correction, since the input impedance of such a CC depends on the frequency and changes the course of the frequency response of the SL.
A full parallel circuit has a constant input impedance and a large output impedance that changes with frequency. A full series circuit has a constant input impedance and a small output impedance, also varying with frequency. For this reason, a full serial loop is most suitable for correcting trunk lines. The T-shaped bridge circuit provides a constant input impedance, but its output impedance is greater than that of full series. Therefore, it is less suitable for correcting LEDs, although it is found quite often in typical equipment.
The degree of complexity of two-terminal networks, and depends on the required correction accuracy. If two-terminal networks and with contain two elements each, moreover, formed parallel connection active resistance and capacitance, by a series connection of active resistance and inductance, then the calculated attenuation characteristic will coincide with the idealized one at two points - at (practically, in the region of lower frequencies) and at. If, - three-element, then the match is obtained in three points. With an increase in the requirements for the accuracy of the frequency response correction, one CC is not enough. Then two or more CCs are used, and additional CCs serve to correct the frequency response unevenness remaining after the introduction of the first CC.
The complication of QC is undesirable for economic reasons. Therefore, they are usually limited by the condition of the coincidence of the idealized and the calculated decay curve of the CC at three points, which are taken as, and one intermediate one. The calculation formulas are greatly simplified if the frequency at which the attenuation of the KK is equal to half the maximum is taken as an intermediate point.
Two-terminal circuits and synthesized based on the following considerations.
In the region of the lowest frequencies of resistance and must be purely active. At the highest calculated frequency, it should vanish and approach infinity. This can be achieved by performing in the form of a serial, and in the form of a parallel oscillatory circuit. The resonant frequencies of the circuits must be equal and coincide with the highest frequency of the operating range. The attenuation of the CC in the low-frequency region is determined by the ratio and:
The steepness of the frequency response of the KK attenuation increases with an increase in the ratio; accordingly, the frequency of half attenuation increases. The losses in the oscillating circuits reduce the correction accuracy at higher frequencies. Therefore, the inductors should have as little active resistance as possible. Capacitors and must have low dielectric losses.
We learned how to calculate the acoustic design with a phase inverter and began to experimentally determine the dependence of the total electrical impedance of the dynamic heads on the frequency. Today we will try to comprehend the measurement results, after which we will consider the methods of amplitude and frequency correction of the emitters.
If you find impedance minima around 3 ohms, don't be discouraged. Some speaker models from well-known brands have dips up to 2.6 ohms, and sometimes even up to 2 ohms! There is, of course, nothing good in this - amplifiers overheat, working on such a load, especially at high volume, distortions grow.
For tube triode amplifiers, the lows in the low frequency and low mids are especially dangerous. If the impedance here drops below 3 ohms, the terminal lamps may fail, but the pentodes are not afraid of this.
It is important to remember that the output impedance of the amplifier is involved in setting the AC filter. For example, if you make a 1 dB rise in the Fc area by connecting the speaker to a transistor amplifier with almost zero output impedance, then when working with a tube amplifier (typical value Rout = 2 ohms), there will be no trace of the afterburner. And the whole frequency response will be different. To get the same results, you have to create another filter.
The ever-evolving listener will eventually come to appreciate the value of good tube amplifiers. For this reason, I usually set up acoustics with a tube end, and when connected to a transistor, I put a 10-watt non-inductive (no more than 4 - 8 mN) resistor with a resistance of 2 ohms in series with the speaker.
If, having a transistor amplifier, you do not exclude the possibility of acquiring a tube amplifier in the future, then when setting up and subsequent operation, connect your speakers through such resistors. When switching to lamps, you do not need to tune the speaker again, you just need to remove the resistors.
In the absence of a generator, a test CD with a recording of test signals for evaluating the frequency response is suitable. In this case, you will not be able to smoothly change the frequency and, most likely, miss the very minimum of impedance. Nevertheless, even an approximate estimate of the Z modulus will be useful, and for this, pseudo-noise signals in one-third octave bands are even more convenient than sinusoidal ones. Such signals are on the test CD of the AV Salon magazine (№7 / 2002). As a last resort, impedance measurements can be dispensed with by limiting the afterburner recoil at the filter cutoff frequency to 1 dB. In this case, the impedance is unlikely to drop by more than 20%. For example, for a 4 ohm speaker, this corresponds to a minimum of 3.2 ohms, which is acceptable.
Please note that you will have to "catch" the parameters of the filter elements required to correct the frequency response yourself. A preliminary calculation is needed in order not to overshoot by a kilometer from the very beginning. Resistors are added to a simple low / mid filter head for some frequency response manipulation that may be required when tuning your speakers. If the average sound pressure level of this speaker is higher than the corresponding parameter of the tweeter head, a resistor must be connected in series with the speaker.
Switching options are shown in Fig. 6 a) and b).
The value of the required decrease in the recoil of the bass / midrange head, expressed in dB, is denoted by N. Then:
where Rd is the average value of the speaker impedance.
Instead of calculations, you can use table 1.
Table 1
1 dB - = 10%, or a change in the level by a factor of 1.1.
2 dB - = 25% - "- 1.25 times.
3 dB - = 40% - "- 1.4 times.
4 dB - = 60% - "- 1.6 times.
5 dB - = 80% - "- 1.8 times.
6 dB - = 100% - "- 2 times.
where Vs is the effective value of the voltage at the amplifier output. Vd - the same, on the dynamics. Vd is less than Vc due to signal attenuation by resistor R1. In addition, N = Nvch - Nnch, where Nnch and Nvch are the sound pressure level developed by the LF and HF heads, respectively.
These levels are averaged over the bands reproduced by the LF and HF drivers. Naturally, Nf and Nf are measured in dB.
An example of a quick estimate of the required R1 value:
For N = 1 dB; R1 = Rd (1.1 - 1) = 0.1 Rd.
For N = 2 dB; R1 = Rd (1.25 - 1) = 0.25 Rd.
For N = 6 dB; R1 = Rd (2 - 1) = Rd.
A more specific example:
Rd = 8 Ohm, N = 4 dB.
R1 = 8 ohms (1.6 - 1) = 4.8 ohms.
Let Рд be the rated power of the LF / MF loudspeaker, PR1 - the allowable power dissipated by R1.
You should not make it difficult to remove heat from R1, that is, you do not need to wrap it with electrical tape, fill it with hot glue, etc.
Peculiarities of preliminary calculation of the filter with R1.
For the circuit in Fig. 6 b) the values of L1 and C1 are calculated for an imaginary speaker, the total resistance of which is: RS = R1 + Rd.
In this case, L1 turns out to be more, and C1 - less than for a filter without R1.
For the circuit in Fig. 6 a) - the opposite is true: the introduction of R1 into the circuit requires a decrease in L1 and an increase in C1. It is easier to calculate the filter according to the diagram in Fig. 6 b). Use this particular scheme.
Additional correction of the frequency response using a resistor.
If, to improve the frequency response uniformity, it is necessary to reduce the filter suppression of signals above the cutoff frequency, you can apply the circuit shown in Fig. 7
R2 in this case gives a decrease in recoil in Fc. Above Fc, the recoil, on the contrary, grows in comparison with the filter without R2. If you need to restore close to the original frequency response (measured without R2), you should reduce L1 and increase C1 in the same proportion. In practice, the R2 range is within:
R2 = (0.1E1) і Rd.
Frequency response correction
The simplest case. On a fairly uniform characteristic, there is a zone of overestimated recoil ("presence") in the midrange. You can apply a corrector in the form of a resonant circuit (Fig. 8).
At resonance frequency 
The circuit has a certain impedance value, in accordance with the value of which the signal on the speaker is attenuated.
Outside the resonance frequency, attenuation is reduced, so the circuit can selectively suppress "presence".
It is convenient to use table 1a:
| Rev. level in dB | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 |
| Relates. rev. level (D) | 1,1 | 1,25 | 1,4 | 1,6 | 1,8 | 2 | 2,2 | 2,5 | 2,8 | 3,16 | 3,55 | 4 |
Example: it is necessary to suppress "presence" with a center frequency of 1600 Hz. The speaker impedance is 8 ohms. Suppression level: 4 dB.
The specific shape of the frequency response of the loudspeaker may require more complex correction.
Examples are shown in Fig. nine.
The case in Fig. 9 a) is the simplest. It is easy to select the parameters of the correcting contour, since the "present" has a "mirror-like" shape of the possible filter characteristic.
In fig. 9 b) another possible option is shown. It can be seen that the simplest contour allows you to "exchange" one large "hump" into two small ones with a small dip in the frequency response in addition.
In such cases, you must first increase L2 and decrease C2. This will broaden the suppression band to the desired level. Then you should bypass the circuit with a resistor R3, as shown in Fig. 10. The value of R3 is selected based on the required degree of suppression of the signal supplied to the speaker in the band determined by the parameters of the circuit.
Fig. 10
R3 = Rd (D - 1)
Example: you need to suppress the signal by 2 dB. Speaker - 8 ohms. Refer to Table 1.
R3 = 8 ohms (1.25 - 1) = 2 ohms.
How the correction occurs in this case is shown in Fig. 9 c).
Modern loudspeakers are characterized by a combination of two problems: "presence" in the 1000 - 2000 Hz region and some excess of the upper mids. A possible form of the frequency response is shown in Fig. 11 a).
The method of correction that is most free from harmful "side" effects requires a slight complication of the contour.
The corrector is shown in Fig. 12
The resonance of the L2, C2 circuit is needed, as usual, to suppress the "presence". Below Fp, the signal passes through L2 to the speaker with almost no loss. Above Fp, the signal goes through C2 and is attenuated by resistor R4.
The corrector is optimized in several stages. Since the introduction of R4 weakens the resonance of the L2, C2 circuit, initially you should choose L2 more and C2 less. This will provide excessive suppression on Fp, which will normalize after R4 is injected.
R3 = Rd (D - 1), where D is the amount of suppression of signals above Fp.
D is selected in accordance with the excess of the upper middle, referring to table 1.
The stages of correction are conventionally illustrated in Fig. 11 b).
In rare cases, a reverse effect on the frequency response slope is required using a correction circuit. It is clear that for this R4 must move to L2.
The diagram is shown in Fig. 13.
The problematic frequency response and its correction for this case are shown in Fig. fourteen.
With a certain combination of L2, C2 and R4 values, the corrector may not have much suppression at Fp.
An example of when just such a correction is needed is shown in Fig. 15.
(To be continued)
Room correction has long been used by sound engineers to provide sound customization in studios and concert halls, with the goal of reducing the impact of the particular room on the sound. This is especially important for recording studios, which should not add anything extra to the sound being recorded. In studios, acoustic processing is primarily used, but in halls, multi-band graphic equalizers or digital parametric equalizers are used for this. Sound reproduction problems in different parts of the room are detected by ear or using a measurement microphone, after which correction is performed using the existing equalizers.
Ideally, we should get a similar sound environment at home when listening to various recordings, which could provide the sound that the sound engineer intended. The realities of our life are such that few can afford a complete acoustic decoration of rooms in an apartment or house, and therefore someone invites installers of sound equipment, and someone is trying to set everything up on their own. Unfortunately, this operation requires certain theoretical knowledge, experience, and appropriate equipment. Therefore, for home needs more and more automatic calibration systems began to be used, which repeat the actions of a sound engineer or installer, but only do it automatically, using an external microphone.
In most articles describing various home equipment, the capabilities of calibration systems are practically not considered. And the question of comparing different calibration systems, as far as the author knows, has not been considered by anyone at all. It seems that this is a kind of taboo among the observers. Well, let's change this world a little. In this article, we will try to correct this omission and compare the most common correction systems. At the same time, we will not compare the differences in the sound of the used receivers and their loading capacity, we will not measure distortion, as well as explore additional functionality, even if it concerns sound - this is a completely separate topic.
We will also slightly touch on the topic of manual equalization, which can be useful for owners of receivers who do not like them. automatic system and for owners of amplifiers.
A bit of theory
With the development of digital technologies in audio equipment of the middle and even budget classes, we were able to carry out preliminary correction of the reproduced sound at the selected listening point through acoustic systems (AC), taking into account the peculiarities of a particular room, its size, the arrangement of speakers and surrounding objects (furniture, curtains, carpets etc.).
The task of such systems is, at a minimum, to correct the amplitude-frequency characteristic (AFC) at the listening point separately for each speaker, as well as the total one while simultaneously sounding several speakers in a multichannel home theater system. In addition to the frequency response, in order to match the simultaneous operation of several speakers, it is also required to synchronize the phase-frequency response (PFC), and to improve the perception of music, it is also necessary to provide a minimum group delay time (CLT).
The task becomes more complicated if speakers of different manufacturers or model lines are used, the speakers themselves have an uneven frequency response, and the room does not have minimal acoustic treatment to reduce the influence of multiple reflections. It adds complexity if the speaker and the listening point are chosen incorrectly: some of the frequencies are amplified at the listening point, which causes bubbling or an unpleasant color of timbres, and some of the frequencies are mutually subtracted, forming a dip in the frequency response, which leads to depletion of musical timbres and, again, additional color sounding.
The principle of operation of such correction systems is to make changes to the original signal at the stage of processing by a digital sound processor (DSP) in such a way that at the listening point you get the most even playback parameters, devoid of the influence of the room and the features of specific speakers in a home theater.
It should never be forgotten that with the help of preliminary sound correction alone it is impossible to solve all the problems of sound reproduction through the speakers in the living room, and it is very desirable to initially perform a number of measures for the correct placement of the speakers, the choice of the listening point, minimal processing of the room to eliminate unwanted reflections of signals from walls, floor and ceiling. And only after that, when everything possible for the current room has been done, you can start correcting the signal as the final stage of adjusting the sound of the audio system in the room in use.
Automatic correction
In modern receivers, sufficiently efficient processors are installed inside, which, by measuring the response of each speaker at the listening point, can automatically adjust the sound correction both in amplitude and in phase, which should lead to minimal deviations in the frequency response graph and in-phase operation of different speakers.
The advertising materials of all systems describe that this system, using an extension microphone, analyzes all the parameters of sound reproduction in a particular room and makes all the necessary corrections to ensure the best sound. That is, at first glance, all systems are quite equivalent, and when choosing an audio / video receiver, we no longer need to pay attention to which calibration system is installed, but more attention should be paid to the number of channels, amplifier power, connectivity of mobile devices, etc. The only visible difference between systems of different levels is the presence or absence of correction of the subwoofer channel (as a rule, in cheap receivers, the subwoofer channel cannot be corrected for frequency response).
In practice, it turns out that different systems have completely different effects on sound optimization, and the end result is quite different when using one system or another. Even one manufacturer has several classes of such systems that have different sound editing capabilities.
Meanwhile, everything similar systems provide a minimum comfortable level: determine the number of connected speakers, the distance to them and the gain level for each speaker. The parameters can be set manually, but with the help of a microphone everything is done more accurately and faster.
Testing task
The objective of the review is a practical test of various automatic sound calibration systems for the characteristics of the room in the same conditions: on the same speakers, in the same living room with speakers from different manufacturers at the fronts, center and rear seats. Of course, it is better to have speakers of one company and one series, but often, due to various reasons, there are situations when the system has been assembled for some time and it turns out to be somewhat inconsistent.
Testing was carried out both with the "out of the box" setting, and with manual change of various parameters to achieve optimal frequency response.
What should an ideal calibration system provide? Uniform frequency response at the listening point during sound reproduction both through any one speaker, and through any number of simultaneously operating speakers. For surround sound, it is important that when the musical image is moved in the created sound space, its tonality does not change, which can only be achieved by a good coincidence of the frequency response and phase response at the listening point.
What can prevent you from getting a flat frequency response at the listening point?
- The formation of standing waves due to multiple reflections of sound waves from the walls. At the listening point, both the amplification of the original amplitude (the antinode of the standing wave) and the weakening of the amplitude (the node of the standing wave) can occur.
- SBIR effect is the interaction between direct sound from the speaker and reflections from the near edges of the room.
- Attenuation of the signal amplitude due to damping of waves from different speakers, the phases of which differ at the listening point.
- The initial frequency response curve of the speakers themselves.
Test room and instrumentation
An example of a test room for listening to music. Not the real test room described in this article!
The test room is a living room measuring 5.8 x 3.1 x 2.7 m (L x W x H) with one entrance door and one door to the balcony. The front speaker is located on a short wall, 70 cm from the window. The listening position is on a leather sofa, 2 meters from the front system and 3 meters from the rear wall. There is a medium-pile carpet on the floor between the sofa and the front speaker, and a thick curtain on the window. The subwoofer is located between the left speaker and the B-pillar.
The following acoustic systems are installed in the room:
All full-range speakers are two-way bass reflex, the subwoofer is equipped with one 15 "speaker with electromagnetic feedback in a closed box.
The following equipment is used for measurement:
- Measuring microphone Behringer ECM 8000
- XLR-to-USB interface Shure X2U with phantom power for microphone
- Sound level meter CEM DT-815
- ASUS N46vz laptop with built-in sound card(Realtek HD Audio)
- Behringer U-Control UCA202 USB Audio Interface (PCM2902E Chip)
- REW software v5.1 beta 17
- A set of necessary wires, a stand of the "crane" type, a photo tripod
It should be noted that the measuring equipment and software are not professional, therefore, the graphs obtained during the test are recommended to be compared not with others, but only with each other within this test. However, as far as the author knows, this microphone and the REW program are widespread among lovers of good sound, and therefore, under certain restrictions, the results can be compared.
Testing technique
To take measurements of the sound of various speakers at the listening point, a Behringer ECM 8000 measuring microphone is used, which is connected via a Shure X2U adapter to the USB port of an ASUS N46vz laptop. The laptop is running REW v5.1 software. The output from the built-in sound card is connected to one of the analog inputs of the receiver under test.
The REW software generates a test sweep tone that is played by the selected speaker through the receiver. Sound waves are captured by a measuring microphone, the data from which are processed in the REW software, resulting in the formation of frequency response, phase response, group delay deviations, etc.
To understand what exactly was being edited, measurements were also taken from the exits. preamplifier of each channel, so that you can study in detail in which ranges and to what level changes were made in the input signal without the participation of the AU. A Behringer U-Control UCA202 USB audio interface was also used to read data from the preamplifier.
The sound interfaces were preliminarily calibrated according to the frequency response and phase response using a "loop", that is, by giving a signal to the input from the output of the sound card itself. A calibration file downloaded from the Internet for this microphone was used as a calibration file for the measuring microphone.
First, measurements were taken of each speaker at the listening point, as well as combinations of different speakers when working together. For each combination, two measurements were made: with the calibration system turned off, i.e. the initial characteristic, and also with the built-in calibration system sound processing turned on, which allows you to see the result of the correction at the listening point. Each system was calibrated several times to obtain the best listening position.
After completing the obligatory metering program, manual edits were made to achieve the best results, as well as auditioning of audio materials and certain episodes from popular films with big amount volumetric effects.
Initial measurements
To understand which speakers I had to work with, below are the AFCs of all speakers separately, taken in the near field (for pairs, only one of the speakers was measured), that is, when the microphone is located at a distance of about 20 cm from the center of the tweeter. Unfortunately, even with such a measurement, it was not possible to avoid the influence of the characteristics of the room, but it is minimal.
| Vandersteen Model 1C |  |
| The 58 Hz hump is related to the influence of the room on the measurement results. The frequency response is fairly uniform. | |
| KEF Cresta |  |
| Raising up to almost 1 kHz strongly distinguishes this speaker from the background of others, which presents an additional problem for calibration systems in straightening the frequency response on par with the front speakers. | |
| AAD C-100 |  |
| Rythmik F15 |  |
| All speakers on one graph |  |
The simulation of the influence of the room on the final frequency response when connecting the front speakers and a subwoofer is quite close to reality, which was obtained as a result of subsequent measurements:
Calculated frequency response graph at the listening point in the REW room simulator
Calibration Systems under Test
- Audyssey 2EQ (briefly based on old measurements on Onkyo TX-N717 receiver)
- Manual subwoofer channel editing with Behringer FBQ2496
- Manual straightening of front speakers with the Behringer FBQ2496
Audyssey
Audyssey branded systems are divided into several classes that differ in functionality and accuracy.
| Possibilities | MultEQ XT32 | MultEQ XT | MultEQ | 2EQ |
| Filter Resolution | 512x | 16x | 2x | 1x |
| Filter Resolution (Subwoofer) | 512x | 128x | 128x | No |
| Number of measurement positions | 8 | 8 | 6 | 3 |
| Adaptive low frequency correction | there is | there is | there is | No |
| Crossover, polarity, delays, levels | there is | there is | there is | there is |
The operation of the frequency response and phase response correction system is based on a complex FIR (FIR) filter, which allows you to accurately correct the initial frequency response over a set of points. Different classes of Audyssey have different filter resolutions, that is, they provide different sound editing fidelity.
After calibration, the user has access to the following modes, which differ in the final curve:
- Flat (Music for Onkyo) - Smoothest frequency response for near field listening.
- Reference (Movie for Onkyo) - an optimized frequency response for natural sound (according to the system designers) with a drop in the HF region and a small drop at a frequency of 2 kHz. High frequency roll-off is recommended for THX requirements to achieve a home theater sound similar to that of a large theater.
- Bypassing fronts (not used on Onkyo) is a mode for those who have invested a lot of money in front acoustics and room acoustics (or who just like the way fronts sound without additional processing). In this case, the correction of the remaining speakers is tied to the frequency response of the front speakers.
The Audyssey Calibration System has the ability to manual change the calculated filters by correcting the frequency response. Either the user selects one of the modes offered by the machine, or refuses and can use a separate graphic equalizer, which cannot be turned on while one of the Audyssey modes is used. The only way the user can influence the resulting frequency response (at least in the Onkyo implementation) is to slightly change the tone to taste by adjusting the low and high frequencies.
All other parameters, such as levels, delays, crossover settings, etc., can be adjusted manually.
Audyssey 2EQ
Audyssey 2EQ is a basic calibration system, and in addition to typical crossover and delay functions, it performs speaker correction with filters with a basic resolution in the mid / high frequency region without support for subwoofer channel correction.
An example of how the Audyssey 2EQ system works on adjusting the frequency response (output from the Onkyo 717 pre-amplifier)
Changes in the frequency response are carried out only in the region of medium and high frequencies, starting from 1 kHz, which allows us to solve only one problem - equalizing the characteristics of various speakers at these frequencies.
In fact, there is no room correction in this basic system, the effect of which manifests itself mainly in the low frequency region. Correction is only possible manually using the built-in graphic equalizer, which is activated only when turned off in the menu using the settings for frequency response editing by Audyssey. And if you are lucky, and the only hump in your conditions is exactly at 63 Hz, then it will be effectively eliminated manually. In other cases, the wide influence and fixed frequency of the graphic equalizer will not allow you to remove the humps on the frequency response without affecting the neighboring areas. For the subwoofer channel, an equalizer with a lower frequency grid can be used, but again the frequencies are fixed and may not coincide with the problematic ones in your room.
When using a receiver with a 2EQ system, it is possible to recommend purchasing a subwoofer with its own calibration system or using an additional device for frequency response correction between the receiver and the subwoofer, which can be either automatic or manual (parametric equalizer). In this case, at least the lowest frequencies will be reproduced correctly, and everything above the frequency of the subwoofer will be reproduced "as is."
Audyssey MultiEQ XT32 in Onkyo TX-NR818 Receiver
Equalizing main channels
The MultiEQ XT32 is Audyssey's oldest system. It is usually equipped with top-end line of devices, but sometimes XT32 can be found in the middle class of receivers.
Frequency response graph of the left channel before and after calibration in Movie and Music modes:
Red curve - left channel frequency response in Pure Direct mode, blue - in "Cinema" mode, green - in "Music" mode
The frequency response graphs show the excellent work of the XT32 both in flattening the frequency response in the area of humps, and in pulling out dips (as much as possible).
Modes "Cinema" and "Music" differ only at a frequency of about 2 kHz and in the HF region after 6 kHz to the end of the range. For clarity of the differences between the edits, let's see the frequency response of the left channel, taken from the output of the preamplifier:
Red curve - frequency response of the left channel from the preamplifier in Pure Direct mode, blue - in "Cinema" mode, green - in "Music" mode
In the rest of the range, the differences between "Kino" and "Music" are so small that they can be neglected. In the future, all frequency response will be displayed only for the "Cinema" mode.
Frequency response graph of the right channel before and after calibration in the "Movie" mode:
Red curve - frequency response of the right channel in Pure Direct mode, blue - in Cinema mode, green - in Music mode; top graphs from the preamplifier, bottom graphs from the microphone
Here we see an aggressive editing of the frequency response in the low-frequency region and a more relaxed editing in the mid-range and high-frequency range.
Also, one of the features of the XT32 is a manic desire to correct the frequency response even where the speaker practically stops playing. In this case frequency range front speakers start at 38 Hz, but at the expense of the room they still play from 30 Hz. But Audyssey amplifies the signal down to 10 Hz (here only in the left channel), which can overload the speaker and amplifier at high volume levels when a separate subwoofer is not used.
Let's go back to the frequency response of the left and right speakers before calibration and see how the frequency response differs in the problem area of the bass, where the room affects the most:
The red curve is the frequency response of the left channel in Pure Direct mode, the blue curve is the right channel
On the graph, we see that speakers with the same frequency response in the near field are quite different in frequency response at the listening point, since they are located in different places in the room and have a different pattern of reflections due to the lack of complete symmetry in the arrangement.
But after running Audyssey XT32, the difference in frequency response is sharply reduced:
The red curve is the frequency response of the left channel in the "Cinema" mode, the blue one - of the right channel in the "Cinema" mode
Now let's look at a rather complex center channel, which has a strong frequency response unevenness:
Red curve - frequency response of the center channel in Pure Direct mode, blue - in "Cinema" mode
As you can see, Audyssey has perfectly straightened the frequency response in the low frequency range. But in this case, it is not the frequency response itself that is more important, but consistency with the front speakers, so that the center channel does not stand out much with its sound. To do this, let's look at the frequency response graphs of all three front speakers - left, right and center:
Red curve - frequency response of the left channel in the "Cinema" mode, blue - right channel, green - center channel
And again I would like to applaud the Audyssey XT32 system for the work done to correct the characteristics of completely different speakers. In practice, while listening to the “Cinema” mode, the center channel is really so harmoniously combined with the front speakers that sometimes it seems that all the acoustics are from one manufacturer.
In order to understand how a system sounds without calibration, again, just look at the frequency response graphs of three systems in Pure Direct mode:
Red curve - frequency response of the left channel in Pure Direct mode, blue - right channel, green - center channel
Equalizing the subwoofer
Now let's move on to the frequency response of the subwoofer channel and see what the XT32 has to offer us here:
Red curve - frequency response of the subwoofer channel in Pure Direct mode, blue - in Cinema mode
We can see that the frequency response of the subwoofer has been adjusted as well as possible with the current placement.
On the frequency response from the preamplifier outputs, you can consider the correction curve of the subwoofer channel:
Purple curve - frequency response of the subwoofer channel from the pre-amplifier output in Pure Direct mode, blue - in "Cinema" mode
Here again the tendency of Audyssey to equalize the frequency response at any frequency is manifested: even at frequencies above 400 Hz, the system tries to draw out sound that the subwoofer no longer reproduces at all. It's good that this happens when the crossover is running, so it does not cause negative consequences. On the other side of the frequency response on this subwoofer, everything is quite normal, since it is physically capable of reproducing frequencies from 10 Hz. But with another subwoofer, which plays, say, from 30 Hz, problems may arise due to the overestimation of the signal level at the lowest frequencies, below 30 Hz, where the subwoofer no longer reproduces anything. And if an infra-low-frequency filter is not built into it, then the amplifier can idle amplify the signal that the speaker is not able to reproduce. This should be taken into account when playing music or movies at high volume levels.
Rear channel equalization
As for the speakers of the rear channels, everything there is also well straightened, which can be clearly seen on the frequency response graphs of these speakers in the "Cinema" mode:
The red curve is the frequency response of the left rear channel in the "Cinema" mode, blue - the right rear channel
Outcome
We have made sure that Audyssey corrects the frequency response of each speaker well enough, which means that the signal from each individual speaker will be as reliable as possible to the listening point.
But what happens if the same signal is played on both channels - for example, a voice or any other mono signal?
To do this, let's take a look at the complex frequency response of the front speakers together with the subbuffer:
Red curve - frequency response of triphonic in Pure Direct mode, blue - in "Cinema" mode
When the two front speakers worked together with a subwoofer after Audyssey correction in Cinema mode, many problems were fixed, although the difference is not as impressive as comparing the changes in each channel separately.
What is very much lacking in the Onkyo receiver is the ability to save several correction settings for different situations: since the correction in Audyssey is quite detailed, when the situation changes, the current edit becomes not entirely relevant. For example, one could use several settings for the following situations:
- listening with one subwoofer for one person (music for yourself)
- listening with two subwoofers (or a second subwoofer) for one person with an open screen for the projector (cinema for oneself)
- listening with two subwoofers for several people on the couch (cinema for a family)
- test cases for choosing the best position of the measuring points
There is also a lack of visualization of the changes made. In YPAO, we can look at the parametric equalizer settings by copying the settings to manual mode. In MCACC, the adjustments made are visible in the graphic equalizer setup menu. And only in the case of Onkyo receivers, the user is deprived of any opportunity to visually evaluate the changes made, without the help of an external microphone, the assessment of changes is only possible by ear. But this is not a feature of Audyssey, but its implementation by Onkyo. In modern Denon receivers, you can view the correction curve for each channel and evaluate its changes with different measurements.
Advanced MCACC in Pioneer SC-LX56 receiver
Pioneer uses a proprietary multi-channel acoustic calibration system called MCACC in its receivers. In addition to the standard functions for detecting connected speakers, distance to them and gain level, MCACC offers frequency response correction, as well as promises correction for reverberation and group delay time.
After performing Auto Calibration, all settings except phase control are available for manual override. For greater convenience, there are 6 memory locations for storing different settings for different situations.
The complete calibration process takes quite a long time, during which a variety of test signals are reproduced at different stages in a circle. The main measurement is performed at one point of the microphone position, after its completion, the result is automatically recorded in memory, and the main menu appears on the screen. A video of the full calibration process is available.
After completing the calibration, the “Symmetry” option was saved in the M1 memory cell, and the “All ch adj” option was saved in the M2 cell. In the future, the designations M1 and M2 will be used on the graphs, which corresponds to these options.
Equalizing main channels
The mechanisms for adjusting the frequency response are:
- 9-band graphic equalizer for all channels except subwoofer, with frequencies 63, 125, 250, 1k, 2k, 4k, 8k, 16k (Hz).
- Parametric equalizer for correcting standing waves (minus only) with a center frequency of 63 Hz. Used for subwoofer, center channel and all other speakers.
For each pair of speakers, one of the size options is configured - Large or Small. When set to Large, the entire frequency range is supplied to the speaker, when set to Small, only frequencies above the cutoff frequency are supplied, and everything below goes to the subwoofer. It is inconvenient that the cutoff frequency for Small speakers is set to one at all from 50 to 200 Hz, while up to 80 Hz there are only two values: 50 and 80 Hz, which slightly limits the precise setting of the crossover for the used speakers.
An interesting feature is the ability to correct the slope of the target curve in the HF range starting from 2 kHz. In the X-Curve setting, you can select the level of slope decay in dB per octave.
Red curve - left channel frequency response in Pure Direct mode, blue - MCACC M1, green - MCACC M2
Left channel graphic equalizer settings:
Setting up standing wave correction for front speakers:
The change in frequency response on the graph coincides with the settings of the graphic equalizer for the left channel and the standing wave filter. The work of the graphic equalizer is clearly visible along the wide adjustment of the frequency response, in which individual humps and dips are not straightened, but a general correction of the curve is in progress.
Also, for calibration systems, the center channel is complex, which stands out strongly in its frequency response in the low and middle frequencies, and without editing it differs in sound from the front speaker pair.
Frequency response of the center channel before and after calibration:
Unfortunately, the miracle did not happen: all the unevenness was practically preserved in both modes of the MCACC system. This example shows that it is better not to use completely different speakers together with MCACC, since it will not be possible to equalize their frequency response.
Group delay time equalization
One of the main features of MCACC is the fight against reverberations and equalization of the group delay time. There are many beautiful videos on the company's website about this, which show what kind of trouble exists before calibration and that after calibration comes sound nirvana.
Well, now for the most exciting moment - let's see the excess group delay time without correction and after calibration.
Black trace - group delay in Pure Direct mode, red - after MCACC calibration, green - after Audyssey calibration
The miracle did not happen here either: the schedule of excess group delay time remained practically unchanged.
For comparison, this is the same graph of the left channel from the Onkyo 818 receiver in Music mode. It can be seen that in some places the delay is even less, despite the lack of such advertising statements from Onkyo.
It may seem that the measurements were made incorrect, but the receiver itself provides us with graphs before and after calibration, which show that only all the graphs have shifted upward, and in frequency they have remained practically unchanged.
In all other measurements, both of individual speakers and their simultaneous operation, the difference in the correction of the group delay time is also not visible, although when measuring in the M1 and M2 modes, the "Fullband Phase Ctrl" phase control was always selected.
Equalizing the subwoofer
Frequency response graph of the subwoofer channel before and after calibration:
Red curve - frequency response of the subwoofer before calibration, blue - MCACC M1, green - MCACC M2
The chart shows that almost no correction was made. This is not surprising, since there is simply no separate graphic equalizer for the subwoofer (in the same Onkyo 818, in addition to Audyssey, there is a manual graphic equalizer for the subwoofer with a corresponding set of frequencies), and the discrepancies in the graphs are only due to the operation of the standing wave filter on frequency 63 Hz.
Since the subwoofer has a fairly flat frequency response of its own, and its location is optimal, adjusting the cutoff at a frequency of 50 or 80 Hz, you can get a completely flat frequency response of the subwoofer. With a different subwoofer or a different cut, things can be more dire.
Outcome
Frequency response graph in triphonic mode with a cutoff frequency to the subwoofer 80 Hz:
Left and right channel in triphonic mode
With the simultaneous operation of the front speakers and the subwoofer, the resulting frequency response is quite predictable and does not give any surprises in the form of dips and humps that appear.
In general, according to the MCACC system, we can say that it is quite standard and allows you to automatically adjust all the basic parameters, as well as a 9-band graphic equalizer and a 3-band parametric standing wave equalizer.
The approach to the implementation of the MCACC system is quite interesting and has many possibilities for manual editing, analysis of the result, the ability to save settings in several memory cells, but everything is limited by two very serious drawbacks:
- A normal graphic equalizer with fixed center frequencies is used. For a modern calibration system, a more accurate mechanism is needed, and if instead of a graphic equalizer there was a parametric one, then there would be much more tuning possibilities.
- Limiting all settings to 63 Hz from below. It is not clear why, but there is no possibility to adjust anything below 63 Hz, although the main problems of room acoustics lie in this range. Actually, due to this limitation, there is absolutely no equalizer for the subwoofer.
YPAO RSC in the Yamaha RX-A2010 receiver
Yamaha's proprietary YPAO calibration system provides both general functions for setting levels, delay, crossover adjustment, and channel-by-channel frequency response adjustment.
For each channel for manual editing, there are 7 parametric equalizer filters (except for the subwoofer channel, where there are only 4 filters). For each filter, you can set the center frequency, correction level, filter quality factor.
The center frequency of the filter is selected from a list of 28 fixed frequencies: 31.3; 39.4; 49.6; 62.5; 78.7; 99.2; 125.0; 157.5; 198.4; 250.0; 315.0; 396.9; 500.0; 630.0; 793.7; 1.00k; 1.26k; 1.59k; 2.00k; 2.52k; 3.17k; 4.00k; 5.04k; 6.35k; 8.00k; 10.1k; 12.7k; 16.0k (Hz). For the subwoofer channel, only the first 10 frequencies up to and including 250 Hz are used.
The quality factor (Q) is set from 0.5 to 10.08: 0.5; 0.63; 0.794; 1; 1.26; 1.587; 2; 2.52; 3.175; 4; 5,040; 6.35; eight; 10.08.
After automatic calibration, the user is offered three options for equalizer correction:
- Average - the smoothest frequency response
- On the front - all speakers are pulled up to the frequency response of the front speakers
- Natural - optimized frequency response for natural sound (according to the system designers)
During the described testing, as well as testing YPAO in the Yamaha 1071 receiver, it was revealed that there are currently two different calibration systems:
- YPAO RSC (Reflected Sound Control)
Both systems are very similar in appearance and functionality, with one exception: the YPAO system uses only a 7-band equalizer for each channel (4-band per subwoofer) as a frequency response correction mechanism, and in addition to this, the YPAO RSC system uses a more complex filter. for the front speakers and the center channel - presumably a FIR (FIR) filter.
After performing automatic calibration in the YPAO RSC system, a complex filter for adjusting the frequency response is calculated (we will call it simply RSC), and on top of it, using the existing parametric equalizer, adjustments are made to obtain several equalizer options ("Averaged", "On the front", "Natural" ).
When copying the settings of one of the modes into the manual equalizer, we get at the output exactly the same editing as when the corresponding "automatic" mode is working. However, when you reset the manual equalizer, the graph at the preamplifier output is not linear, but contains edits to the RSC filter, which cannot be turned off.
YPAO uses only parametric EQ for all channels, and when you set it to zero, the output is flat, just like YPAO RSC for the surround and subwoofer channels.
Equalizing main channels
Frequency response graph of the left channel before and after calibration:
Red curve - frequency response graph without correction, blue - in "Natural" mode, green - in "Average" mode
The adjustments in the low-pass area are made by the RSC filter, and in the high-pass area there is a difference due to different settings of the parametric equalizer.
Let's look at the graphs from the left channel preamplifier:
Black curve - frequency response graph without correction, red - in "Natural" mode, blue - in "Edge" mode, green - in "Averaged" mode
And here is what the user sees in the parametric equalizer of the left channel by copying the "Natural" setting:
And by copying the "Average" setting:
In the parametric equalizer, only the general edit for LF and HF is automatically adjusted, and the two modes "Natural" and "Averaged" differ only in the edit at HF: for "Averaged", the correction is plus +2.5 dB at a frequency of 12.7 kHz, for " Natural ”correction –0.5 dB at 5 kHz and –1.5 dB at 16 kHz.
In Edge mode, only the RSC filter remains, and the parametric equalizer is reset. If the user copies any equalizer mode, and then resets it, then the RSC curve "On the front" will be at the output of the preamplifier.
Unfortunately, we have not found a way to turn off the RSC filter, so that in fact, only the parametric equalizer is edited. But in practice this is not necessary, since the RSC filter quite correctly corrects the humps on the frequency response and it can be supplemented with manual settings of the parametric equalizer.
Edits in the region up to about 500 Hz have a maximum amplitude of up to 6-7 dB, after which the amplitude gradually decreases down to 3-4 kHz. Editing in the high-frequency region is left to the mercy of the parametric equalizer, which each user can change to suit his preferences.
An unpleasant surprise was the rise in the lowest frequency F3 (cutoff frequency at the level of −3 dB), where the front speakers practically do not play anymore, but the RSC filter tries to stretch the frequency response using maximum amplification up to several hertz. The same can be seen in the Audyssey XT32, which we cannot edit. YPAO has a parametric equalizer on top of the automatic filter, but, unfortunately, they failed to correct this range, since its minimum frequency is only 31.3 Hz. You need to take this into account when setting up home acoustics or connecting a subwoofer - then the graph begins to fall below the cutoff frequency:
Black curve - frequency response graph without correction, red - in the "On the front" mode, blue - an attempt to edit the equalizer at 31.3 Hz, green - when the crossover for the subwoofer is turned on at 80 Hz
But this feature manifested itself only on the front speakers, for the speakers there is no boost at the lowest frequencies for the central channel.
Frequency response graph of the right channel before and after calibration (for clarity, both the frequency response from the microphone and the preamplifier):
Red curve - graph of frequency response of the right channel before correction, green - in "Average" mode, blue - in "Natural" mode
Now let's move on to the central speaker, which is quite complex due to its peculiar frequency response:
Red curve - graph of frequency response of the central channel before correction, green - in "Average" mode, blue - in "Natural" mode
Unfortunately, all the features of the frequency response remained on the graph even after all the filters were working, i.e. YPAO failed to align the frequency response curve and bring it closer to the front speakers.
But we have a parametric equalizer in stock, with which you can try to correct the frequency response manually. Using the example of the central channel, we will evaluate the capabilities of the parametric equalizer in editing the frequency response in the LF / MF region. Several edits, and in the end:
The red curve is the graph of the frequency response of the central channel before correction, the blue curve is in the "On the front" mode with manual adjustment of the equalizer
After manual editing, the frequency response graph leveled off quite well, which is also noticeable by ear: the center is now not so much distinguished by its sound.
And if you superimpose the frequency response graph of the central speaker on the frequency response graph of the left speaker, you can see that the frequency response now does not differ so much:
Red curve - left channel frequency response graph, blue - center channel after manual correction
On the graph from the preamplifier, you can see the difference in the filters:
The red curve is the graph of the frequency response of the central channel from the output of the preamplifier before correction, the blue one is in the “Averaged” mode, the green one is in the “On the front” mode with manual adjustment of the equalizer in the band up to 2 kHz
Editing the parametric equalizer for the center channel on the screen looks like this:
Equalization of rear channels and subwoofer
Now let's move on to the rear channels. After looking at the frequency response graphs from the preamplifier, it becomes clear that only the usual parametric equalizer works on the rear without the complex RSC filter:
Red curve - frequency response graph in the "Natural" mode, blue - in the "On the front" mode, green - in the "Average" mode
Either Yamaha engineers decided that additional precision is not needed for the rear channels, or the processing power of the built-in DSP processor is still lacking.
We can find exactly the same graphs on all channels of the usual YPAO system (without the RSC prefix), where only the parametric equalizer is used as a frequency response adjustment tool (for example, in the Yamaha RX-V1071 receiver).
Unfortunately, not only the rear speakers, but also the subwoofer channel were deprived of the complex RSC filter:
Green curve - frequency response graph of the subwoofer channel from the pre-amplifier before correction with a 200 Hz crossover, blue - in the "Natural" mode
Accordingly, in automatic mode, the frequency response of the subwoofer practically does not change:
The red curve is the frequency response graph of the subwoofer channel before correction with a crossover of 200 Hz, blue - in the "Natural" mode, green - in the "Average" mode
Attempts to correct the frequency response of the subwoofer channel with a parametric equalizer did not give much result, since at the lower frequencies the step of the center frequency of the equalizer is quite large:
Black curve - frequency response graph of the subwoofer channel before correction with a 200 Hz crossover, red - after manual editing with a parametric equalizer
To assess the capabilities of the equalizer in the sub channel, a second subwoofer was connected and attempts were made to correct the frequency response with manual settings. But even in this case, due to the limited set of center frequencies of the equalizer, a significant change was also not achieved, therefore, to completely edit the subwoofer channel, it is better to use a subwoofer with built-in calibration or a separate external device(below will be described the use of an external parametric equalizer).
But for both subwoofers, setting the crossover to a frequency of less than 80 Hz avoids large fluctuations in the frequency response, which for many will be quite acceptable in terms of the result obtained.
Manual frequency response editing using Behringer FBQ2496
Since we had the appropriate equipment, it was decided to give an example of the possibility of manual sound correction using an inexpensive external parametric equalizer for comparison with automatic equalization systems.
Manual subwoofer channel editing
As such a device, a fairly popular digital parametric equalizer was chosen to adjust the frequency response of the subwoofer as part of the Behringer FBQ2496 feedback suppressor. The FBQ2496 has 20 filters per two channels. For each filter, the center frequency is set fairly accurately from 20 Hz to 20 kHz.
In the low-frequency range, the step is from a fraction of a hertz (at the beginning of the range) to several hertz: 20.00; 20.23; 20.46; 20.70; 20.94; 21.18 ... 60.49; 61.10; 61.80; 62.52; 63.25 ... 120.5; 121.9; 123.3; 124.7; 126.2 ... (Hz).
In the HF region, the step is already tens and hundreds of hertz: 5.024; 5.082; 5.141; 5,200; 5,260; 5.321 ... 19.099; 19,321; 19,544; 19,771; 20 (kHz).
To set up the subwoofer, the original frequency response was removed, the editing range was selected from 10 to 120 Hz, and the filters were automatically generated in the REW program, after which they were loaded into the equalizer via the MIDI interface.
Filter settings for adjusting the frequency response of the subwoofer
In addition to the automatically generated filters based on the measurement results, two more filters were added with the following parameters:
- Frequency 44.2 Hz, gain −2 dB, Q factor 0.5
- Frequency 153 Hz, gain -6.5 dB, Q-factor 0.16
The resulting frequency response curves when the subwoofer is operated with a cutoff at 200 Hz:
Green curve - before the equalizer, blue - the equalizer on 12 filters
In the range up to 67 Hz, the frequency response graph turns into an almost straight line, and further up to 120 Hz the deviations do not exceed 3 dB. In the future, it is better to set the crossover frequency to 60 or 80 Hz.
However, you need to understand that only the subwoofer channel was tuned, and when it works together with the front speakers, if the frequency response in the low-frequency region does not correct them, you will have to adjust the settings depending on the imposition of the signal from the subwoofer and front speakers in the region of the selected frequency section.
Manual straightening of front speakers
In the case when the system does not use a subwoofer and music is heard only through the front speakers connected to the stereo amplifier, it is possible to use the parametric equalizer to correct the sound.
The range up to 1 kHz was selected for the test. Measurements were made in the REW program of the initial frequency response of two speakers, automatic generation of filters along the target line at a level of 75 dB was performed, then the filters were loaded into the equalizer via the MIDI interface. No further edits were made to the equalizer settings. All 20 filters were used for the left channel, and only 17 for the right channel.
The frequency response graph from the outputs of the equalizer shows that the filters are formed of a rather complex shape, in places of narrow Q-factor, which required the use of a large number of filters for each channel.
Changing the frequency response for the left speaker:
Red curve - frequency response of the left speaker without equalizer operation, blue - with equalizer on, black - from the equalizer output
Changing the frequency response for the right speaker:
Green curve - frequency response of the right speaker without equalizer operation, blue - with equalizer on, black - from the equalizer output
Here we see that for each individual speaker, the frequency response graph in the region up to 1 kHz has become smoother, and only dips remain, which should not be pulled out due to a change in the signal amplitude.
Audyssey MultiEQ XT32 in ARC2 (Advanced Room Correction 2)
Previously, only "iron" solutions were considered, when the correction system was built into the receiver or an external parametric equalizer was used. But there are also software solutions, allowing you to correct the signal in accordance with the characteristics of the room acoustics.
The disadvantage of this method is the attachment to the computer as a signal source, as well as the processing of only 2 stereo channels. The advantage is the flexibility of settings and the ability to use it in conjunction with any integrated amplifier.
The ARC2 (Advanced Room Correction 2) system is built on the basis of the Audyssey MultiEQ XT32 solution and allows not only to take measurements at several points, but also to see the resulting frequency response for each channel, as well as to correct the target curve by choosing any presetting or manually editing it to your liking.
You can use the VST plug-in in any player that supports VST extensions, as well as to play any sounds in Windows, provided that several programs are installed. For this you will need:
- ASIO4All
- Virtual audio cable
- ASIO FX Processor
Having configured the output of all sound to a virtual cable, turn on the ARC2 VST plug-in in the ASIO FX Processor program and output the sound to the line-out of the sound card.
To measure frequency response using an external microphone, you need a sound card with ASIO support and a sampling rate of 48 kHz.
Frequency response of the left channel of the ARC2 system:
Red curve - frequency response of the left speaker without ARC2, blue - with ARC2 on, green - with ARC2 with "Full Range Bass Correction" enabled
The result of the ARC2 is similar to what we see after Audyssey XT32 in the Onkyo receiver. The difference is that we can tweak the target curve in real time and get the result right away.
You can use the "Full Range Bass Correction" option for equalization in the area of the lowest frequencies, select one of the predefined curves, and edit up to 4 custom curves. In our case, when using a measurement microphone with IK000008 calibration instead of IK000002, we had to change the curve in the HF region:
After correction in both channels at the output, we get two flat frequency response:
Green curve - AFC of the left speaker with ARC2 on, blue - right speaker with ARC2 on
If we compare the frequency response graphs at the preamplifier output from the Onkyo receiver when the Audyssey setting is on “Cinema” mode and from the sound card output when the ARC2 is in operation, we can see that they are almost completely identical and differ only by a slight offset of the microphone when measuring:
The red curve is the frequency response of the left speaker when the ARC2 is on, the blue curve is the left speaker when the Audyssey XT32 is working in the Onkyo receiver
The ARC2 software package can be viewed as a kind of specific equalizer, where the user does not set the editing curve, but immediately the required frequency response curve at the listening point, and the system generates the necessary filter to provide the specified value according to the preliminary measurement data from the microphone at the required listening area.
Final comparison of "iron" equalization systems
I would like to immediately warn you that the receivers were tested at different times, so you need to compare the graphs of different receivers with the understanding that the measuring microphone could be slightly shifted (although it was always set strictly along the line and then it was adjusted by test measurements in Pure Direct mode to match the frequency response with previous measurements). The position of the measurement microphone has a greater impact on the mid and high frequencies, where every 5 mm of displacement can significantly change the picture. In the low-frequency region, such movements are practically imperceptible, and only movements of a few centimeters are anything but critical.
To demonstrate the differences, we present the left channel frequency response graphs for each receiver without the use of correction by correction systems:
Blue curve - YPAO left speaker frequency response without correction, green - MCACC systems, red - Audyssey systems
As you can see, in the low-frequency range, the differences are minimal, and in the rest of the range, too, are insignificant, therefore, taking into account the fact of these small differences, let's start comparing the frequency response from different systems.
Unfortunately, none of the tested receivers in any form shows the original graphs of the measured frequency response (at least in a simplified form with 1/6 octave smoothing) for a visual assessment of problem areas and the possibility to partially solve them by searching for the optimal location of the speakers and listening places. All the necessary data is present after the measurement, and the processors used and the quality of the image output to the TV allow you to display the frequency response graph, but for some reason none of the manufacturers does this.
Consider editing the left channel MCACC and YPAO:
Blue curve - the frequency response of the left speaker of the YPAO system at the "Average" setting, the green curve - the MCACC system at the M2 setting
In the low frequency area, everything is very similar, since the editing of both systems is minimal, but the YPAO graph looks a little better due to stretching out some of the dips. At frequencies below 40 Hz, the YPAO system tries to stretch the frequency response due to additional boost, which looks quite nice, and at a low volume it is even pleasant to the ear, but it is not recommended to play music at high volume with such a correction due to the possible overload of the amplifying part and distortion from the woofer.
Audyssey & YPAO Left Channel Frequency Response Graph:
Blue curve - YPAO left speaker frequency response with “Average” setting, red - Audyssey system with “Cinema” setting
Audyssey's bass correction is more accurate and the frequency response graph is more linear by cutting peaks and pulling out dips. Just like YPAO, Audyssey tries to pull the frequency response below 40Hz by amplifying the signal. At frequencies around 6 kHz, Audyssey has a boost that sounds more "open" to the ear. The rest of the charts are very similar.
Let's move on to the central channel, as the most interesting for analyzing the operation of the frequency response correction system (due to the initially large non-uniformity of the frequency response of this channel in the system under test):
Red curve - frequency response of the center channel of Audyssey system, green - YPAO systems in Natural mode, blue - MCACC systems in M2 mode
The graphs after the MCACC and YPAO systems have a fairly large unevenness in the frequency range from 100 Hz to 700 Hz, which is perceived by ear as a coloration of the sound relative to the front speakers. The graph after Audyssey is the smoothest and, as we discussed in the description of Audyssey MultiEQ XT32, practically coincides with the frequency response of the front channels.
However, for YPAO, manual correction was performed using a parametric equalizer, and now their difference with Audyssey is quite insignificant and appears only in the range from 100 to 180 Hz:
The red curve is the frequency response of the center channel of the Audyssey system, the green curve is the YPAO systems with manual adjustment of the equalizer
Next, let's compare several simultaneously sounding speakers in order to assess how correct the editing turned out to be for reproducing a signal not from one speaker, but from several at once - this is any monophonic signal, a voice or an instrument located in the center.
Frequency response in triphonic mode (fronts + sub with cutoff at 80 Hz) of MCACC and YPAO systems:
The red curve is the frequency response in the triphonic mode of the YPAO system at the “Average” setting, the green curve is the MCACC system at the M2 setting
The frequency response in the triphonic mode after correction by the MCACC and YPAO systems is very similar, especially in the low-frequency range, where both systems practically do not control the subwoofer channel and then together repeat all the humps and dips. The treble boost in YPAO can be easily changed with the parametric equalizer.
Trifonik (fronts + sub with cutoff at 80 Hz) Audyssey and MCACC:
Green curve - frequency response in triphonic mode of MCACC system with M2 setup, blue - Audyssey XT32 system with "Cinema" setup
The frequency response correction by the Audyssey XT32 system is very noticeable in the low-frequency range, where there is practically a “shelf” in the subwoofer channel, and then all the humps are cut off and some dips are stretched out.
Trifonik (fronts + sub with cutoff at 80 Hz) Audyssey and YPAO:
Red curve - frequency response in triphonic mode of YPAO system with “Average” setting, blue - Audyssey XT32 system with “Cinema” setting
Again, we see excellent Audyssey performance in the subwoofer channel and in the rest of the low-frequency range.
A difficult test is the simultaneous reproduction of the signal by all speakers - front, center, rear and subwoofer. In this case, all parameters are important: editing the frequency response of each channel, correctly setting the distance to the speakers, gain levels for each signal, phase coincidence. When playing a test signal simultaneously in all speakers, the difference in the final frequency response turned out to be quite decent:
The red curve is the frequency response of the simultaneous operation of all 5.1 channels of the YPAO system at the "Average" setting, the blue curve is for the Audyssey XT32 systems at the "Cinema" setting, the green curve is for the MCACC systems at the M2 setting
The graphs of the MCACC and YPAO systems practically coincide in the frequency range from 100 Hz to 800 Hz, then up to 3 kHz in YPAO there is a slight dip - apparently, due to the fact that the rear channels are corrected only minimally. In the area of subwoofer operation, the difference is about 7 dB, which is still difficult to explain. Perhaps the difference is due to measurement errors, or some channels were set for the MCACC system to Large (without a cut to the subwoofer), or, perhaps, the systems work out differently the decomposition of a stereo signal into 5 channels simultaneously.
The frequency response graph of the Audyssey system is distinguished by a flat "shelf" in the range of the subwoofer, but then there is a drop of about 7 dB and then a more or less direct frequency response with dips at 197 and 356 Hz, but without a significant rise at 165 Hz, as in others systems, which is most likely associated with the features of the central channel. The blockage in the 2 kHz region is a feature of the "Cinema" mode and is practically absent in the "Music" mode.
Outcomes
- Audyssey MultiEQ XT 32 for the smoothest frequency response curves of all channels, including the subwoofer
- YPAO RSC for the good work of the sophisticated RSC filter in correcting low-pass problems
- MCACC for visibility of changes made
- YPAO for full range performance
- Audyssey 2EQ for correcting the frequency response of dissimilar speakers in the high frequency range
- YPAO (all) for flexible parametric equalizer per channel
- MCACC for 9-band graphic EQ and 3-band standing wave EQ
- Audyssey (all) for graphic equalizer including subwoofer channel EQ (Onkyo implementation)
- Audyssey MultiEQ XT 32 and YPAO RSC. It is difficult to unambiguously choose a leader, since one system perfectly straightens the frequency response in the entire range, and the second, although it straightens the frequency response worse, has the ability to additionally edit the result obtained using a parametric equalizer for personal preferences.
- MCACC. A good set of features is limited only by the editing tools used.
- YPAO. Automatic tuning only slightly corrects the frequency response of the channels, which requires a mandatory change in the parametric equalizer settings to obtain an acceptable result.
- Audyssey 2EQ. Failure to edit below 1 kHz does not allow room influence to be corrected.
When using a computer as a source and listening only to stereo recordings, the best option would be to use Audyssey MultiEQ XT32 in the ARC2 software, since this solution combines two features at once: excellent operation of the machine and the ability to manually edit.
Audyssey 2EQ
Pros: basic system of basic parameters calibration.
Minuses: the absence of any correction in the region below 1 kHz, which does not allow correcting problems related to the characteristics of the room.
Audyssey MultiEQ XT32 (in receiver)
Pros: the most powerful system for equalizing the frequency response of all channels in the full range (both for the characteristics of the room and for dissimilar speakers, including rear and subwoofer), simplicity for the end user.
Minuses: the impossibility of editing the result of the correction, there is no way to set the parameters before starting the measurements, there is no way to save several results of the correction, stretching the frequency response outside the range of the speaker.
Audyssey MultiEQ XT32 (included with ARC2 program)
Pros: the most powerful system for equalizing the frequency response of all channels in the full range for the characteristics of the room, the ability to manually edit the target curve.
Minuses: requires a computer as a source, processing only stereo output, the complexity of setting up a pass-through path to output all sounds from the computer.
Advanced MCACC
Pros: the ability to edit the automatic equalizer, several memory cells for different settings and measurement results, a visual representation of information about the changes made, the accuracy of setting the center frequency of the parametric equalizer of the standing wave filter (starting from 63 Hz).
Minuses: no equalizer for the subwoofer, setting of the standing wave filter only from 63 Hz, the worst result of correcting the frequency response in the low frequency area, one subwoofer crossover frequency for all channels.
YPAO (normal)
Pros: the ability to edit the result by adjusting the parametric equalizer.
Minuses: impossibility to fine-tune the frequency response of the subwoofer, some user skill is required to fine-tune the frequency response using manual adjustment of the equalizer, a large step of the center frequencies of the parametric equalizer and a maximum of 7 bands per channel.
YPAO RSC
Pros: combination of a complex RSC filter to correct problems in the area of low and midrange with the ability to edit the result by adjusting the parametric equalizer.
Minuses: inability to fine-tune the frequency response of the subwoofer, non-switchable editing of the RSC filter in the manual equalizer mode, no RSC filter for the rear channels and subwoofer, some user skill is required to fine-tune the frequency response using manual equalizer adjustment, a large step of the center frequencies of the parametric equalizer and a maximum of 7 bands for each channel.
3.2. High-frequency and low-frequency correction of the frequency response of the resistor amplifier
To correct the frequency response of a real amplifier in order to bring it closer to the frequency response of an ideal amplifier (see Fig. 3.1), special correction schemes are used in the LF and HF regions.
The RF circuit - correction of the frequency response using the correcting inductance Lk is shown in Fig. 3.8.
The principle of operation of this circuit is based on an increase in the HF resistance of the collector circuit (Rк + jwLк). An increase in this resistance with increasing w makes it possible to increase the gain of the RF stage. A necessary condition for the efficiency of this circuit is the high resistance of the external load resistance Rн> Rк. Otherwise, a small resistance Rн will shunt the collector circuit, while the amplification of the stage will be determined by the value of Rн and depend little on Rк and Lк. The equivalent circuit of a casacada with HF-correction at 1 / Yi> Rn> Rk is shown in Fig. 3.9, from which it follows that the HF frequency response of the corrected amplifier is close to the frequency response of a parallel oscillatory circuit.
Consequently, with a non-optimal choice of the parameters of the correcting inductance Lk, a rise may appear on the frequency response of the amplifier, causing distortion of the amplified signals. AFC and RH of an amplifier with HF-correction with optimal and non-optimal parameters of the correcting inductance Lk are shown in Fig. 3.10.
1. Lk< Lопт 2.Lк = Lопт 3.Lк >Lopt
It can be seen that HF correction affects only the HF region (the region of small times - pulse fronts). When Lc> Lopt, the rise time is the smallest, however, an overshoot occurs at the output pulse signal.
The LF correction circuit for the amplifier frequency response is shown in Fig. 3.11, where Rf and Cf are the LF correction elements, which simultaneously perform the role of the LF filter in the power supply circuit of the transistor VT1.
The principle of operation of the LF correction circuit is based on an increase in the resistance of the collector circuit in the LF region, therefore, as in the inductive HF correction circuit, this scheme effective only at high-resistance load Rн> Rк. The capacitance of the capacitor Ср is chosen so that at medium and high frequencies 1 / wСф<< Rф (то есть Сф шунтирует Rф), поэтому цепь Сф, Rф практически не оказывает влияния на работу усилителя на СЧ и ВЧ. На НЧ сопротивление Сф становится больше сопротивления Rф, это увеличивает сопротивление коллекторной цепи и как результат - понижает нижнюю граничную частоту полосы пропускания усилителя. При этом отношение Rф/Rк определяет максимально возможный подъем усиления с понижением частоты w, который однако, реально всегда бывает меньше по причине снижения усиления на НЧ из-за разделительного конденсатора Ср.
Frequency response and RH of the amplifier with optimal and non-optimal parameters of LF correction (1 - no correction, 2 - optimal correction, 3 - overcorrection) are shown in Fig. 3.12.
4. DESCRIPTION OF THE LABORATORY INSTALLATION.
The laboratory setup includes:
1) laboratory model;
2) laboratory power supply;
3) a universal voltmeter (type B7-15, B7-16).
4) generator of low-frequency signals (type G3-56, GZ-102).
Lab layout contains:
a) the investigated AC resistor amplifier with an emitter follower at the output to ensure a high-resistance load of the amplifier (see Fig. 4.1.).
b) built-in pulse signal generator (with the ability to adjust the amplitude and duration of pulses), located on the upper part of the laboratory model body.
The laboratory model is powered from a constant voltage source En = + 12V. The external view of the front panel with the schematic diagram of the laboratory layout printed on it is shown in Fig. 4.2.
5. ORDER OF WORK
5.1. Investigation of the influence of the blocking capacitor on the characteristics of the amplifier.
a) Assemble the installation according to the diagram in Fig. 5.1. Put all switches in their original 1 position.
Set the value of Uout within 10 ... 30 mV to ensure the linear operation of the amplifier. Investigating the dependence of Uout on the frequency f of the input signal (with a constant value of Uin), obtain and construct the frequency response of the amplifier at 2 values of the capacitance Cp (switch S4). When examining the frequency response, it is recommended to first estimate the frequency domain of uniform amplification, where the number of samples can be reduced to 3 ... 4. In the frequency regions of the frequency response (LF and HF), the number of reference points should be increased to 4 ... 5.
b) Connect a pulse signal from a square-wave generator to the input of the amplifier under study (see Section 4). Monitor the output voltage of the amplifier using an oscilloscope. Draw from the oscilloscope screen on one graph the shape of the pulses at the output of the amplifier (PX amplifier) for two values of Cp.
Measure the falloff of the flat part of the pulse top (in%) for two values of Cp.
Draw conclusions about the influence of the blocking capacitor Cp on the characteristics of the amplifier.
5.2. Study of the influence of collector resistance on the characteristics of the amplifier.
Using the scheme and methods of clause 5.1. measure the nominal gain Ko, remove the frequency response and HR of the amplifier for 2 values of Rk. Construct the frequency response and HR of the amplifier for two values of Rk.
Draw conclusions about the influence of the collector resistance on the characteristics of the amplifier.
5.3. Investigation of the effect of LF correction.
Switch S4 to the position corresponding to the lower value of Cp. Investigate the frequency response and HR of the amplifier for 3 values of the LF correction parameters. Construct the frequency response and frequency response of the amplifier for various parameters of the low-frequency correction.
Draw conclusions about the influence of Rf, Cf on the characteristics of the amplifier.
5.4. Investigating the impact of RF correction
Switch S1 to position Rк max, and switch S5 to position 1.
Investigate the frequency response and HR of the amplifier for 3 values of the correcting inductance Lk. Construct the frequency response and frequency response of the amplifier for various parameters of the inductive high-frequency correction.
Draw conclusions about the influence of Lk on the characteristics of the amplifier.
5.5. Preparation of a report on laboratory work.
The report should contain:
a) AC resistor amplifier circuit with LF and HF correction;
b) measurement results, tables and graphs required by laboratory tasks;
c) conclusion on the correspondence of the obtained results to theoretical data.
6. CONTROL QUESTIONS
1. Elements of temperature stabilization of the operating point of the transistor and their choice.
2. The work of the resistor casacada in the low-frequency range.
3. The work of the resistor casacada in the HF region.
4. Influence of the expanding capacitor Cp on the characteristics of the amplifier.
5. Influence of collector resistance Rk on the upper cutoff frequency and nominal gain.
6. The principle of operation of inductive HF - correction of the resistor amplifier.
7. Amplifier frequency response with optimal and non-optimal parameters of the HF - correction elements.
8. HF amplifier with optimal and non-optimal parameters of the HF - correction elements.
9. The principle of operation of the LF - correction of the resistor amplifier.
10. Amplifier frequency response with optimal and non-optimal parameters of the LF elements - correction.
11. HF amplifier with optimal and non-optimal parameters of the elements of the LF - correction.
7. LI T E R A T U R A.
1. Ostapenko GS Amplifying devices. - M.: Radio and communication, 1989, subsections 1.4, 1.5, 3.2, 4.8.
2. Voishvillo GV Amplifying devices. - M.: Radio and communication, 1983, subsections 4.1.1, 4.7.3, 5.3.1, 5.3.3.
3. Mamonkin I. G. Amplifying devices. - M.: Communication, 1977, subsections 6.3, 7.3, 11.3.
