It is hardly an exaggeration to say that every radio amateur has a tester of the M-83x family. Simple, accessible, cheap. Quite sufficient for an electrician.
But for the radio amateur it has a flaw when measuring alternating voltage. Firstly, low sensitivity, and secondly, it is intended for measuring voltages with a frequency of 50 Hz. Often a novice amateur does not have other instruments, but wants to measure, for example, the voltage at the output of a power amplifier and evaluate its frequency response. Is it possible to do this?
On the Internet, everyone repeats the same thing - “no higher than 400 Hz.” Is it so? Let's get a look.
For testing, a setup was assembled from an M-832 tester, a GZ-102 sound generator and
lamp voltmeter V3-38.
Judging by the available data, numerous devices of the M-83x or D-83x family are assembled according to almost the same scheme, so there is a high probability that the measurement results will be close. In addition, in this case, I was of little interest in the absolute error of this tester; I was only interested in its readings depending on the signal frequency.
The level was selected around 8 Volts. This is close to the maximum output voltage of the GZ-102 generator and close to the voltage at the output of an average power UMZCH.
It would be better to make another series of measurements with a powerful ULF loaded onto a step-up transformer, but I don’t think that the results will change dramatically.
For convenience of estimating the frequency response in dB, a level of 0 dB was selected at the 10 V limit of the V3-38 voltmeter. When the signal frequency changed, the level was slightly adjusted, but the changes did not exceed fractions of dB and can be ignored.
results
In the table below TO- coefficient by which the measurement result of the tester at a given frequency must be multiplied, taking into account the decline in the frequency response.
To obtain tabulated results in dB, the voltage level obtained for each frequency was set at the output of the generator, and the difference in dB was read and entered into the table. Some inaccuracies due to 0.5 dB rounding of tube voltmeter readings and rounding of the last digit of tester readings. I think that in this case a systematic error of 1 dB is quite acceptable because it is imperceptible to the ear.
Conclusion
So what happened?The frequency response of the tester is correct not up to 400 Hz, but up to 4...6 kHz; above that the decline begins, which can be taken into account using the table and, therefore, obtain relatively reliable results in the range of 20...20000 Hz and even higher.
In order to assert that the amendments are suitable for all testers, you need to collect statistics. Unfortunately, I don’t have a bag of testers.
We must not forget that the tester measures alternating voltage using a half-wave rectifier circuit with its disadvantages, such as the ability to measure only sinusoidal voltage without a direct component; at a low measured voltage, the error will increase.
How can I improve the M-832 tester for measuring alternating voltages?
You can install an additional “200-20 V” limit switch and another shunt resistor. But this requires disassembling and modifying the tester; you need to understand the circuit and have a calibration device. I think this is inappropriate.Better make a separate attachment that amplifies and rectifies the voltage. The rectified voltage is supplied to the tester, which is turned on to measure DC voltage.
But this is a topic for another article.
In descriptions of low-frequency amplifiers published in Radio magazine and other radio engineering literature, it is customary to indicate their rated power, nonlinear distortion factor, sensitivity and frequency response. Based on these four main parameters, one can already judge the quality of the amplifier and its suitability for certain purposes.
What are these parameters of a low-frequency amplifier? Rated output power (P nom) is the power, expressed in watts or milliwatts, delivered by the amplifier to the load, at which nonlinear distortions correspond to those specified in the description. With further increase in power, distortion increases significantly. The power at which distortion reaches 10% is considered to be maximum (P max).
Nonlinear distortions. In the process of amplifying any signal, even a purely sinusoidal one, due to the nonlinearity of the characteristics of transistors, vacuum tubes, transformers and a number of other equipment elements, harmonics appear in the amplified signal - oscillations whose frequencies are 2, 3 or more times higher than the fundamental frequency. This is nonlinear or harmonic distortion, which increases as the power supplied by the amplifier to the load increases. They are rated by harmonic distortion factor.
Harmonic distortion coefficient (K g), measured with a sinusoidal input signal of constant frequency, is expressed as a percentage of the total voltage of all harmonics U g to the output voltage U out
The permissible Kg is established by the relevant standards (GOST). For example, for amplifiers of low-frequency radios, radios, tape recorders, and electric phones, it can be 5-7%, for household tape recorders - 5%. The higher the class of radio equipment, the lower its Kg should be.
Sensitivity. The term sensitivity usually means the voltage of the low-frequency signal in millivolts that must be applied to the input of the amplifier in order to obtain the rated output power at the load. The sensitivity of most record playback amplifiers is 100-200 mV, and the sensitivity of household tape recorder recording amplifiers, measured from the microphone input, is 1-2 mV.
The frequency response of an amplifier is the dependence of the output signal voltage on frequency at a constant input voltage Uin. For a number of reasons, low-frequency amplifiers amplify signals of different frequencies differently. Usually the lowest (f n) and the highest (f b) are the worst amplified, so the frequency characteristics of the amplifiers are uneven and have dips or rollovers at the edges. The extreme frequencies at which there is a decrease in gain by 30% (-3dB) are considered the boundaries of the amplified frequency band; they are indicated in the amplifier’s passport data. The frequency response or amplified frequency band of low-frequency amplifiers of network radio receivers can be from 100 to 10,000 Hz, and of amplifiers of small-sized transistor receivers - from 200 to 3500 Hz. The higher the class of the amplifier, the wider the amplified frequency band.
In addition to these parameters, there are some others, but they are secondary or arising from the main ones.
But a radio amateur installed, tested and adjusted the amplifier. How to measure its main parameters in order to compare them with the given ones?
Measurements of the parameters of LF amplifiers are usually carried out using special high-precision measuring equipment. However, in amateur conditions this can be done using simple measuring instruments, for example, described in our magazine in 1971 and 1972. under the heading Radio Amateur Laboratory. You will need a low-frequency generator, a transistor AC millivoltmeter and rectifiers to separately power these devices. The amplifier being tested usually has its own power supply. You also need a load equivalent R e - a wire resistor whose resistance is equal to the resistance of the voice coil of the loudspeaker used in the amplifier, or a special device described in the article Universal load equivalent, published in Radio No. 12, 1973.
The set of instruments of the Amateur Radio Laboratory does not include a nonlinear distortion meter (NID), so measurements of this parameter of the amplifier will have to be carried out using a simplified method, additionally using any low-frequency electronic oscilloscope, for example LO-70. In this case, measurements begin with taking the amplitude characteristics of the amplifier - the dependence of the output voltage U out of the amplified signal on the input voltage U in, measured at a frequency of 1000 Hz (1 kHz) with a constant load R n = R e.
So, let's start reading the amplitude characteristics of the amplifier. The connection diagram of the measuring instruments with the amplifier being tested is shown in Fig. 1, a (power circuits not shown). A signal with a frequency of 1000 Hz from the output of the low-frequency generator (LF) is supplied to the input of the low-frequency amplifier (ULF) using a shielded two-core cable. We ground the cable braid and one of its cores at the amplifier input. We connect the millivoltmeter to the generator output control sockets. We smoothly increase the amplitude of the generator signal to a voltage of 0.3 V. In this case, the actual signal voltage at the amplifier input will be 30 mV, since it is removed from the generator attenuator, which attenuates the signal 10 times (1: 10). Having measured the input voltage Uin, switch the millivoltmeter to the measurement limit of 10 V and measure the output voltage Uout at the load equivalent R e (Fig. 1, b). Suppose the voltage U out is 1.2 V. We draw up a table (Table 1) and write down the measurement results in it: U in = 30 mV, U out = 1.2 V. Next, we increase the input voltage in steps of 10 mV, and write down the measurement results to the table. And so on until the proportionality of the increase in output voltage U out is violated. In this case, the cutting off of the tops of the sinusoid, noticeable to the eye, should be observed on the oscilloscope screen (Fig. 1, c). Clipping occurs due to symmetrical limitation of the amplitude of the output signal and is accompanied by an increase in distortion to approximately 10%. This means the amplifier has reached maximum power (P max). Then we slightly reduce Uin until the sine wave distortion disappears (see Fig. 1, b) and consider that the amplifier now delivers the rated power P nom. The output voltages corresponding to P max and P nom, for example 4.1 and 3.6 V, are highlighted in the table.
Now, using the data in table. 1, we build the amplitude characteristic of the amplifier (Fig. 2). To do this, along the horizontal axis to the right of zero we mark the input voltages Uin in millivolts, and along the vertical axis upward - the output voltages Uout in volts. We mark all measured values of Uout with crosses on the graph and draw a smooth line through them. This line is straight up to point a, and then noticeably deviates downward, which indicates a violation of the direct relationship U out / U in and a sharp increase in distortion.
Knowing the voltage U out and the load equivalent resistance R e, you can calculate the output power P out of the amplifier for various voltages U out.
The output power Pout is calculated using the formula resulting from Ohm’s law:
For example, with P n = 6.5 Ohm and Uout = 1.0 V
at Uout corresponding to 1.8 V, Pout ≈ 0.5 W, etc. In Fig. 2, parallel to the U out axis, a second vertical axis is drawn, on which the calculated output powers P out are marked.
The inflection of the amplitude characteristic usually corresponds to the rated power P nom of the amplifier, in our example 2 W (maximum power P max ≈ 2.5 W). If the inflection of the characteristic is not clearly expressed, it is clarified using an oscilloscope by repeated measurements. Then take the arithmetic mean value U out, at which the distortions of the sine wave on the oscilloscope screen become indistinguishable to the eye.
The numerical value of the harmonic distortion coefficient Kg can be measured using a band-stop filter tuned to the fundamental frequency of 1 kHz. The filter is connected between the output of the low-pass amplifier and the millivoltmeter (Fig. 3). First, U out is measured at the first position of switch B. Let us assume that it is equal to 3.6 V (3600 mV). Then, setting the switch to the second position to turn on the filter, measure the harmonic voltage U g. Let's say it is 72 mV. After this, the harmonic coefficient is calculated using the previously given formula:
Now, using the amplitude characteristic, we determine the sensitivity of the amplifier. Since Uin at P nom is equal to 90 mV, therefore, the nominal sensitivity of the amplifier is also equal to 90 mV,
The connection diagram of devices with an amplifier for measuring the frequency response remains the same (see Fig. 1). The initial frequency of the input signal is the same - 1000 Hz. Using the Generator Amplitude knob, we set the voltage Uin equal to 20 mV, which we subsequently maintain constant at all frequencies (this voltage, which is almost five times less than the nominal sensitivity of the amplifier, was chosen for the convenience of reading measurement results on the scale of the avometer dial instrument). Then, switching the voltmeter to the output of the amplifier, we measure the voltage at the load equivalent R e. We record the measurement results in the table. 2 in two lines: in the first - the frequency f of the input signal, in the second - the output voltage U out. In the header of the table we write the name of the amplifier, the load equivalent resistance R e, the input voltage U in at which we make measurements (in this example, 20 mV).
Having recorded the measurement results at a frequency of 1000 Hz, we switch the low-frequency generator to the frequency, it is equal to 72 mV. After this, the harmonic coefficient is calculated using the previously given formula:
Now, using the amplitude characteristic, we determine the sensitivity of the amplifier. Since Uin at P nom is equal to 90 mV, therefore, the nominal sensitivity of the amplifier is also equal to 90 mV.
The frequency response of the amplifier is measured at an output power significantly lower than the rated power, which eliminates any overload of the amplifier. The frequency characteristics of industrial receiver amplifiers, for example, are measured at an output power of 50 and even 5 mW.
If the amplifier is relatively simple and does not have any tone controls, then the volume control is set to maximum and its position is not changed while taking the frequency response. If there is a fine-compensated volume control, the frequency response is measured at maximum, minimum and several, at the designer’s request, intermediate positions of the volume control.
The connection diagram of devices with an amplifier for measuring the frequency response remains the same (see Fig. 1). The initial frequency of the input signal is the same - 1000 Hz. Using the Generator Amplitude knob, we set the voltage Uin equal to 20 mV, which we subsequently maintain constant at all frequencies (this voltage, which is almost five times less than the nominal sensitivity of the amplifier, was chosen for the convenience of reading measurement results on the scale of the avometer dial instrument). Then, switching the voltmeter to the output of the amplifier, we measure the voltage at the load equivalent R e. We record the measurement results in the table. 2 in two lines: in the first - the frequency f of the input signal, in the second - the output voltage U out. In the header of the table we write the name of the amplifier, the load equivalent resistance R e, the input voltage U in at which we make measurements (in this example, 20 mV).
Having recorded the measurement results at a frequency of 1000 Hz, we switch the low-frequency generator to a frequency of 500 Hz. We check the input voltage of 20 mV with a voltmeter, then measure the output voltage of the amplifier as accurately as possible at the load equivalent R e. Next, we make measurements in exactly the same way at frequencies of 250, 150, 100, 75, 50 Hz and record the measurement results in a table (amateur amplifiers are usually not checked at a frequency of 25 Hz). After this, we carry out a repeated control measurement at a frequency of 1000 Hz to check the stability of the amplifier and measuring instruments.
Then we make measurements at higher frequencies. After a control frequency of 1000 Hz, we apply signals with frequencies of 2.5 to the amplifier input; 5; 7.5; 10; 15 kHz (measurements at a frequency of 20 kHz are made only when testing high-end amplifiers). We record the measurement results in a table and use them to calculate the ratio of the output voltages Uin to the control frequency voltage U1000. We write the resulting relationships in the corresponding row of the table.
For example. At frequencies of 50 Hz and 15 kHz, the output voltage U out = 300 mV. Therefore, the relationship
At frequencies of 100 Hz and 10 kHz we have the ratio
Now, having all the preliminary data, we proceed to drawing the frequency response of the amplifier (Fig. 4). Usually, special logarithmic paper is used for this purpose (auditory perception of sounds of different frequencies and loudness obeys the logarithmic law). However, to construct the frequency response, you can use any squared paper or graph paper. It is marked as shown in Fig. 4. First, we plot the frequency values along the horizontal ordinate axis. In Fig. 4, the top row of numbers corresponds to the fixed frequencies of the low-frequency generator of the Amateur Radio Laboratory. The bottom row of numbers, highlighted in color, corresponds to the frequencies recommended by GOST when taking characteristics using industrial measuring equipment.
Then along the vertical axis, having previously made 8-10 equally spaced marks on it, the ratio U f /U 1000 in decibels. Since the drop or rolloff of the frequency response we measured does not exceed 6 dB, we draw the zero line at the level of the 6th mark and put the numbers 0, -1, -2... -6 dB on the left. We also draw a control frequency line of 1000 Hz. Further, using the data in table. 2, sequentially place marks on the measuring frequencies from 50 Hz to 15 kHz. Since the characteristic has declines at the edges, we put the marks in decibels down from the zero line. For example, at a frequency of 50 Hz there was a decline of 6 dB, therefore, we set the mark at a level of - 6 dB. For a frequency of 75 Hz, we place the mark at a level of - 3 dB, etc. A smooth line drawn through these marks will be the frequency response. The horizontal line at -3 dB, which corresponds to the generally accepted tolerance for flatness in the frequency response, intersects the response at frequencies of 75 Hz and approximately 12 kHz. Therefore, the amplified frequency band, or bandwidth of the amplifier under test, is 75-12,000 Hz with a ripple of 3 dB.
High-quality bass amplifiers, in addition to volume controls, usually have two separate tone controls - for low and high frequencies. The frequency characteristics of such amplifiers are measured at least three times. First, both tone controls are set to positions corresponding to the greatest rolloff of the extreme low and high frequencies. The resulting characteristic may take the form of a curve indicated in Fig. 5 with the number 1. Then the knobs of both tone controls are turned to the other extreme position, corresponding to the maximum increase in low and high frequencies, and measurements are made at an input voltage that is ten times (20 dB) less than the nominal one. This characteristic may take the form of curve 2 (Fig. 5).
After this, the handles of both regulators are set to the middle positions and the third measurement is taken. If the obtained characteristic corresponds to or is close to curve 3, then the measurements are completed. If it differs significantly from this curve, then, by testing, find the positions of the control knobs at which the characteristic is most straightforward, and the corresponding marks are made on the control knobs.
From the graph in Fig. It is clearly seen that for a bass amplifier having such characteristics, the limit of tone control at the lowest frequency of 63 Hz (according to GOST) is +6 and -6 dB, and at the highest, equal to 12 kHz, from approximately +5 to -10 dB .
The most important characteristic of periodic processes is frequency, which is determined by the number of complete cycles (periods) of oscillations per unit time interval. Need for frequency measurement arises in many areas of science and technology and especially often in radio electronics, which covers a vast area of electrical oscillations from infra-low to ultra-high frequencies inclusive.
To measure the frequency of power supplies of electrical radio devices, electromagnetic, electro- and ferrodynamic frequency meters with direct assessment on the scale of a ratiometric meter, as well as tuning fork frequency meters, are used. These instruments have narrow measurement limits, typically within +-10% of one of the nominal frequencies of 25, 50, 60, 100, 150, 200, 300, 400, 430, 500, 800, 1000, 1500 and 2400 Hz, and operate at rated voltage 36, 110, 115, 127, 220 or 380 V.
Very low frequencies (less than 5 Hz) can be approximately determined by counting the number of complete oscillation periods over a fixed period of time, for example, using a magnetoelectric device connected to the circuit under study and a stopwatch; the desired frequency is equal to the average number of periods of oscillation of the instrument needle in 1 s. Low frequencies can be measured by the voltmeter method, the bridge method, and also by comparison with a reference frequency using acoustic beats or an electron beam oscilloscope. Frequency meters based on the capacitor charge-discharge and discrete counting methods operate in a wide range of low and high frequencies. To measure high and ultra-high frequencies (from 50 kHz and above), frequency meters based on resonant and heterodyne methods are used. At microwave frequencies (from 100 MHz and above), the method of directly estimating the wavelength of electromagnetic oscillations using measuring lines is widely used.
If the oscillations under study have a shape other than sinusoidal, then, as a rule, the frequency of the fundamental harmonic of these oscillations is measured. If it is necessary to analyze the frequency composition of a complex vibration, then special devices are used - frequency spectrum analyzers.
Modern measuring technology makes it possible to measure high frequencies with a relative error of up to 10 -11; this means that a frequency of approximately 10 MHz can be determined with an error of no more than 0.0001 Hz. Quartz, molecular and atomic oscillators are used as sources of highly stable reference frequencies, and tuning fork oscillators are used in the low frequency range. Frequency stabilization methods used at broadcasting stations make it possible to maintain the frequency with a relative error of no more than 10 -6, so their carrier frequencies can be successfully used as reference frequencies in frequency measurements. In addition, through the radio stations of the State Time and Frequency Service of the USSR, oscillations of a number of standard frequencies (100 and 200 kHz, 2.5; 5; 10 and 15 MHz), which represent an unmodulated carrier, periodically interrupted by the supply of call signs and precise time signals, are regularly transmitted.
In many cases of radio engineering practice, when measuring low frequencies, an error of up to 5-10% can be allowed, and when measuring high frequencies - up to 0.1-1%, which simplifies the requirements for the circuit and design of the frequency meters used.
Measuring frequency with a voltmeter
The simplest is an indirect method of measuring frequency, based on the dependence of the resistance of reactive elements on the frequency of the current flowing through them. A possible measurement scheme is shown in Fig. 1.
Rice. 1. Circuit for measuring frequency using a voltmeter
A chain of a non-reactive resistor R and a capacitor C with low losses, the parameters of which are precisely known, is connected to the source of frequency oscillations F x. A high-resistance AC voltmeter V with a measurement limit close to the input voltage value alternately measures the voltages U R and U C on the elements of the chain. Since U*R = I*R, and U C = I/(2πF x C) (where I is the current in the circuit), then the ratio U R /U C = 2πF x RC, which follows:
F x = 1/(2πRC) * U R /R C
The voltmeter's input resistance V must be at least 10 times the resistance of each element in the chain. However, the influence of a voltmeter can be eliminated if it is used only as an indicator of the equality of voltages U R and U C , achieved, for example, by a smooth change in resistance R. In this case, the measured frequency is determined by a simple formula:
F x = 1/(2πRC) ≈ 0.16/(RC),
and with a constant capacitance of the capacitor C, the variable resistor R can be equipped with a scale with a report in the values of F x.
Let us estimate the possible order of the measured frequencies. If resistor R has a maximum resistance R M = 100 kOhm, then at C = 0.01 μF, 1000 and 100 pF, the upper measurement limit will be 160, 1600 and 16000 Hz, respectively. When choosing R M = 10 kOhm and the same capacitance values, these limits will be equal to 1600 Hz, 16 and 160 kHz. The effectiveness of the method depends on the accuracy of the selection of denominations and the quality of the elements of the RC chain.
Capacitive frequency meters
For practical purposes, direct-indicating frequency meters are most convenient, allowing continuous observations of the frequency of the vibrations being studied on the dial meter scale. These include, first of all, capacitive frequency meters, the operation of which is based on measuring the average value of the charge or discharge current of a reference capacitor, periodically recharged by the voltage of the measured frequency f x. These instruments are used to measure frequencies from 5-10 Hz to 200-500 kHz. With an acceptable measurement error of approximately 3-5%, they can be performed using simple schemes, one of the options of which is shown in Fig. 2. Here, transistor T1, operating in switch mode, is controlled by frequency voltage f x, which is supplied to its base from the input potentiometer R1. In the absence of an input signal, transistor T1 is open, since its base is connected through resistors R3 and R2 to the negative pole of the power source. In this case, a voltage drop U is created across the resistor R5 of the divider R5, R2; the latter, due to the presence of a large capacitance capacitor C2, is fixed as the supply voltage of the transistor cascade and with rapid periodic changes in the transistor mode almost does not change. When installing the switch IN in the “U-” position, the And meter, connected in series with the additional resistor R6, forms a voltmeter that measures the constant voltage U on capacitor C2, which is maintained at a certain level, for example 15 V, with the help of trimming resistor R2. Instead of the one discussed, a standard circuit can be successfully used parametric voltage stabilization on the zener diode, which does not require systematic monitoring.
Rice. 2. Circuit of a capacitive frequency meter
During the positive half-cycle of the input voltage of frequency f x, transistor T1 closes and the voltage at its collector increases sharply to the value U; in this case, one of the capacitors C is quickly charged to a voltage close to U, the charging current of which flows through the meter AND and diode D2. During the negative half-cycle, transistor T1 opens, its resistance becomes very low, which leads to a rapid and almost complete discharge of capacitor C with current flowing through diode D1. During one period of the measured frequency, the amount of electricity imparted to the capacitor during charging and released by it during discharge is q ≈ CU. Since the charge-discharge process is repeated with a frequency f x, then the average value I charging current recorded by the meter AND, turns out to be proportional to this frequency:
I = q*f x ≈ C*U*f x .
This allows the meter to be equipped with a linear scale, calibrated directly in the values of the measured frequencies.
If the total deflection current of the meter I and the constant voltage U are known, then for a given limit value of the measured frequencies f p the capacitor must have a capacitance
C = I and /(U*f n).
For example, with the values of the circuit elements shown in Fig. 2, The frequency meter can be adjusted to operate at upper measurement limits of 100 Hz, 1, 10 and 100 kHz.
In this circuit, the switch on transistor T1 simultaneously performs the functions of an amplifier-limiter, due to which the frequency meter readings depend little on the shape of the input voltage. Any periodic input voltage with an amplitude of approximately 0.5 V and above is transformed into a pulse voltage of almost rectangular shape with a constant amplitude U f which powers the measuring (counting) circuit of the frequency meter. Capacitor C3, a shunt meter, smoothes out the ripples of the latter's needle when measuring the lowest frequencies of the general range.
Trimmer resistor R7, connected in parallel to the meter, serves to correct the frequency meter scale during its operation. In this case, a reference frequency voltage is supplied to the input of the frequency meter from a measuring generator or an alternating current network (50 Hz) and by adjusting resistance R7, the meter needle is deflected to the corresponding division of the frequency scale. This adjustment is repeated several times, alternating it with the above-mentioned setting of the supply voltage U, carried out using resistor R2.
An input voltage less than 0.3-0.5 V may not be sufficient to turn off transistor T1 for most of the positive half cycle; then capacitor C will not have time to charge to voltage U and the frequency meter readings will be underestimated. To increase the sensitivity of the input voltage to 20-50 mV, the electronic switch is sometimes preceded by an amplification stage, performed according to a circuit with a common emitter.
If the input voltage is excessive, the input transistor may be damaged; this leads to the need to include limiting or adjusting elements at the input, for example, potentiometer R1 in the circuit in Fig. 2. The input voltage should be increased gradually, monitoring the readings of the frequency meter, and when the latter, after a certain interval of increase, stabilize, the frequency f x can be estimated. It is useful to monitor the input voltage in order to set it at the optimal level for a given frequency meter, for example 1.5 V. In this circuit, this takes place in the “U~” position of switch B, when the meter with diodes D1, D2 and resistor R4 form an AC voltmeter current with a measurement limit of approximately 3 V, controlling the voltage taken from potentiometer R1.
Frequency meters made according to circuits similar to the one discussed above give fairly accurate readings only at input voltages close in shape to the voltages (usually sinusoidal) used during debugging and calibration of the device. Universal capacitive frequency meters allow you to measure frequencies of both continuous and pulsed voltages of any shape and polarity in a wide range of frequencies and input voltages 1. In the most general case, their functional circuit contains the following components connected in series: input divider - matching stage - amplifier - Schmitt trigger - differentiating circuit with a filter diode - standby multivibrator - counting circuit. A high-impedance input divider, usually stepped, increases the maximum permissible input voltages to hundreds of volts. An emitter or source follower provides a high input impedance to the device, weakening its influence on the circuits under test. The amplifier reduces the maximum permissible input voltage to tens of millivolts. The frequency oscillations f x amplified by it periodically trigger a Schmitt trigger, which generates rectangular pulses with a repetition frequency f x .
Rice. 3. Scheme of a universal capacitive frequency meter
Because the duration of these pulses depends on the frequency and amplitude of the input signal, they are not suitable for accurate frequency measurements. Therefore, with the help of a differentiating RC circuit, each rectangular trigger pulse is converted into a pair of pointed pulses of different polarities. One of these pulses, occurring on the falling edge of the rectangular pulse, is filtered by a diode, and the second, corresponding to the rising edge of the rectangular pulse of the flip-flop, is used to trigger the standby multivibrator. The latter produces rectangular pulses of strictly defined duration and amplitude, the repetition frequency of which is obviously equal to f x. As a result, a counting circuit with switchable capacitors of various ratings, rectifier elements and a dial meter ensures the measurement of frequency f x with complete independence of the reading from the amplitude and shape of the input voltage. In order to reduce the measurement error (not exceeding 1% in the best samples), the optimal duration of the multivibrator pulses is set at each frequency limit, approximately equal to half the period of the highest frequency of this measurement limit. If the universal frequency meter is powered from an alternating current network, then parametric stabilization of the rectified voltage must be carried out, and the network frequency of 50 Hz or its double value of 100 Hz (pulsation frequency) is used as a reference for scale correction.
In specific devices, the considered functional diagram is implemented in various versions. In Fig. Figure 3 shows a diagram of a relatively simple universal frequency meter with upper measurement limits of 200, 2000 and 20,000 Hz, in which the meter can be used AND with a total deviation current of 1-3 mA. The device contains an input step divider R1-R3, an amplifier on transistor T1, a Schmitt trigger on transistors T2 and T3, a differentiating circuit C3, R13 with diode D2, which passes only pulses of positive polarity, and a standby multivibrator on transistors T4, T5. A special feature of the frequency meter is the absence of special rectifier elements. The AND meter is included in one of the arms of the multivibrator, which is opened for a fixed time interval by differentiated trigger pulses, and registers the average value of the collector current, proportional to the frequency f x. The upper limits of measurements f p are determined by the duration of the multivibrator pulses, which are set by selecting the values of capacitors C4-C6 using trimming resistors R18-R20. Since in this circuit all the RC counting chains are interconnected, they should be adjusted in the following order: C4-R18, C5-R19 and C6-R20, followed by re-adjusting all limits with resistors R18-R20.
The measurement error of the frequency meter is determined mainly by the adjustment accuracy and stability of the standby multivibrator, therefore the supply voltage of the latter is stabilized by resistor R12 and zener diode D1. Using trimming resistor R4, the optimal bias is selected based on transistor T1 (4-5 V). If there is a high-frequency measurement limit (for example, up to 200 kHz), to increase the speed of the trigger and multivibrator, it is useful to connect small capacitors (tens of picofarads) in parallel with resistors R10 and R15.
Since the amplifier on transistor T1 operates in amplitude limiting mode, at input voltages up to 10-20 V you can do without an input voltage divider; In this case, a limiting resistor should be turned on at the input.
Electronic counting (digital) frequency meters
Electronic frequency counters are universal devices in their capabilities. Their main purpose is to measure the frequency of continuous and pulsed oscillations, carried out in a wide frequency range (from approximately 10 Hz to 100 MHz) with a measurement error of no more than 0.0005%. In addition, they make it possible to measure periods of low-frequency oscillations, pulse durations, the ratio of two frequencies (periods), etc.
The operation of electronic counting frequency meters is based on a discrete counting of the number of pulses arriving over a calibrated time interval to an electronic counter with a digital display. In Fig. Figure 4 shows a simplified functional diagram of the device. The voltage of the measured frequency f x in the amplifier-forming device is converted into a sequence of unipolar pulses repeated with the same frequency f x . For this purpose, a system of a limiting amplifier and a Schmitt trigger is often used, supplemented at the output with a differentiating circuit and a diode limiter (see and Fig. 3). A time selector (an electronic switch with two inputs) passes these pulses to the electronic counter only during a strictly fixed time interval Δt, determined by the duration of the rectangular pulse acting on its second input. When registering m pulses with a counter, the measured frequency is determined by the formula
For example, if during a time Δt = 0.01 s 5765 pulses are noted, then f x = 576.5 kHz.
The frequency measurement error is determined mainly by the calibration error of the selected counting time interval. The master component in the system for forming this interval is a highly stable quartz oscillator, say, with a frequency of 100 kHz. The oscillations it creates with the help of a group of series-connected frequency dividers are converted into oscillations with frequencies (f 0) 10 and 1 kHz, 100, 10, 1 and 0.1 Hz. which correspond to periods (T 0) 0.0001; 0.001; 0.01; 0.1; 1 and 10 s (the last one or two of the indicated values of f 0 and T 0 are missing for some frequency meters).
Oscillations of the selected (via switch B2) frequency f 0 (the numerical value of the latter is a multiplier to the counter count) are converted into rectangular oscillations with a repetition frequency f 0 using a Schmitt trigger. Under their action, an interval pulse of duration Δt = T 0 = 1/f 0 of a strictly rectangular shape is formed in the control device. This pulse causes the previous counter readings to be reset, and then (with a delay of several microseconds) arrives at the selector and opens it for a time Δt to pass pulses with a repetition frequency f x. After closing the selector, the number of pulses m passed by it is recorded by the counter indicator, and the measured frequency is determined by the formula f x = m*f 0 .
Rice. 4. Simplified functional diagram of an electronic counting (digital) frequency meter
The selector control circuit can be started manually (by pressing the “Start” button); in this case, the control device sends a single pulse of duration Δt to the selector and the counter produces a one-time measurement result with an unlimited indication time. In automatic frequency measurement mode, time relay pulses are periodically repeated and measurement results are updated at selected time intervals.
The frequency meter can serve as a source of oscillations of a number of reference frequencies f 0, obtained using a quartz oscillator, multiplier and frequency dividers and taken from a special output. The same oscillations, applied to the input of the frequency meter, can serve to check the correctness of the meter readings.
The frequency meter counter is assembled from 4-7 recalculating decades on trigger circuits and digital indicator lamps. The number of decades determines the maximum number of significant figures (digits) in the measurement results. The possible counting error, called discreteness error, is one unit in the least significant digit. Therefore, it is desirable to select a counting time interval Δt that uses the maximum number of counter digits. So, in the example discussed above, with Δt = 0.01 s (f 0 = 100 Hz), four digits of the counter and the measurement result f x = 576.5 kHz + -100 Hz were sufficient for the count. Let us assume that the measurements are repeated at Δt = 0.1 s (f 0 = 10 Hz) and a count of m = 57653 pulses is obtained. Then f x = 576.53 kHz +-10 Hz. An even smaller discreteness error (+-1 Hz) will be obtained at Δt = 1 s (in this case, the counter must have at least six decades).
When expanding the measurement range of a frequency meter towards high frequencies, the limiting factor is the speed of the recalculation decades. When implementing trigger circuits on high-frequency silicon transistors (for example, type KT316A), which have a charge resorption time in the base of approximately 10 ns, the upper limit measurable frequency can reach tens of megahertz. In some instruments, when measuring high frequencies exceeding, for example, 10 MHz, they are first converted to a frequency less than 10 MHz (for example, a frequency of 86.347 MHz to a frequency of 6.347 MHz), using the heterodyne method (see).
The factor limiting the lower limit of the measured frequency is the measurement time. If, for example, we set the largest counting time interval for many frequency meters to Δt = 1 s, then when the counter registers 10 pulses, the measurement result will be the frequency f x = 10 = +-1 Hz, i.e. The measurement error can reach 10%. To reduce the error, let's say, to 0.01%, it would be necessary to count pulses over a period of time Δt = 1000 s. Even more time is required to accurately measure frequencies equal to 1 Hz or less. Therefore, in electronic frequency counters, the measurement of very low frequencies f x is replaced by the measurement of their oscillation period T x = 1/f x. The oscillation period measurement circuit is formed when the switch is installed IN 1 to the “Tx” position (Fig. 4). The voltage under study, after conversion in the Schmitt trigger, acts on the control device, in which a rectangular pulse of duration T x is formed, maintaining the time selector in the open state; during this time, the counter registers pulses generated from oscillations of one of the reference frequencies f o, determined by the setting of the switch AT 2. For the number m of marked pulses, the measured period
For example, with m = 15625 and f 0 = 1000 Hz, the period T x = 15.625 s, which corresponds to the frequency f x = 1/T x = 0.054 Hz. In order to reduce their error, it is advisable to make measurements at the highest possible frequency f o (excluding, of course, overloading the meter). If period T x< 1 с (f x >1 Hz), then it may be rational to use frequency oscillations f 0 equal to 1 or 10 MHz, obtained after frequency multipliers. In this case, the lower limit of measured frequencies can be expanded to 0.01 Hz.
The measurement of the ratio of two frequencies f 1 / f 2 (f 1 > f 2) corresponds to setting switches B2 to the “Off” position, and B1 to the “f x” position. A voltage of lower frequency f 2 is applied to the “f o” terminals, and its period determines the counting time interval Δt. The frequency voltage f 1 supplied to the input is converted into pulses, the number of which (m) is recorded by the counter during the time Δt = 1/f 2. The desired frequency ratio f 1 / f 2 = m (with an error of up to unity). Obviously, this method makes sense to find the ratio of only significantly different frequencies.
The disadvantages of electronic frequency counters include the complexity of their circuits, significant dimensions and weight, and high cost.
Oscillographic frequency measurement methods
The measured frequency can be determined by comparing it with a known reference frequency f o . This comparison is most often made using a cathode ray oscilloscope or beat methods.
Cathode ray oscilloscopes are used to measure oscillation frequencies of primarily sinusoidal waveforms over a frequency range of approximately 10 Hz to a value determined by the upper limit of the deflection channel bandwidth; the measurement error is practically equal to the calibration error of the oscillation source (generator) of the reference frequency f 0 . Most often, measurements are carried out with the oscilloscope scan turned off, using the connection diagram shown in Fig. 5. Voltages of the measured and known frequencies are applied directly or through amplifiers to various pairs of deflecting plates of the CRT (depending on which input of the oscilloscope these voltages affect, we will denote their frequencies by f x and f y). If these frequencies are related to each other as integers, for example 1:1, 1:2, 2:3, etc., then the movement of the electron beam becomes periodic and a stationary image called a Lissajous figure is observed on the screen. The shape of this figure depends on the ratio of amplitudes, frequencies and initial phases of the compared oscillations.
Rice. 5. Frequency measurement scheme using the Lissajous figure method
In Fig. Figure 6 shows the formation of a Lissajous figure when the deflecting plates of the tube are exposed to two sinusoidal oscillations of the same frequency and equal amplitudes, but having different initial phases. This figure has the appearance of an inclined ellipse, which, with phase shifts between oscillations of 0 and 180°, is compressed into a straight inclined line, and with phase shifts of 90° and 270°, it turns into a circle (we conventionally assume that the sensitivity to deflection of both pairs of plates is the same). If the voltage amplitudes of the frequencies f x and f y are not equal, then in the latter case, instead of a circle, an ellipse with axes parallel to the planes of the deflecting plates will be observed on the screen.
Rice. 6. Construction of an oscillogram with the ratio of compared frequencies f x /f y = 1
If the frequency ratio f x /f y (or f y /f x) is equal to two, then the figure on the screen takes the form of a figure eight, which, with initial phase shifts of 90 and 270°, contracts into an arc. (The initial phase shift is always evaluated relative to the period of the higher frequency voltage). From the table shown in Fig. 7, it is clear that the greater the number of the fraction characterizing the ratio of the compared frequencies, the more complex the Lissajous figure observed on the screen.
During the measurement, the frequency of the reference oscillator f 0 (equal to f x or f y) is smoothly changed until one of the Lissajous figures of the simplest possible shape appears on the screen. This figure is mentally crossed by lines xx and y, parallel to the planes of the deflecting plates X1, X2 and Y1, Y2, and the number of intersections of each line with the figure is counted. The ratio of the obtained numbers is exactly equal to the ratio of the frequencies f x:f y, provided that the drawn lines do not pass through the nodal points of the figure or tangent to it, and the shape of the compared oscillations is close to sinusoidal.
Rice. 7. Figures observed on the screen at different frequency ratios f x / f y
Having determined the ratio f x:f y and knowing one of the frequencies, for example f y, it is easy to find the second frequency.
Let us assume that at a known frequency f y = 1000 Hz the figure shown in Fig. is obtained on the screen. 5. From the construction shown in the drawing it is clear that this figure corresponds to the frequency ratio f x:f y = 3:4, from which f x = 750 Hz.
Due to some instability of the compared frequencies, the integer or fractional-rational relationship established between them is constantly violated, which leads to a gradual change in the shape of the observed figure, successively passing through all possible phase states. If we fix the time Δt during which the figure undergoes a full cycle of phase changes (from 0 to 360°), then we can calculate the difference between the compared frequencies |f x - f y | = 1/Δt, the sign of which can be easily determined experimentally by slightly changing the frequency f 0 . At high frequencies, even a very small instability of one of the frequencies causes such rapid changes in the Lissajous figure that it becomes impossible to determine the frequency ratio. This limits the upper limit of measurable frequencies to approximately 10 MHz.
Rice. 8. Circuit for measuring frequency using the circular scanning method with brightness modulation
When the integer ratio of the compared frequencies exceeds 8-10, or their fractional ratio with numbers in the denominator or numerator greater than 4-5, due to the complication of the Lissajous figure, the possibility of error in establishing the true frequency ratio increases. Accurate determination of relatively large integer frequency ratios (up to 30-50) can be carried out using the circular scanning method with image brightness modulation (Fig. 8). In this case, a voltage of a lower frequency f 1, using two identical phase-splitting RC circuits, is converted into two voltages of the same frequency, mutually shifted in phase by 90°. When these voltages are applied to the Y and X inputs of the oscilloscope, respectively, and the ratio of their amplitudes is adjusted by resistors R and the gain controls of the Y and X channels, the light spot on the screen will move along a curve close to a circle; the latter is set clearly visible using the brightness control. A voltage of a higher frequency f 2 is applied to the input of the modulator M (or channel Z) and it will periodically increase and decrease the intensity of the electron beam, and therefore the brightness of individual sections of the scan curve on the screen. With an integer frequency ratio f 2: f 1 = m, achieved by changing one of them, the curve of the observed circle becomes dashed, it consists of f motionless luminous segments of equal length, separated by dark intervals. When the integer relation is violated, rotation of the dashed circle is observed, at high speed of which the circle appears solid.
The considered method can also be used to measure the repetition frequency f p of pulse oscillations. In this case, a circular sweep is carried out using the reference frequency voltage f 0, and with a brightness control it is set visible or invisible depending on the polarity (negative or positive, respectively) of the pulse oscillations supplied to the modulator. The latter will create dark breaks on the scan line in the first case, and luminous dots in the second. By smoothly changing the frequency fo (from its minimum possible value), one achieves one stationary or slowly moving pulse trace on the scan line, with f p = f 0.
The frequency fp of pulse oscillations can be measured using the diagram in Fig. 5 when applying a sinusoidal voltage of the reference frequency f 0 to the X input, and a pulse voltage to the Y input of the oscilloscope. The scanning frequency f 0 = f x is gradually increased, starting from its lowest value, until a fairly stable image of a single pulse appears on the screen, which occurs when f p = f 0 . This measurement technique eliminates the possibility of error, since a single pulse will be observed on the screen for other, larger than unity, integer frequency ratios f 0:f n.
Frequency measurement using beat methods
The source of oscillations of the reference frequencies is usually a measuring generator with a smooth or stepless setting, the frequency of which f 0 can be set equal to the measured frequency f x . If the frequencies f 0 and f x are sound, then their equality can be approximately judged by listening in turn to the tones of the vibrations they create using telephones or a loudspeaker.
The measurement error is reduced almost to the calibration error of the measuring generator if electrical oscillations of both compared frequencies are simultaneously applied to the phones in accordance with the diagram in Fig. 9, a. If the frequencies f 0 and f x are close to each other, then when the corresponding oscillations are added, acoustic beats arise, which manifest themselves in a periodic increase and decrease in the intensity of the T f tone heard in phones. Beat frequency
can be determined by listening to the number of increases or decreases in tone intensity over a fixed period of time. In order for the beats to appear quite sharply, the amplitudes of the oscillations of the frequencies f 0 and f x must be set approximately the same; this follows from consideration of Fig. 9, b, where the average curve of oscillations pulsating with frequency F is the result of adding the upper and lower oscillation curves corresponding to frequencies f 0 and f x.
Rice. 9. To the principle of measuring low frequencies using the acoustic beat method
By changing the generator settings, the frequency f 0 is brought closer to the frequency f x , which is detected by an increase in the beat period. When the compared frequencies coincide, the beats disappear and a monotonous tone is heard in the phones. Instead of telephones, an AC voltmeter can be used as a beat indicator; This is especially useful when measuring frequencies above 5 kHz, the tone of which is not clearly audible on phones.
At high frequencies, comparison of frequencies f 0 and f x is most often carried out using the zero beat method. In Fig. 10 shows the simplest measurement scheme. Frequency oscillations f 0 and f x are simultaneously introduced into the diode circuit D through coupling coils L1, L2 and L. As a result of detecting the total oscillation, a pulsating current appears in the diode circuit, containing components of the fundamental frequencies f 0 and f x, as well as components of higher harmonics and combination frequencies f 0 + f x and |f 0 - f x |. If the frequencies f 0 and f x are close to each other, then the difference beat frequency F = |f 0 - f x | may be within the range of audio frequencies and the tone of this frequency will be heard in Tf telephones, shunted from high-frequency currents by capacitor C.
Rice. 10. To the principle of measuring high frequencies using the zero beat method
If you change one of the frequencies, for example f o , bringing it closer to another frequency f x , the tone in the phones will decrease and if these frequencies are equal, zero beats will be observed, detected by the loss of sound in the phones. Thus, frequency measurement is reduced to determining the frequency of the reference oscillator at which zero beats occur. As can be seen from the graph in Fig. 11, a, when moving away from the point of zero beats, the difference frequency F increases both with an increase and a decrease in the generator frequency f 0.
Rice. 11. Graphs of the dependence of the beat frequency on the settings of the reference frequency generator
The frequency measurement error is determined mainly by the frequency calibration error f 0 of the reference oscillator. However, when making accurate measurements, it is necessary to take into account a possible error of several tens of hertz, due to the fact that the human hearing system does not perceive tones with a frequency below a certain frequency F n; the values of the latter for different people range from 10-30 Hz. To eliminate this error, you can connect a magnetoelectric current meter in series with the telephones T f, the needle of which, at a very low difference frequency F, will pulsate with this frequency. When approaching zero beats, the oscillations of the needle slow down and are easy to count in a fixed period of time.
The connection between the reference oscillator and the source of the measured frequency should not be strong in order to avoid the occurrence of the “locking” phenomenon, leading to an increase in measurement error. If there is a strong connection between two generators, the difference in frequency settings of which is small, one of the generators can impose its frequency on the other and both generators will create oscillations of the same frequency. In this case, the beat frequency F changes in accordance with the graph in Fig. 11, b, i.e. in the entire “capturing” area it turns out to be zero and there is no sound in the phones.
As a sensitive indicator of zero beats, you can use a cathode-ray oscilloscope, preferably with an open input on channel Y. In this case, instead of telephones, a resistor with a resistance of 50-200 kOhm is connected as a load of the detector circuit (Fig. 10), the voltage from which is supplied to the input U of the oscilloscope. When scanning is turned on, the voltage curve of the beat frequency F is visible on the screen. As it approaches zero beats, the period of this voltage will increase and at f 0 = f x only a horizontal scanning line is visible on the screen. If measurements are carried out with the scan turned off, then the vertical line observed on the screen at f 0 = f x turns into a point.
The operation of quartz calibrators and heterodyne frequency meters is based on the principle of measuring high frequencies using the zero beat method.
Quartz calibrators
Of the high-precision instruments used to measure high frequencies, the simplest are quartz calibrators. They allow you to check the scales of radio receiving and radio transmitting (generating) devices at a number of points corresponding to strictly defined (reference) frequencies.
Rice. 12. Functional diagram of a quartz calibrator
The functional diagram of a quartz calibrator is shown in its most complete version in Fig. 12. The main component of the device is a quartz oscillator operating in such a mode that the oscillations excited by it have a shape that is sharply different from sinusoidal, and therefore contain, in addition to the component of the fundamental frequency f 0, a large number of harmonics, the frequencies of which are 2f 0, 3f 0, 4f 0, etc., and the amplitudes gradually decrease with increasing frequency. It is usually possible to use for measurements from tens to several hundred harmonics, which have the same high stability (usually within 0.01 - 0.001%) as the frequency f 0) stabilized by a quartz resonator (quartz) in the absence of special devices (for example, thermostats) increasing the stabilization effect.
The oscillations excited by the quartz oscillator are supplied to the communication socket (or clamp) An, which, together with a small conductor or pin attached to it, plays the role of a receiving or transmitting antenna, depending on the nature of the device’s use. For shielding purposes, the device is usually placed in a metal casing.
When checking the scales of radio receivers, the calibrator serves as a source of oscillations of a number of reference frequencies emitted through the communication wire. The receiver is sequentially tuned to various harmonics of the quartz oscillator and the corresponding scale points are determined. If the receiver operates in telegraph mode, then its tuning to the generator harmonic is recorded by zero beats with the frequency of the second local oscillator, heard in telephones or a loudspeaker connected to the receiver output. The scales of direct amplification receivers are checked with feedback brought to generation. To check the calibration of receivers operating only in telephone mode, for example, broadcasting ones, the oscillations of a quartz oscillator must be modulated with an audio frequency, which requires the introduction of an oscillation generator with a frequency of 400 or 1000 Hz into the calibrator (in devices with mains power supply, a voltage with a frequency of 50 or 100 Hz is sometimes used for modulation Hz). In this case, the receiver is tuned to the harmonics of the quartz oscillator according to the highest volume of the tone reproduced by the loudspeaker, or, more precisely, according to the maximum readings of the voltmeter connected to the output of the receiver.
If the quartz calibrator is also intended for checking the scales of high-frequency oscillators, for example radio transmitters, then it is supplemented with a detector (mixer), the input of which is connected to the An communication socket and the output of the quartz oscillator. The oscillations of the transmitter being tested, induced in the communication conductor, create beats with the harmonic of the quartz oscillator closest to them in frequency; As a result of detection, oscillations of the difference beat frequency are released, which, after amplification, are heard in telephones. The transmitter is sequentially tuned to the frequencies of a number of harmonics of the generator using zero beats and thereby determining the corresponding points on the transmitter frequency scale.
The main disadvantage of quartz calibrators is the ambiguity of measurement results, since zero beats only allow one to establish the fact that the measured frequency is equal to one of the harmonics of the quartz oscillator without fixing the number of this harmonic. To avoid errors in establishing the frequency of the harmonic that creates zero beats, it is desirable that the device under study has a frequency scale approximately calibrated using some device with an unambiguous frequency estimate (resonant frequency meter, measuring generator, etc.), the measurement accuracy of which can be small.
The frequency difference between adjacent reference points of the calibrator is equal to the fundamental frequency of the quartz oscillator f 0 . In order to cover the main broadcasting bands, the frequency f 0 is often taken equal to 100 kHz, which ensures that the scales of radio devices can be checked up to frequencies of the order of 10 MHz (λ = 30 m). To expand the range of measured frequencies towards shorter waves and eliminate errors in determining the frequency of the harmonics used, it is possible to operate a quartz oscillator at two stabilized and 10-fold fundamental frequencies, usually equal to 100 and 1000 kHz. Each of these frequencies has its own grid of reference points. The principle of sharing both fundamental frequencies can be understood from the following example. Let's assume that the transmitter tuning is being checked at a frequency of 7300 kHz. Then the calibrator is initially turned on at the fundamental frequency of 1000 kHz. The transmitter is tuned according to zero beats to the frequency closest to the desired one, a multiple of 1000 kHz, i.e., to a frequency of 7000 kHz. At this frequency, the possibility of error is practically eliminated, since the reference points are located rarely, every 1000 kHz. The calibrator is then switched to the fundamental frequency of 100 kHz; with precise adjustment of the quartz, zero beats should be preserved. The transmitter tuning is smoothly changed towards the required frequency and the scale points corresponding to zero beats at frequencies 7100, 7200 and 7300 kHz are marked sequentially.
If it is necessary to reduce the interval between adjacent reference frequencies, then frequency dividers are used, which are usually implemented using a multivibrator circuit synchronized at a subharmonic of the input signal. Thus, using two division stages with division factors equal to 10, with a fundamental frequency of a quartz oscillator of 1 MHz, it is possible to obtain oscillations with fundamental frequencies of 100 and 10 kHz and a large number of harmonics. Then the scale point corresponding, for example, to a frequency of 7320 kHz will be identified by sequentially passing reference points at frequencies 7000, 7100, 7200, 7300, 7310 and 7320 kHz. With a fundamental quartz frequency of 100 kHz, two dividers can produce oscillations with fundamental frequencies of 10 and 1 (or 2) kHz, but their harmonics at high frequencies will be very weak. Oscillations of combination frequencies with small intervals between reference points, but having significant intensity, can be obtained by mixing oscillations of several fundamental frequencies.
Rice. 13. Scheme of a universal quartz calibrator
In Fig. Figure 13 shows a diagram of a simple quartz calibrator suitable for measuring the frequency of generator and radio receiving devices. A quartz oscillator on transistor T2 excites oscillations of the fundamental frequency of 100 or 1000 kHz, depending on the setting of the switch AT 2. Precise adjustment of the fundamental frequencies to the nominal values is carried out by the tuning cores of the coils L1 and L2. The distortion of the oscillation shape, necessary to obtain a large number of harmonic components, is achieved by connecting diode D1 between the emitter and the base of transistor T2. If it is necessary to modulate these oscillations, switch B1 starts the low-frequency generator on transistor T1. Beat detection is carried out by diode D2, high-frequency components of the rectified current are filtered by capacitor C9.
The beat frequency voltage, amplified by transistor T3, creates sound vibrations in Tf phones.
Rice. 14. Schematic of a quartz calibrator with a frequency divider
In Fig. Figure 14 shows a diagram of a quartz calibrator designed for calibrating the frequency scales of radio receivers. A quartz oscillator on transistors T1 and T2 excites frequency oscillations of 100 kHz. Precise adjustment of the frequency to the nominal value can be done by selecting the capacitance of capacitor C2 or by using a small-capacity tuning capacitor connected in parallel with the contacts of the quartz holder. The parameters of the multivibrator on transistors T3, T4, which serves to divide the frequency by 10 times, are selected such that in the free self-oscillation mode it generates oscillations with a frequency slightly less than 10 kHz. Then, when exposed to oscillations of a quartz oscillator, it will be synchronized at a frequency of 10 kHz; this must be carefully checked when setting up the device: between oscillations of adjacent harmonics of a frequency of 100 kHz at 9 points on the scale of the device being tested, harmonics of a frequency of 10 kHz should appear. The abundance of harmonics is facilitated by a reduction in the duration of the pulses using differentiating chains C3, R6 and C6, R12, as well as amplification of the pulses by a pulse amplifier on transistor T5 switched on at the output.
When operating quartz calibrators, it should be taken into account that due to aging, the natural frequency of quartz resonators changes slightly over time.
Heterodyne frequency meters
Heterodyne frequency meters are used for precise frequency measurements in a smooth high-frequency range. In principle, a heterodyne frequency meter differs from a quartz calibrator, made according to the functional diagram in Fig. 12, only in that instead of a quartz oscillator it uses a local oscillator, that is, a low-power generator with a continuously variable tuning frequency. The presence of a mixer allows the device to be used not only for calibrating the frequency scales of radio receivers, but also for measuring the frequency of generators using the zero-beat method. Indication of zero beats is carried out by telephones, oscilloscope and electronic light indicators, as well as dial meters.
The measurement error of a heterodyne frequency meter is mainly determined by the stability of the local oscillator frequency and the error of its setting. Therefore, they often prefer to perform local oscillators using vacuum tubes. Increased frequency stability is facilitated by the correct choice of circuit and design of the local oscillator, the use of parts with a low temperature coefficient, the inclusion of a buffer stage between the local oscillator and output circuits, stabilization of supply voltages, and long-term heating of the device under current before measurements. To increase the smoothness of adjustment and accuracy of frequency setting, the local oscillator tuning capacitor is usually controlled through a vernier mechanism with a large retardation (up to 100-300 times). Direct frequency reading on the scale of a variable capacitor is carried out only in the simplest designs; in most instruments, the scale is uniform with a very large number of divisions (up to several thousand), and the reading on it is converted into frequency using tables or graphs.
In order to reduce the number of frequency subranges and increase frequency stability, local oscillators usually operate in a narrow range of relatively low frequencies (with an overlap coefficient of two), and for measurements both the fundamental frequencies of the generated oscillations and a number of their harmonics are used; the occurrence of the latter is ensured by selecting the operating mode of the local oscillator or buffer amplifier. For example, in a widely used frequency meter of the Ch4-1 type with a general range of measured frequencies from 125 kHz to 20 MHz, the local oscillator has two smooth subranges of the main frequencies: 125-250 kHz and 2-4 MHz. In the first sub-band, when using the first, second, fourth and eighth harmonics, it is possible to smoothly cover the frequency band 125-2000 kHz; in the second subband, when using the first, second, fourth and partially fifth harmonics, the frequency band of 2-20 MHz overlaps. Thus, each position of the local oscillator tuning knob corresponds to three or four operating frequencies, the values of which can be determined from the calibration table. For example, frequencies of 175, 350, 700 and 1400 kHz are measured with the same local oscillator setting at the fundamental frequency f g = 175 kHz.
The ambiguity of local oscillator tuning frequencies creates the possibility of error in establishing the harmonic with which oscillations of the measured frequency f x create beats. Therefore, when starting measurements, it is necessary to know the approximate value of the frequency f x. However, the latter can also be determined by calculation using the heterodyne frequency meter itself.
Let us assume that when changing the local oscillator setting, zero beats with a frequency f x are obtained at two adjacent values of the fundamental frequencies f g1 and f g2 of the same local oscillator subrange. It is obvious that the frequency f x is simultaneously a harmonic of both of these frequencies, i.e.
f x = n*f g1 = (n+1)*f g2 .
where n and (n + 1) are the numbers of harmonics, respectively, for the fundamental frequencies f g1 and f g2 (with f g2< f г1).
Solving the resulting equality for n, we find
n = f g2 /(f g1 -f g2).
Therefore, the measured frequency
f x = n*f g1 = f g1 *f g2 / (f g1 -f g2).
For example, if zero beats are obtained at fundamental frequencies f g1 ≈ 1650 kHz and f g2 ≈ 1500 kHz, then approximately f x ≈ 1650*1500/(1650 - 1500) = 16500 kHz.
When measuring frequency, you should beware of errors caused by the possibility of beats occurring between the oscillations of the local oscillator and the harmonic of the measured frequency; Therefore, measurements should be carried out with a weak connection between the frequency meter and the generator under study. The measurement error also increases when the device is exposed to modulated vibrations; in this case, beats with the main (carrier) frequency will be heard against the background noise of beats with side frequencies.
Heterodyne frequency meters of the type considered provide measurement of high frequencies with an error of approximately 1%. Reducing the measurement error to 0.01% or less is achieved by adding a quartz oscillator to the frequency meter, which makes it possible to check and correct the local oscillator scale at a number of reference points before starting measurements.
An expanded functional diagram of a high-precision heterodyne frequency meter is shown in Fig. 15. The local oscillator has two subranges, the adjustment of which is carried out by trimming capacitors C3 and C4. The frequency of the fundamental oscillations is set by a direct-frequency variable capacitor C1. The level of the input (output) signal is controlled by potentiometer R. The crystal oscillator creates harmonic-rich oscillations, the fundamental frequency of which is often taken to be 1 MHz. The type of operation of the device is selected without disrupting interstage connections by turning on or off the power of individual components. When switch B2 is set to position 3 (“Quartz”), the local oscillator is turned off and the crystal oscillator is turned on; in this case, the frequency meter can be used as a quartz calibrator for frequency measurements on generator harmonics. In switch position 1 (“Loterodyne”), on the contrary, the crystal oscillator is turned off and the local oscillator is on. This is the normal operating mode of the frequency meter.
Rice. 15. Functional diagram of a high-precision heterodyne frequency meter
The local oscillator frequency scale is checked by setting switch B2 to position 2 (“Check”), when both the local oscillator and the generator are turned on at the same time, the oscillations of which are supplied to the detector. At a certain ratio of frequencies or harmonics of these vibrations, sound beats arise, the frequency of which is determined by the formula
F = |m*f g - n*f k |,
where f g and f k are the fundamental frequencies of the local oscillator and quartz oscillator, respectively, and m and n are integers corresponding to the numbers of interacting harmonics.
The beat frequency turns out to be zero (F = 0) for a number of frequencies in the local oscillator range that satisfy the condition
f g =(n/m)*f c.
These frequencies are called reference frequencies and are specially highlighted in calibration tables. Let us find, for example, the reference frequencies (f 0) of the local oscillator range 2000-4000 kHz, if the fundamental frequency of the quartz oscillator f k = 1000 kHz:
at m = 1 and n = 2, 3 and 4 f 0 = 2000, 3000 and 4000 kHz; at m = 2 and n = 5 and 7 f 0 = 2500 and 3500 kHz;
at m = 3 and n = 7, 8, 10 and 11 f 0 = 2333, 2667, 3333 and 3667 kHz, etc.
It should be taken into account that as the numbers of interacting harmonics increase, the amplitude of the beats decreases.
If the calibration of the local oscillator scale is violated, then when its tuning knob is set to one of the reference frequencies and the quartz oscillator is turned on, instead of zero beats, audio frequency oscillations are created, which, after amplification, are heard in telephones. For correction (calibration), a small-capacity capacitor C2 is used, connected in parallel to the main adjustment capacitor C1: with its help, before starting measurements, zero beats are achieved at the reference point closest to the frequency being measured.
Let's look at the procedure for setting up a heterodyne frequency meter using the following example. Suppose you want to check the correctness of the transmitter scale at a frequency of 10700 kHz. Referring to the calibration table of the frequency meter, we find that this frequency corresponds to the fundamental frequency of 10700/4 = 2675 kHz. Using the table or scale of the main points, we determine that the nearest reference frequency is 2667 kHz. Then, on the scale of capacitor C1, we set the frequency to 2667 kHz and, by placing switch B2 in the “Check” position (2), we use corrector C2 to achieve zero beats. Then we set switch B2 to the “Loterodyne” position (1) and, having set the local oscillator frequency to 2675 kHz, we check the transmitter scale at this frequency.
When measuring an unknown frequency f x, the local oscillator scale is calibrated at the reference point closest to the expected value of this frequency, and then in the measurement mode, zero beats are set by adjusting the local oscillator frequency.
When calibrating the local oscillator scale, as well as when measuring the frequency of generators, the modulator must be turned off; When measuring the tuning frequency of receivers, the low-frequency unit of the device is not needed. Use a switch to turn off unused frequency counter components. AT 3.
Heterodyne frequency meters of various types of industrial production collectively cover the range of measured frequencies from 100 kHz to 80 GHz with a measurement error within +-(5*10 -4 ...5*10 -6). At very high frequencies it is difficult to obtain zero beats. Therefore, in microwave frequency meters, a low-frequency frequency meter (for example, a capacitive one) is sometimes used as an indicator; it is used to determine the difference beat frequency F, the size of which is corrected in the measurement results.
A very small measurement error in a very wide frequency range (from low to ultra-high) is achieved by combining two frequency meters: a heterodyne and an electronic counter. The latter, in addition to its independent use in its inherent frequency range, can be used to accurately measure the local oscillator tuning frequency when zero beats are achieved; in this case, a quartz oscillator, calibration tables and graphs turn out to be unnecessary.
Resonant frequency meters
Features of resonant frequency meters used to measure high and ultra-high frequencies are simplicity of design, speed of operation and unambiguity of measurement results; The measurement error is 0.1-3%.
A resonant frequency meter is an oscillatory system that is tuned into resonance with the measured frequency f x of the oscillations that excite it, which come from the source under study through the coupling element. The resonant frequency is determined by the readings of a calibrated tuning device. The resonance state is recorded using a built-in or external indicator.
Frequency meters that measure frequencies from 50 kHz to 100-200 MHz are made in the form of an oscillatory circuit made of elements with lumped constants: an inductor L 0 and a variable capacitor C 0 (Fig. 16). E.M.F. is induced in the frequency meter circuit. measured frequency f x , for example, due to inductive coupling with the oscillation source through a coil L 0 or a small whip antenna connected to the An socket. With a low-power source, the connection with the latter can be capacitive through a coupling capacitor C St (with a capacity of several picofarads) and a coupling conductor. By changing the capacitance of the capacitor C 0, the circuit is tuned into resonance with the frequency fx according to the maximum readings of the resonance indicator. In this case, the measured frequency f x is equal to the natural frequency of the circuit:
f 0 = 1/(2π*(L0C0) 0.5),
determined by the scale of the capacitor C 0.
With a fixed inductance L 0, the range of measured frequencies is limited by the overlap coefficient, which is understood as the ratio of the maximum tuning frequency of the frequency meter f m to the lowest frequency f n when the circuit capacitance changes from the initial value C n to the maximum C m. The initial capacitance of the circuit C n is composed of the initial capacitance of the capacitor C 0, installation capacitances and capacitances of permanent or tuning capacitors included in the circuit in order to obtain the required overlap coefficient or for other purposes (Fig. 17). If it is necessary to expand the range of measured frequencies, the frequency meter is equipped with several coils of different inductance, replaceable (Fig. 16) or switchable (Fig. 17). In the latter case, it is advisable to short-circuit unused coils (if they are not shielded) to prevent them from sucking energy from the frequency meter circuit at tuning frequencies close to the natural frequencies of these coils; in this case, communication with the source of oscillations is carried out through the communication socket An or through an external communication coil L St of one or several turns, connected to the circuit with a flexible high-frequency cable (Fig. 17).
Resonance indicators allow you to record the state of resonance by the maximum current in the circuit or the maximum voltage on the circuit elements. Current indicators should be low-resistance, and voltage indicators should be high-resistance; then the losses they introduce into the circuit will not cause a noticeable dulling of the resonant characteristics of the circuit.
Rice. 16. Diagram of a resonant frequency meter with a current indicator and replaceable loop coils
Thermoelectric milliammeters with a total deflection current of up to 10 mA are sometimes used as current indicators, connected in series to the frequency meter circuit (Fig. 16); When operating such a frequency meter, you should very carefully establish a connection with the measurement object and avoid overloading the thermal device when approaching resonance. The simplest current indicator can be a miniature incandescent light bulb L; In this case, the measurement error naturally increases.
In modern frequency meters, voltage indicators are most often used - high-frequency voltmeters with dial meters; They provide high indication accuracy with good overload resistance. The simplest such indicator (Fig. 17, a) consists of a point diode D and a sensitive magnetoelectric meter AND, shunted from high-frequency components of the rectified current by capacitor C2. A frequency meter with a dial meter can be used as a field strength indicator when taking radiation patterns of transmitting antennas.
Rice. 17. Circuits of resonant frequency meters with voltage indicators and switchable loop coils
If the oscillations under study are modulated, then a high-impedance telephone T f can serve as an indicator (Fig. 17, a). In this case, resonance is noted by the highest volume of the tone of the modulating frequency. This frequency meter is suitable for auditory quality control of radiotelephone transmitters.
Resonant frequency meters are characterized by sensitivity, i.e., the minimum value of high-frequency power supplied to them, which provides a clear indication of resonance; usually it is in the range of 0.1-5 mW, and when using an incandescent light bulb it increases to 0.1 W. In order to increase sensitivity, a transistor DC amplifier with a high input resistance is sometimes introduced into the resonance indicator (after the detector); The simplest circuit of such an amplifier is shown in Fig. 17, b.
At ultrahigh frequencies, circuits made of elements with lumped constants become ineffective due to a sharp decrease in their quality factor. In the frequency range from 100 to 1000 MHz, fairly good results are achieved in frequency meters with mixed-type circuits having lumped capacitance and distributed inductance (Fig. 18). As an inductance element L0, a curved section (turn) of silver-plated copper wire or tube with a diameter of 2-5 mm is used. Switch B determines the measurement subrange. The frequency meter is adjusted by changing the working length of the inductance coil L0 using a rotary contact slider. The upper limit of the measured frequencies is limited by the value of the installation capacitance C m. Communication with the source of the oscillations under study is carried out through the communication loop L1.
Rice. 18. Scheme of a resonant frequency meter with a mixed-type circuit
In Fig. Figure 19 shows a diagram of a wide-range single-limit frequency meter with an overlap coefficient in the range of 5-10; here the inductance element of the circuit is a metal plate Pl, bent into an arc and connected to the stator St of a variable capacitance capacitor. A slider slides along the plate, mechanically and electrically connected to the rotor Rot of the capacitor. When the rotor is turned, both the capacitance of the circuit and its inductance simultaneously increase (or decrease). Such frequency meters, along with a wide measurement range, have a fairly high quality factor with small dimensions. In the ranges of meter, decimeter and centimeter waves, instruments that use oscillatory systems with distributed constants - sections of transmission lines and volumetric resonators - are used to measure the parameters of electromagnetic oscillations.
Rice. 19. Scheme of a wide-range single-limit resonant microwave frequency meter
To increase the stability of the calibration characteristic, the elements of the frequency meter circuit must have a strong and rigid structure and be made of materials with a low temperature coefficient. The greatest error due to the influence of external factors occurs when measuring the highest frequencies of each subband, when the capacitance of the capacitor C 0 is small. To reduce this error, sometimes the initial capacitance of the circuit is increased by connecting a permanent or tuning capacitor in parallel with capacitor C0 (C1 in Fig. 17, a). At the same time, the frequency overlap coefficient decreases, which helps reduce the frequency measurement error, but at the same time increases the number of required subbands. The measurement error is also reduced if the tuning element is controlled through a vernier device with a slowdown of several tens of times. In industrially manufactured devices, the vernier handle is often equipped with a scale divided into 100 divisions, and on the main scale of the frequency meter setting organ, divisions are applied, marking the number of complete turns of the vernier handle. When both scales are used together, it is possible to obtain several thousand reference points; their corresponding frequencies are determined using tables or graphs.
Adjustment of the frequency meter, excited by a source of frequency oscillations f x , causes a change in the current in its circuit in accordance with the resonance curve of the latter (Fig. 20). The higher the quality factor of the circuit, the sharper its resonance curve and the smaller the possible error in recording the resonance. To achieve a high quality factor, the circuit elements must have low losses, and the connection of the circuit with the resonance indicator and the source under study should be as weak as possible.
The connection with the indicator can be reduced by using, for example, a capacitive voltage divider (Fig. 17, b) with a capacitance ratio C2/C1 >> 1. However, it should be taken into account that weakening the connection with the circuit leads to the need to increase the sensitivity of the indicator or strengthen the connection with the source being studied.
When using a direct-frequency capacitor in a frequency meter, an almost uniform frequency scale can be obtained. Resonant frequency meters are calibrated using standard heterodyne frequency meters, and in the microwave range, measuring lines are used for this. Approximate calibration can be performed by having a measuring generator or transmitter with a smooth frequency range.
Rice. 20. Resonance characteristic of a resonant frequency meter
During measurements, a frequency meter or its coupling element is brought into the radiation zone of the source being studied. By selecting their relative position, a connection is established such that at resonance the indicator needle is approximately in the middle of its scale.
If the sensitivity of the frequency meter is low, it is necessary to strengthen the connection with the source of oscillations; this leads to a flattening of the resonant characteristic of the frequency meter, which makes it difficult to accurately record the resonance state. To reduce possible errors, the two-count method is used. After approximately adjusting the frequency meter to resonance with the measured frequency f x change in capacitance C 0, the circuit is detuned first in one direction and then in the other direction from the resonant frequency until the same indicator reading (I 1-2) is obtained within approximately 50-70% resonant value I m (Fig. 20). Since steep slopes of the resonance curve are used, it is possible to determine the circuit tuning frequencies f 1 and f 2 corresponding to the current with great accuracy. Measured frequency f x = (f 1 + f 2)/2.
If the vibrations under study are non-sinusoidal, then it is possible to adjust the frequency meter to one of the harmonics. In this case, the frequency meter will detect tuning to a number of other frequencies that are multiples of the main oscillation frequency. The latter will be determined as the lowest of the series of resonant frequencies found.
If the E.M.F. induced in the frequency meter circuit is insufficient for the normal operation of the resonance indicator, then the measurement can be performed using the reaction method (absorption, absorption): tuning to resonance is determined by the effect of the frequency meter on the generator mode, from which the measuring circuit absorbs some energy . A fairly strong connection is established between the circuits of the generator and the frequency meter and the setting of the latter is smoothly changed. At resonance, the DC component of the anode (or collector) current of the generator reaches a maximum, and the DC component of the control grid (or base) current drops sharply, which can be detected by connecting a sensitive DC meter to one of these circuits. The frequency meter does not affect the frequency of the generated oscillations, because during resonance it introduces only active resistance into the generator circuit.
A resonant frequency meter is a passive device, since its operation is based on the absorption of energy from the source of the measured frequency. Therefore, it is unsuitable for directly measuring the tuning frequency of radio receivers and isolated oscillating circuits. However, the carrier frequency of the radio station to which the receiver is tuned can be measured quite accurately by the reaction method. To do this, the frequency meter circuit is connected to the antenna circuit of the receiver by means of a coupling coil included in this circuit or by approaching a magnetic antenna. The frequency meter setting is changed until resonance is obtained, which is detected by a sharp drop in the volume of sound signals reproduced by the receiver.
I like the method you proposed, but... In the first case you need an oscilloscope, in the second you need to “assemble a simple circuit.” I have neither one nor the other...
So I found (it seemed to me) a way. which was shown on video on YouTube
I did everything exactly as shown there: I applied a frequency of 50 Hz, connected an AC voltmeter in parallel to the output of the amplifier, measured the current strength with a current clamp, and in one of the wires going to the speaker... I didn’t understand what I got in the end. Current = 1 ampere, voltage - 10 Volts... Why then does the speaker “resonate” to its fullest? I expected to see something in the range of 300 watts there. For example, 6 amperes * 50 volts (parameters approximately corresponding to a resistance of 8 ohms) = 300 watts. This is somehow understandable.
I don’t quite understand your comments about the “coordinated load” - I don’t have enough knowledge...
I read all the manuals, but this does not solve my problem - to determine how much power goes to the speaker system.
I was “delighted” to learn that it can be measured using a Voltmeter and Current Clamps, but... I already wrote how it ended for me
Sorry there's a lot of text
And I need to understand these powers for the next case. When I feed an “unequalized” signal from an amplifier to an acoustic system, no questions arise: the powers of the amplifier and acoustics are comparable and everything is audible even by ear (I imagine 300 watts by ear).
But when I equalize (using a crossover) and “distribute” the signals to different acoustic systems (I remove low frequencies from the Front acoustics - portals and send them to subwoofers), then it is absolutely unclear to the ear what powers went where. At this time, approximately 2.5-3 kW are thundering in the studio.
A particular problem arises in the subwoofer. In general, it is not clear by ear how much power is supplied to it: Subwoofer - 800 Watt, amplifier - 1.5 kW (both at 8 Ohms). So, here it was necessary to measure what exactly goes to the speaker... And here, as you understand, I had problems with which I turned to you.
I hope this clarifies for you the problems that I came to you with.
Thanks in advance
In other words, my question briefly sounds like this:
Is it possible to use a Voltmert and an AC current clamp to measure the power that an amplifier puts out to a speaker? And if possible, how?
Vladimir, your approach is incorrect, that’s why I removed the link to the video. You are trying to measure the power delivered by a reactive load, not a resistive load. And for this it would be necessary to synchronously measure the peak values of current and voltage with a sampling frequency many times higher than the signal frequency. After this, it would be necessary to multiply the values of each resulting pair and calculate the root mean square values from the resulting sequence.
In principle, such devices exist and are inexpensive. Called Wattmeters or Power Meters. They operate on the basis of ADCs and microprocessors capable of performing such calculations. The issue price is about $15.
But, all these budget Power Meters are designed to measure the power of household appliances and are designed to work as an adapter between the network and the load. Their minimum permissible measured voltage is 80-90 Volts. A device capable of operating in a wider range of voltages and signal frequencies will cost an order of magnitude more.
When I was engaged in a similar craft, there was no trace of such devices yet. And I saw a thermal type Wattmeter (they also measured something at that time) only once in my life in one of the laboratories in the city. In addition, in repair practice, the use of an active load is even preferable, since, say, with a power of 2x150 watts, it would be difficult to arrange four-hour bench tests of the amplifier on real speakers.
Is it possible to use a Voltmert and an AC current clamp to measure the power that an amplifier puts out to a speaker? And if possible, how?
I wrote to you above that you would then need to know at what amplitude of the output voltage the signal would begin to be limited. There, voltage squared is a parabola. Even with a small mistake, the result will be very different. In addition, the speaker is a reactive load. Current and voltage are out of phase.
Dear Admin (unfortunately, I don’t know your name).
If I have tired you, you can ignore my message and even delete it. But I really want to understand this issue.
From everything you told me, I can’t understand what’s wrong with the measurements presented in the video and most importantly WHY, when I take exactly the same measurements, I see completely different indicators. By the way, if you use your method No. 2, then in theory they will differ from mine by 1.44 (root of 2) from those that I would see on my voltmeter. But I don’t even see such voltages when I connect an 800-watt subwoofer (in your picture it’s 28 volts). Yes, I'm connecting a speaker, not a resistor. But this cannot change the indicators by an order of magnitude.
I agree that the measurement is not correct (but I’m not looking for absolute accuracy) and the questions remain:
1. I am feeding a 50 Hz sine wave, not a piece of music. Therefore, there is no great need to make measurements with very high sampling.
2. I don't measure peaks. My subwoofer amplifier (1.5 kW) is much more powerful than the subwoofer (800 watts) and is unlikely to start peaking... The speaker on the subwoofer will “fail” first, which is what I actually want to avoid - this is the main goal - to understand approximately what power is flowing to the speaker.
3. I’m trying to understand from the voltage at the output of the amplifier at what volume level it has already given the 600 watts it needs out of its 1.5 kW per channel to the speakers? That is, I do not catch the peaks, but the moment when adding volume is already dangerous for the speaker. For example, go out and not exceed the 600 Watt level. With an 8 ohm speaker (even including reactance) this should be approximately 8 amps and 80 watts. Not 10 watts, which I see during my measurements.
4. I measure amperes not with cheap (as in the video) current clamps, but with those that calculate True RMS (root mean square values). The ammeter, roughly speaking, “doesn’t know” that I’m also measuring voltage. Therefore, it does not matter to him that the current and voltage are out of phase. It should show a current corresponding to 8 or 10 amperes. The question is why I don’t see this current on the device! This is what completely confuses me... And I start looking for those who might know “some secret”
Sorry if you are already tormented by my questions...
Thank you.
Vladimir, it’s not difficult for me to answer and I told you about measurement methods with which you can get a reasonable result.
Logically, when calculating power based on speaker resistance and voltage across it, you should get inflated results compared to real ones. How exactly you measure and what you can take as a reference point is completely unclear to me. In technology, such concepts as “when it is dangerous to increase the volume” cannot be used. At the same time, we do not yet know how accurate the readings of your instruments are.
An AC voltmeter can be checked by measuring the line voltage. Then, you can solder the voltage divider and test the device at other AC voltage measurement limits. Using resistors whose values are known, you can check the ohmmeter and ammeter by making simple calculations. Of course, these are not metrological tests, but at least some kind of verification.
I hope that the sine wave actually reaches the speaker undistorted.
1. You apparently did not understand why that very sampling is needed. When the phases of the current and voltage do not coincide, then only the peak power can be measured in a very short period of time. The shorter this interval, the more accurate the measurement. In the next period, the peak values may change and you need to measure again. For example, when the voltage of a sinusoidal signal reaches its maximum, the current will not be maximum at all due to that same phase shift. Therefore, it is incorrect to make such measurements with conventional instruments.
2. It shouldn't be this way. The amplifier power should not exceed the maximum continuous power of the speakers. But it is even preferable that the speakers be one and a half times more powerful. Moreover, the power values must be in the same units. Nowadays many different terms have been invented that are misleading. It is best to use RMS (Root Mean Square) power.
3. See point 2. Then you can set any power by ear.
4. You are mistaken. Power is the product of current and voltage, so it is very important what current and what voltage affects the load at any given time. It does not matter in cases where the current is constant, or when the phases of alternating current and voltage coincide.
Dear Admin
I agree that there may be some incorrectness in these measurements... In the absence of an oscilloscope, there is no need to talk about accuracy, but... As a person with a “mathematical mind”, I understand that there should be some kind of dependence even with not entirely correct measurements.
Let's try the reverse method
If we take your second method (Measuring the output power of the amplifier using a voltmeter) as a basis, then we can assume some alternative approaches.
Let’s say that I don’t have a “simple scheme” at hand to catch the peaks... Let’s remove this component. According to theory, on the voltmeter in your case on your amplifier, I should see not 28 Volts, but 28/1.41 = 19.9 Volts or something close to that. Right?
As far as I can see, you are very good at theory
What will we see on the volt meter if we replace the resistor with a speaker system with a resistance of 8 ohms? Not a cheap one, but a well-made one, with high damping, which does not imply a radical deviation from its declared characteristics. I think there will definitely be something within the same 19.9 volts (definitely not 10 or 30). We are talking about a 100 Watt amplifier in your case.
Now about my case. I take an amplifier that, according to the passport, at 8 Ohms produces a nominal 1.5 kW. I admit that he can give less, but not by much. This is a fairly powerful and expensive studio amplifier. I connect a subwoofer to it (800 watts at the same 8 Ohms). I feed a 50 Hz sine wave to the amplifier input and turn the volume level up to half. I understand that the sound cannot be described in words, but I act as an analyst: the sound increases (by ear) evenly. Somewhere in the middle (I think it’s within 500-600 watts) the windows in the studio begin to shake, the big drums “bounce”, the microphones jump on the table. This is what I called a strange term for you “when you turn up the volume” dangerous,” meaning that the speaker may already be damaged... But let’s discard this lyric for the sake of the purity of the experiment...
So, a practical experiment: 50 Hertz, half the volume of a 1.5 kW amplifier, 800 watt speaker and a voltmeter connected to the amplifier output (or speaker terminals). How many volts will the volt meter THEORETICALLY show?
Perhaps this is incorrect, perhaps not entirely clear, but it will EXACTLY show some stable number (as in your case, 28 volts is frozen on the screen).
Maybe it seems to me, but this number, in the absence of an oscilloscope and other capabilities, will help me VERY ROUGHLY understand what is happening in speaker systems.
Question: within what limits should this number be in theory?
THANK YOU
P.S. Checking your measuring instruments is a very sound idea. It immediately came to mind. I borrowed other devices from friends, checked the indicators, etc. No anomalies were detected. It’s a pity no one has an oscilloscope. But I continue to search...
Vladimir, I am not against your approach and agree that the relative power in the load can be measured using a voltmeter. It is on this principle that the operation of overload indicators is based, which are often built into household speaker systems to prevent their failure when connecting a ULF of unknown power. But these indicators are designed to work with a specific load. You can also build a table of correspondences, where in one column there will be the power values obtained by the method specified in the article, and in the other - the corresponding voltage values at a specific speaker. But for this you need to have a certain reference point in metrological terms.
I have already written about your experiments above and I can only repeat where you need to start.
1. Check the voltmeter. (Have you checked the accuracy of the voltmeter readings?)
2. Check the ohmmeter. (Have you checked the accuracy of the ohmmeter reading?)
3. Measure the speaker resistance. For example, two four-volume coils could be connected in parallel rather than in series, that is, there are not 8 ohms, but only 2 ohms. (Did you take this measurement?)
4. Measure the voltage on the speaker at different positions of the volume control.
6. At 500 Watts of power, you should get, at an 8 Ohm active load, an effective voltage value of about:
U = √(P*R) = √(500*8) ≈ 63(Volts RMS)
On a reactive load of 8 Ohms, in theory, it should be a little more, maybe 70 or 80 Volts RMS. But I have not conducted such comparative experiments.
And one last thing. There are no miracles. Our professor Preobrazhensky proved this. If you are sure, say, that the power is enormous, but the output voltage is too low, then an error has crept in somewhere, either in the measurement and calculation methodology, or in the operation of the measuring equipment. Ohm's law usually helps to understand where the error is hidden.
1. — I checked. At the amplifier input, it honestly shows 223 Volts. For a change, I poked it into other devices and connected another volt meter nearby. No anomalies were detected.
2. - I checked. The current clamps (Uni-T UT204) made a slight mistake at milliampere measurements, but at higher currents (from 0.5 amperes) they work like clockwork. I connected a “regular” volt/ampere meter (up to 10A) nearby - it shows the same thing. Actually, I took current clamps, taking into account the fact that the current in the cable could theoretically be >10A, but I didn’t find it there
3. — I looked at the characteristics and installed the subwoofer. . There is one subwoofer speaker without filters based on the MAG 1880 speaker. The power of 800 watts was apparently stated for the subwoofer version. The "rest" resistance is 6 ohms. They state 8, apparently also for the “active” state? But all this would not make significant (many times!) changes in the measurements...
4. - tried...
it was 10.6 volts (and 1 amp) at mid-volume. The audible volume did not fit in my head with 10-15 watts
On a reactive load of 8 Ohms, in theory, it should be a little more, maybe 70 or 80 Volts RMS.
But I have not conducted such comparative experiments.
There are no miracles... Ohm's law usually helps to understand where the error is hidden...
- Or miracles do happen, but Oma’s zak somehow didn’t help me
I really thought that some “electro-wizard” would tell me - “Dude, you need to immediately apply such and such a coefficient and multiply everything by 7!!!”... But, alas... There are no such significant coefficients that you and confirmed Thank you. Ohm's law should, even with the errors of my measurements, still correspond to the declared powers...
Well. or 12 watts, or rather 10 volts and 1 ampere supplied to the speaker - this is VERY loud!
Then I can’t even imagine what the speaker should produce if 50 volts are applied to it
In any case, THANK YOU for the ideas, algorithm and TIME...
In the near future I will try again to get to the studio with the instruments and try everything there again
Vladimir, maybe you made a mistake with the order of the number on the instrument scale. To measure the ULF output, did you use the same measurement range as when measuring the network voltage?
In the end, ask Klyachin, since you have already reached him. He is a guru in these things and has probably encountered various measurement errors.
Measurement methods. Frequency measurement is carried out by comparing it with the frequency of the frequency-setting process, taken as a unit (the frequency-setting process can be a reference, exemplary or working, depending on the measure that reproduces it). This type of measurement is one of the important tasks of measuring technology. In electronics, radio engineering, automation and other related industries, signals of a wide variety of frequencies are used - from fractions of a hertz to thousands of GHz.
There are analogue and digital methods of frequency measurement. Analog method is an indirect measurement method based on comparing the measured frequency with the frequency of another source (usually a reference one) using an oscilloscope, heterodyne and resonant methods.
For comparison, it is necessary to have a reference generator, the accuracy of which is at least 5 times higher than the accuracy of the controlled source, and a device for frequency comparison. Often such a device is an oscilloscope.
To measure frequencies that are multiples of a known frequency, use Lissajous figure method. A voltage of known frequency frev of the reference source is applied to one input of the oscilloscope (for example, input X) , and the voltage of the measured frequency fmeas - on the second (for example, input Y). The frequency of the reference generator is adjusted until a stable image of the simplest interference figure is obtained on the screen: a straight line, a circle or an ellipse. The appearance of one of these figures indicates equality of frequencies (ratio fmeas:frev = 1:1). When the frequencies are not equal to each other, but are multiples, more complex figures are observed on the oscilloscope screen.
The frequency ratio is determined in the following way. Two straight lines are mentally drawn through the image of the figure: horizontal and vertical. Number ratio T intersections of a horizontal line with a figure to a number P intersection of a vertical line with a figure is equal to the ratio of the frequency supplied to the input of channel Y to the frequency supplied to the input of channel X:
Rice. 3.1
If the frequencies being compared are multiples, but their ratio is large, use circular scanning method with brightness modulation. A reference voltage frev is applied simultaneously to both inputs of the oscilloscope with a phase shift of 90°, achieved using a phase shifter. The gain of both channels is adjusted so that the beam draws a circle on the screen. The voltage of the measured frequency is supplied to the brightness control channel. The frequency of the reference source is adjusted until a stationary image of a dashed circle is obtained on the screen (Fig. 3. 1). The number of bright arcs or dark spaces between them uniquely determines the ratio N = fmeas / frev (7:1 in Fig. 3. 1).
If the ratio of the frequencies fmeas and frev differs slightly from an integer, i.e. fmeas = Nfrev Fp (frequency Fp is relatively small), then the figure rotates, and the direction of rotation shows the sign of the frequency divergence (it is easiest to determine experimentally, fixing the direction of rotation for a certain established relationships f 'meas > Nfo6p and f 'meas > Nfo6p). The degree of discrepancy (and the resulting frequency measurement error) can be determined as follows: count the number d arcs passing through a certain radial line on the screen in a fixed period of time. Then the discrepancy Fp = d /t.
Digital method(discrete counting method) occupies a dominant position in modern measuring technology. It has many advantages: a very wide range of frequencies that can be measured with one instrument (for example, from 10 Hz to 32 GHz); high measurement accuracy; receiving a reading in digital form; the ability to process measurement results using a computer, etc.
Rice. 3.2
The problem of measuring frequency using a digital method is the inverse of the problem of measuring period. If, when measuring a period, the time interval t x = Tx was filled with timestamps T 0 , then when measuring frequency the reference time interval T 0 is filled with pulses with a period T x = 1/ f x . To do this, the signal under study is converted into a periodic sequence of short pulses, the moments of their appearance correspond to the moments of transition of the sinusoidal signal through the zero level with a derivative of the same sign. Thus, the pulse repetition period is equal to the period of the signal under study. From two adjacent reference frequency pulses that are separated by a time interval T 0 , a strobe pulse is generated - a temporary gate with a duration t = T 0 . Number of pulses entering the gate P = t/T x . Obviously, the desired frequency will be determined from the relation fx = p/t.
The measurements turn out to be indirect. To get direct readings, in frequency meters. Constructed according to a circuit with hard logic (without a microprocessor), the duration of temporary gates is set t = With, where p = 0; ±1; ±2; . . . (on the instrument panel the gate duration switch is indicated by the inscription MEASUREMENT TIME). At p=0 (t = 1c) fx = n Hz;
If t== 1ms, then fx == P kHz.
Digital frequency meter. Modern digital frequency meters are multifunctional devices. They measure the frequency of sinusoidal and pulse signals, the repetition period of signals, the duration of pulses, time intervals specified by two pulses from the same or different sources, frequency variation, the ratio of two frequencies; They count the number of pulses received at the input, etc. Shown in Fig. 3. 3 block diagram refers to frequency measurement mode. The operation of the circuit is as follows.
A periodic signal, the frequency of which needs to be measured, is supplied to the input of the device (usually designated by the letter A). After amplification or attenuation in the input block, the signal is fed to the shaper, where it is converted into a periodic sequence of pulses with a repetition rate f x . These pulses are supplied to input 1 of the time selector and pass through it to the counter, if at the input 2 The selector has a strobe pulse. The strobe pulse is generated from the voltage of a high-frequency quartz oscillator. Since the period of its output signal is small, to obtain the required duration of the strobe pulse, a frequency divider is provided in the circuit (on the front panel of the device it is designated as a PERIOD MULTIPLIER). The divider is a set of decades, each of which reduces the pulse repetition rate by 10 times. Division ratio q depends on the number of included decades. From a periodic sequence of pulses generated at the output of the divider, the automation unit (time gate circuit) generates a strobe pulse (time gate) with a duration t == T 0 , supplied to the input 2 time selector and determining the duration of the count.
Rice. 3.3
Let's consider the process frequency ratio measurements Fx1 / Fx2 (Fx1 >Fx2). The higher frequency Fx1 is supplied to the input of the frequency meter, and the lower frequency Fx2 is supplied through an additional shaper to the automation unit (in this case, the quartz oscillator and divider are turned off). Pulses with period Tx1 during period Tx2 pass through the time selector and are counted. Number of pulses m = Tx2 / Tx1 = Fx1 / Fx2 . To improve measurement accuracy, frequency Fx2 supplied through a divider (only the crystal oscillator is turned off).
Errors in frequency measurements are similar to those considered in the analysis of time interval measurements.
