JP4389344B2 - Analog isolation circuit - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an analog isolation circuit. Specifically, when floating measurement is performed to measure the potential difference between two points that are not ground potentials, the direct current is changed from DC to several hundred megahertz (MHz) to insulate the measurement object from the grounded voltage measuring device. An improved isolation circuit that isolates a wide range of analog signals. In particular, an isolation circuit suitable for isolating an input signal of a digital oscilloscope (DSO) is provided.
[0002]
[Prior art]
Floating measurement that measures the potential difference between two points that are not grounded is often required in the fields of motors, power supplies, and the like. However, this floating measurement point is often at a high potential, and it is not easy to measure the potential difference between two points of the floating measurement object that is not completely and accurately grounded. In particular, accurate measurement is difficult when a small differential signal is superimposed on a large voltage amplitude of the same phase or when the signal under measurement is a high frequency.
[0003]
Such a problem often occurs during measurement with an oscilloscope. Conventionally, the oscilloscope is not grounded but is measured in a so-called floating state. However, the following problems arise when an oscilloscope is not installed.
[0004]
The first problem is that the outer metal part of the measuring device has the same potential as the object to be measured, which may cause an electric shock.
[0005]
The second problem is that the oscilloscope ground capacitance causes linking in the measured signal waveform, or the measured point that is floating in direct current and / or high frequency at the moment the probe touches the measured point. As a result of being instantaneously grounded through, there is a risk of damaging the circuit under test on the floating side and the oscilloscope that is the measuring instrument.
[0006]
In order to prevent such inconvenience, it is necessary to ground the oscilloscope and to electrically insulate the oscilloscope from the measurement point on the floating side by an analog insulation circuit. In order to solve such a problem, the present applicant has disclosed an analog isolation circuit that performs balanced modulation and demodulation with a high-frequency sine wave signal in the following document.
[0007]
Reference 1. Japanese Patent Laid-Open No. 11-72515 "Broadband analog isolation circuit"
[0008]
FIG. 9 is a block diagram of the circuit disclosed in Document 1, FIG. 10 is a circuit diagram of a Gilbird multiplication circuit used in FIG. 9, and FIG. 11 is a waveform diagram of each part of FIGS. It is shown.
[0009]
Input signals to be measured for floating are applied to the input terminals 5a and 5b in FIG. The input signal is input to a multiplication circuit 20 that operates as a modulator via the buffer amplifier 7. Here, the analog input signal is modulated by the carrier signal (sine wave) from the insulating circuit 12. The sine wave output from the oscillator 18 passes through the insulation circuit 12 and becomes a carrier signal of the multiplication circuit 20. The multiplication circuit 20 multiplies the input signal and the carrier signal and outputs the result.
[0010]
The output of the multiplication circuit 20 is input to the multiplication circuit 60 that operates as a demodulator through the insulation circuit 11. The carrier signal output from the oscillator 18 becomes a demodulation carrier signal of the multiplication circuit 60. The multiplication circuit 60 multiplies the output of the insulating circuit 11 and the demodulation carrier signal and outputs the result.
[0011]
FIG. 10 shows a specific circuit example of the multiplication circuit 20 on the floating side in FIG. 9 and the multiplication circuit 60 on the non-floating side which is the same circuit as that. This is known as a Gilbert type multiplication circuit and can be easily integrated into an IC.
[0012]
In FIG. 10, differential input terminals 1a and 1b are respectively connected to the bases of transistors 31 and 32 in a differential current amplification stage, and resistors 41 and 42 having a negative feedback function between the emitters of transistors 31 and 32 are provided. Is connected. The middle point of the resistors 41 and 42 has a negative power source V on the floating side._EEA constant current source connected to is connected.
[0013]
This constant current source is composed of transistors 37, 38 and resistors 45, 46, 47. Resistors 45, 46 and 47 determine the current value of the current mirror. Transistors 37 and 38 constitute a current mirror. Since the cascode-connected transistors 33, 34, 35, and 36 are required to have the same characteristics, the Gilbert multiplication circuit 20 is preferably realized by a monolithic IC. For example, an NEC transistor array μPA101G is formed. Use.
[0014]
The emitter of the transistors 33 and 35 is cascode-connected to the collector of the transistor 31, and the emitters of the transistors 34 and 36 are cascode-connected to the collector of the transistor 32. The bases of the transistors 33 and 35 are connected to the bases of the transistors 34 and 36, respectively.
[0015]
The collectors of the transistors 33 and 34 are connected to the collectors of the transistors 36 and 35, respectively, the load resistors 43 are connected to the collectors of the transistors 34 and 35, and the load resistors 44 are connected to the collectors of the transistors 33 and 36. The middle point of the load resistors 43 and 44 is the positive power supply V on the floating side._CCIt is connected to the. The transistors 31 to 38 form a Gilbert type multiplication circuit 20 on the floating side. The Gilbert type multiplication circuit 20 outputs the result (product) of multiplication of the signal voltage between the terminals 1a and 1b and the high-frequency sine wave voltage between the terminals 3a and 3b.
[0016]
The collectors of the transistors 34 and 35 and the collectors of the transistors 33 and 36 are connected to the insulating circuit 11 via terminals 4a and 4b, respectively.
[0017]
Since the output (terminal 9) of the multiplication circuit 60 includes noise having a frequency component twice that of the carrier signal, it is input to the low-pass filter 67 and removes noise having a frequency component twice that of the carrier signal. An output signal from the low-pass filter 67 is obtained at the output terminal 10 through the buffer amplifier 69. Since the input side and the output side are insulated by the insulation circuits 11 and 12, analog signals can be insulated.
[0018]
Here, when an insulating transformer is used as the insulating circuits 11 and 12, a capacitor is used, a radio wave is transmitted between opposing antennas, an optical fiber is inserted and light is transmitted and received oppositely, a space is inserted. Reference 1 discloses a case in which light is transmitted oppositely and received.
[0019]
FIG. 11 shows an example of simulation when the input signal in the circuits shown in FIGS. 9 and 10 is a step waveform, the frequency of the oscillator 18 is 1 GHz, and the cutoff frequency of the low-pass filter 67 is 100 MHz. The time axis is graduated at intervals of 5 ns.
[0020]
10A is a differential input waveform between the input terminals 1a and 1b in FIG. 10, FIG. 10B is a waveform between the terminals 4a and 4b for obtaining the output of the multiplication circuit 20 in FIG. The waveform of the terminal 9 that obtains the output is shown, and (d) is the output waveform of the output terminal 10. As can be seen from FIG. 11, the analog signal is not distorted from DC to high frequency while insulating the floating side (left side of the isolation transformers 11 and 12 in FIG. 9) from the non-floating side (right side of the isolation transformers 11 and 12 in FIG. 9). It can be seen that can be obtained.
[0021]
In the analog isolation circuit based on the synchronous detection using the sine wave as described above, since the multiplication is performed by the carrier signal of the sine wave at the time of modulation and demodulation using the multiplication circuits 20 and 60, the multiplication circuit that operates as a demodulator. It is necessary to remove twice the frequency of the carrier signal from the output from 60. The removal of the carrier signal will be described next.
[0022]
The differential input signal applied between the terminals 1a and 1b is E_A, A high frequency sine wave signal which is a carrier signal applied between the terminals 3a and 3b is represented by E_B, The result of multiplication on the floating side between terminals 4a and 4b is E_CThen there is the following relationship.
[0023]
E_A= F (t) (1)
E_B= Sin (ωt) (2)
E_C= E_A× E_B= F (t) sin (ωt) (3)
[0024]
The signal E of the formula (3) obtained from the insulating circuit 11_CIs input to the multiplication circuit 60. On the other hand, the high frequency sine wave signal which is the carrier signal of the expression (2) output from the oscillator 18 supplies a high frequency sine wave signal whose polarity is inverted to the multiplication circuit 60 to the multiplication circuit 60. Similarly to the multiplication circuit 20, the multiplication circuit 60 outputs the multiplication result of the high-frequency sine wave signal, which is a carrier signal, to the terminal 9 (FIG. 11 (c)).
[0025]
Output signal E of multiplication circuit 60 obtained at terminal 9_D= E_C× E_BIs expressed as:
E_D= F (t) sin²(Ωt) = (1/2) f (t) (1-cos (2ωt)) (4)
[0026]
When the frequency component of the input signal f (t) is lower than the frequency of the high-frequency sine wave signal sin (ωt), the cutoff frequency of the low-pass filter 67 is set lower than 2ωt, so that the carrier frequency of Expression (4) Since the term of cos (2ωt), which is a double component, becomes 0, only the term (1/2) f (t) remains, and this is obtained at the output terminal 10 to reproduce the input signal (FIG. 11). (D)). Although the input terminals 5a and 5b and the output terminal 10 are completely insulated by the insulating transformer 11, analog signal insulation over a wide band from DC to high frequency is possible.
[0027]
Here, it is necessary to set the cutoff frequency of the low-pass filter 67 to a value that can sufficiently cut off the sinusoidal carrier signal. For example, the case where it is assumed to be configured with a Gaussian filter used in a time measuring device such as an oscilloscope is shown below. The attenuation characteristic H (ω) of the amplitude of the Gaussian filter is expressed by the cutoff frequency f_cIs expressed by the following formula.
H (ω) = exp (0.00878 ω²/ F_c ²(5)
[0028]
The cutoff frequency f of the low-pass filter 67 depends on how much the leakage of the carrier frequency ω is reduced from this H (ω)._cIs decided. Actually, the cut-off frequency f of the low-pass filter 67_cDetermines the band of the analog insulation circuit, so the cutoff frequency of the low-pass filter 67 (band of the analog insulation circuit) f_cThe carrier frequency ω is determined depending on how much the carrier frequency ω leaks.
[0029]
As a result of the experiment, when the bandwidth is 100 MHz, the carrier frequency can be practically about 1 GHz. This is because when the carrier frequency (1 GHz) is reached, the frequency characteristics of each circuit and the oscilloscope band to be observed are limited, so the oscilloscope itself acts as a low-pass filter. Need not follow the relational expression expressed by the equation (5). However, in order to obtain a sufficient attenuation characteristic, a carrier signal having a high frequency about 10 times the bandwidth of the oscilloscope is required. In order to realize an analog isolation circuit with a bandwidth of 100 MHz, a carrier signal of about 1 GHz is required, and each circuit requires a broadband configuration. In addition, when this frequency is reached, new problems such as emission of electromagnetic waves occur.
[0030]
Other analog insulation circuits include those using photocouplers. However, the photocoupler can isolate an analog signal with high accuracy with a bandwidth of 1 MHz at most, which is insufficient for high-speed waveform observation. A broadband analog isolation circuit using a photo coupler and an isolation transformer is proposed in the following document.
[0031]
Reference 2. US Patent 5,517,154 “SPLITPATH LINEAR ISOLATION CIRCUIT APPARATUS AND METHOD”
[0032]
The technique disclosed in Document 2 uses a cross-insulation so that the low frequency range of about DC to 100 kHz is insulated with a photocoupler, and the high range of about 100 kHz to 100 MHz is insulated with an insulating transformer to flatten the overall frequency characteristics. The analog signal up to about DC to 100 MHz is insulated by making it over.
[0033]
The analog isolation circuit disclosed in Document 2 has the following problems.
[0034]
The first problem is that since the bandwidth of the photocoupler is 1 MHz at most, the crossover frequency can only be increased to about 100 kHz, and the low-frequency cutoff frequency of the insulating transformer cannot be increased. Therefore, there is a problem that the high-frequency characteristics are generally not good because the insulating transformer is not downsized and the insulating transformer that can be used up to a low frequency is large. Therefore, it is difficult to further widen the frequency bandwidth.
[0035]
The second problem is that the low and high frequencies are isolated and crossed over separately, so that crossover distortion occurs and that precise adjustment of the low and high frequency gains is required. It is.
[0036]
[Problems to be solved by the invention]
In the analog isolation circuit based on the synchronous detection method that modulates at the carrier frequency of the sine wave disclosed in Document 1, the carrier signal frequency is the same as that of the analog isolation circuit in order to remove the double carrier signal from the output by a low-pass filter or the like. More than 10 times the bandwidth is required. Therefore, there is a problem that the analog insulation circuit requires a wide-band circuit configuration, and the wiring or the like needs to use a transmission line capable of transmitting a wide-band signal.
[0037]
In addition, when a pulse transformer is used as the insulating means, the size can be reduced. However, it is difficult to maintain the breakdown voltage at this frequency. If a structure that ensures a withstand voltage is used, the coupling coefficient becomes small, and the transmission efficiency of the pulse transformer is sacrificed. Furthermore, a high frequency of about 1 GHz causes a new problem such as the emission of electromagnetic waves, and there arises a problem that countermeasures such as shielding are required for these problems.
[0038]
Even in the analog insulation circuit disclosed in Document 2, since the band of the photo coupler is at most about 100 kHz, the low-frequency cutoff frequency of the insulation transformer cannot be increased. Therefore, the insulation transformer cannot be reduced in size. In addition, an insulating transformer that can be used up to a low frequency has a problem that high-frequency characteristics are not good. Since the low frequency region and the high frequency region are separately insulated, there is a problem that the circuit configuration is complicated and gain adjustment is required.
[0039]
The present invention has been made in view of such an unsolved problem, and provides a wide-band analog insulation circuit that can be insulated without a waveform distortion from a direct current to a high frequency with a simple configuration without using a high frequency carrier signal. To do.
[0040]
[Means for Solving the Problems]
  A multiplication circuit that multiplies the floating-side analog input signal and the floating-side high-frequency sine wave signal to obtain a multiplication result;
  A first insulation circuit for obtaining a multiplication result as a non-floating multiplication result on a non-floating side by insulating the multiplication in a DC manner;
  A sampling circuit for sampling a non-floating multiplication result with a clock and restoring an analog input signal to obtain a non-floating signal;
  An oscillator that oscillates a non-floating high-frequency sine wave signal;
  Non-floating side high frequency sine wave signalPeak point ofA clock circuit for obtaining a clock synchronized with
  A non-floating high-frequency sine wave signal is galvanically isolated and a second insulating circuit for applying the floating high-frequency sine wave signal to the multiplication circuit is provided.
[0041]
  Since the synchronous detection method of Document 1 performs modulation and demodulation with a high-frequency sine wave, a filter having a frequency twice that of the carrier signal is required. On the other hand, the analog isolation circuit of the present invention modulates at a sinusoidal high frequency, but demodulates using a sampling circuit. This eliminates the need for a low-pass filter that removes the carrier signal. The sinusoidal frequency as the carrier signal does not need to be sufficiently high with respect to the band of the insulating circuit. Therefore, each circuitBroadbandThere is an advantage that no configuration is required. The pulse transformer has the advantage that it can be used for a commercially available high-speed LAN.
[0042]
  Furthermore, the mainstream of oscilloscopes has recently shifted to a digital storage oscilloscope (DSO), which incorporates an AD converter and an oscillator for a sampling clock of the AD converter. When the present invention is used for DSO input isolation, the DSO AD converter and oscillator can be used as the sampling means and the sine wave oscillator in the present invention, respectively. Therefore, by adding two sets of insulation means and one multiplication circuit, there is an advantage that DSO input insulation becomes possible.In literature 1, two broadband multiplication circuits are required, whereas in the present invention, only one is required, and the broadband property as in literature 1 is not required.
[0043]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 shows an embodiment of the present invention. Components corresponding to those in FIG. 9 are given the same symbols. The analog signals input to the floating side input terminals 5 a and 5 b are connected to the buffer amplifier 7. The output of the buffer amplifier 7 is connected to the multiplication circuit 20. The multiplication circuit 20 is the same as that shown in FIG. The multiplication circuit 20 is connected to an oscillator that oscillates a sine wave through an insulation circuit 12.
[0044]
The multiplication circuit 20 is connected to the insulation circuit 11. The insulation circuit 11 is connected to the AD converter 51. The output of the oscillator 18 is connected to the clock circuit 15. The clock that is the output of the clock circuit 15 is applied to the A / D converter 51 and the D / A converter 52. The digital output of the A / D converter 51 is connected to the D / A converter 52.
[0045]
The insulating circuits 11 and 12 insulate the floating side from the non-floating side. The floating side includes a buffer amplifier 7 and a multiplication circuit 20 for buffering input signals from the input terminals 5a and 5b. The non-floating side includes an oscillator 18, a clock circuit 15, an A / D converter 51, and a D / A converter 52 connected to the output terminal 10. Insulating circuits 11 and 12 are, for example, insulating transformers.
[0046]
Each frequency of the analog input signal f (t) applied between the input terminals 5a and 5b and the sine wave (carrier) output from the oscillator 18 is ω. The analog input signal f (t) is input to the multiplication circuit 20 via the buffer amplifier 7. A sine wave (carrier) from the oscillator 18 is input to the multiplication circuit 20 via the insulation circuit 12. Here, the analog input signal is modulated by a sine wave (carrier) from the insulating circuit 12. The result of multiplying the analog input signal by the sine wave carrier is the same as the modulation signal obtained by modulating the input signal with the carrier.
[0047]
The multiplication circuit 20 is a modulation signal obtained by multiplying the analog input signal f (t) by the carrier signal sin (ωt) from the insulation circuit 12.
E_C= E_A× E_B= F (t) sin (ωt)
Is output. This is the E shown in equation (3)._CIt is. A modulation signal output as a result of multiplication from the multiplication circuit 20 is input to the A / D converter 51 via the insulation circuit 11.
[0048]
2A and 2B show the input and output waveforms of the clock circuit 15, respectively. The sine wave of (a) is input to the clock circuit 15 from the oscillator 18, and the clock circuit 15 outputs the clock of (b) that drives the A / D converter 51 and the D / A converter 52.
[0049]
  The clock timing is the same as that of (a) from the oscillator 18.1 period 2πAgainst sine wave1/4 period, ie π / 2 onlyRunning late. The A / D converter 51 performs A / D conversion at the rising edge of the clock (b) from the clock circuit 15, that is, at the peak of the sine wave. The modulation signal input to the A / D converter 51 is expressed by Equation (3).
        E_C= F (t) sin (ωt)
It is represented by Here, the A / D converter 51 always responds to the sine wave from the oscillator 18.π / 2A / D conversion is performed with a delay of, that is, sin (ωt) = 1.
[0050]
  The sampling timing of the A / D converter 51 is
        Tsm = (2π / ω) m +π / 2
It can be expressed as. Here, m is an integer. Since sin (ωt) = 1, the output V of the A / D converter 51_omIs
    V_om= F (T_sm) sin (T_sm) = F (T_sm)
This is equivalent to A / D conversion of the input signal applied between the input terminals 5a and 5b, despite being modulated by the carrier signal sin (ωt).
[0051]
In FIG. 3, the output corresponding to the input of the signal to the multiplication circuit 20 is shown in an arbitrary scale. When the carrier frequency of the output from the multiplication circuit 20 is 100 MHz, the rise time of the analog input signal f (t) to the multiplication circuit 20 is a pulse waveform of 40 ns. Since the A / D converter 51 performs A / D conversion at the rising edge of the clock, it can be seen that the input waveform is reproduced.
[0052]
The present invention reproduces an input signal from a modulation signal modulated with a carrier signal by sampling a modulation signal obtained by modulating the input signal with a carrier signal by an A / D converter 51 in synchronization with the carrier signal. ing. Here, it largely depends on the characteristics of the A / D converter used. In general, the A / D converter has a resolution at which A / D conversion can be performed depending on the input signal (signal change rate). It is known that the A / D converter has a resolution that can be A / D-converted as the slew rate (change rate) of the input signal increases.
[0053]
  The slew rate of the modulation signal input to the A / D converter 51 in the present invention is the modulation signal V input to the A / D converter 51._oIt is obtained by differentiating (t). V_oSlew rate V 'differentiated from (t)_o(t) is, F (t) derivative is f '(t)It is represented by Formula (6).
    V '_o(t) =f ′ (t) sin (ωt) + ω · f (t) cos (ωt)
(6)
Here, as described above, the timing at which the A / D converter 51 performs A / D conversion is as follows.
    T_sm= (2π / ω) m +π / 2
Every. Sin (T_sm) =sin (ωt) = 1, cos (T _sm ) = Cos (ωt) = 0It is.
[0054]
  From this, the slew rate V ′ viewed from the A / D converter 51_o(t) isConsidering sin (ωt) = 1, cos (ωt) = 0,
    V '_o(t) =f ′ (t) sin (ωt) + ω · f (t) cos (ωt)= F '(t)
Thus, it can be seen that the slew rate of the modulation signal is the same as the slew rate f ′ (t) of the input signal f (t). However, in practice, the time interval at which the A / D converter 51 samples the input signal is not infinitely small but is a finite value, and the influence of this sampling time interval can be considered.
[0055]
In FIG. 4, the input and output of signals to the multiplication circuit 20 are shown in an arbitrary memory. Assuming that the input frequency of the signal to the multiplication circuit 20 is a sine wave of 25 MHz and an amplitude of 0.5 V, and the frequency of the carrier included in the output of the multiplication circuit 20 is 100 MHz, the memory on the horizontal axis is 12.5 ns. is there. An output from the multiplication circuit 20 is applied to the A / D converter 51 via the insulation circuit 11.
[0056]
  As the characteristics of the A / D converter 51 to be used, it is assumed that the resolution is 8 bits, the input signal is a sine wave having a frequency of 25 MHz and an amplitude of 0.5 V, the sampling frequency is 100 MHz, and the conversion accuracy is 6 bits. Signal input at this time (25MHzThe maximum slew rate of the sine wave is 0.0785 V / ns at the time of zero crossing (A portion circled by a broken line in FIG. 4) at which the rate of change of the sine wave is maximum.
[0057]
Here, since the resolution is assumed to be 6 bits, 1 LSB (6 bits) is about 0.0156 V. Therefore, at least, the A / D converter 51 can perform A / D conversion within the time of (0.0156 / 0.0785) ns = 198 ps. The A / D converter 51 can perform sampling and A / D conversion within about 198 ps. A portion A circled by a broken line indicating the maximum slew rate portion of a waveform obtained by modulating the input frequency with a carrier of 25 MHz and the input with a frequency of 100 MHz will be described.
[0058]
FIG. 5 shows an enlarged waveform of a portion A circled by a broken line in FIG. It can be seen that the slew rate of the input (solid line) and the output (dashed line) of the signal to the multiplication circuit 20 is almost the same at the input and output in the range of 200 ps near the zero cross. When the A / D converter 51 is used, the slew rate seen from the A / D converter 51 is the same even if it is modulated by the carrier, so that the resolution of the A / D converter 51 is not deteriorated. Yes.
[0059]
FIG. 6 shows the relationship between the signal input and output of the multiplication circuit 20. This is a case where the signal input DC is 0.5V. The carrier frequency at which the output of the multiplication circuit 20 is received via the insulation circuit 11 and sampled by the A / D converter 51 is 100 MHz. Since it is carrier-modulated, it cannot be seen from the A / D converter 51 as DC, but since sampling is performed at each peak point of the carrier, the sampled point is equivalent to the case where direct current is sampled.
[0060]
FIG. 7 shows an enlarged waveform diagram of a portion B circled by a broken line in FIG. Since the time for A / D conversion is about 198 ps, the signal change during this period is 0.001 V, which is about 0.3 LSB in terms of LSB (8 bits), and an 8-bit resolution can be obtained. As described above, the output does not deteriorate due to the performance of the A / D converter 51.
[0061]
In the above description, from the relationship between the sampling frequency and band of the DSO generally called, as an example, the sampling frequency is 100 MHz, and the band is ¼ of 25 MHz. The sampling frequency is 100 MHz. Sometimes the bandwidth is not limited to 25 MHz or less. It will be apparent from the above description that the band is not limited to 25 MHz if the present invention is applied to a DSO capable of equivalent sampling.
[0062]
The digital output of the A / D converter 51 is input to the D / A converter 52, converted into an analog signal, and output to the output terminal 10. Since the output from the D / A converter 52 includes a high frequency that is three times or more the sampling carrier frequency, it is necessary to insert a low-pass filter to filter the high frequency if necessary.
[0063]
In the case of the low-pass filter 67 shown in FIG. 9 showing the conventional example, the object of removal is a frequency component twice that of the carrier, whereas in the case of the present application shown in FIG. Since it is a target, the effect on the signal to be restored is very small.
[0064]
When the present invention is applied to a digital oscilloscope, the signal is processed and displayed in a digital format, so that the D / A converter 52 and the low-pass filter are not necessary. The A / D converter 51, the oscillator 18, and the clock circuit 15 that are used in a digital oscilloscope (DSO) can also be used. It is possible to insulate the input side of the digital oscilloscope (DSO) only by adding the insulation circuits 11 and 12 and the multiplication circuit 20.
[0065]
FIG. 8 shows an embodiment of the present invention. Components corresponding to those in FIGS. 1 and 9 are given the same symbols. In the embodiment of FIG. 8, TDK-made insulating pulse transformer TLA-6T207 for high-speed LAN is used as the insulating circuits 11 and 12. The loss is 1 dB or less from 100 kHz to 100 MHz. The −3 dB cutoff frequency is about 100 MHz. When this insulating transformer is used, a modulation signal (carrier) from 100 kHz to 200 MHz can be used.
[0066]
The analog signal input between the input terminals 5a and 5b is connected to the buffer amplifier 7. The output of the buffer amplifier 7 is connected to the multiplication circuit 20. The multiplication circuit 20 is applied with a sine wave (carrier) from an oscillator 18 that oscillates a sine wave through the insulation circuit 12. The output of the multiplication circuit 20 is connected to the sample and hold (SH) circuit 53 through the insulation circuit 11. The output of the oscillator 18 is applied to the clock circuit 15B. The clock circuit 15B applies a sampling clock to the sample and hold (SH) circuit. The output of the sample and hold (SH) circuit 53 is connected to the output terminal 10 via a low-pass filter 54.
[0067]
The difference from the configuration of FIG. 1 is that a sample and hold (SH) circuit 53 and a low-pass filter 54 are provided in place of the A / D converter 51 and the D / A converter 52. The operation of this part will be described.
[0068]
Modulation signal waveform V applied to the sample and hold (SH) circuit 53 via the insulating circuit 11_oBecomes the input signal f (t) modulated to the carrier signal sin (ωt). In other words, the sample and hold circuit (SH) 53 has V_o= F (t) sin (ωt) is input. The sample and hold (SH) circuit 53 holds the input waveform at the rising edge of the clock from the clock circuit 15B. The modulation signal input to the sample and hold (SH) circuit 53 is f (t) sin (ωt).
[0069]
  Here, the waveform is held when the sample and hold circuit always has sin (ωt) = 1. That is,
    T_sm = (2π / ω) m +π / 2
A sample and hold (SH) circuit 53 samples each time. m is an integer. Output V of sample and hold (SH) circuit 53_omIssin (T _sm ) = 1
    V_om= F (T_sm) Sin (T_sm) = F (T_sm)
It becomes.
[0070]
That is, by sampling in the sample and hold (SH) circuit 53, the modulated signal modulated by the carrier can demodulate the output of the same waveform as the input analog signal applied between the input terminals 5a and 5b. Since the output from the sample-and-hold (SH) circuit 53 includes harmonics more than three times the sampling frequency, the harmonic component is removed via the low-pass filter 54 and the input f (t) is demodulated and output. Output to terminal 10. This embodiment is suitable for obtaining an analog output.
[0071]
In the case of the low-pass filter 67 shown in FIG. 9 showing the conventional example, the object of removal is a frequency component twice that of the carrier, whereas in the case of the present application shown in FIG. Since it is a target, the influence of the low-pass filter 54 on the signal to be restored is extremely small.
[0072]
Here, as the insulating circuits 11 and 12, an insulating transformer or a capacitor may be used, radio waves may be transmitted between opposing antennas, optical fibers may be inserted to face each other, and light may be transmitted and received, or a space may be inserted. It will be apparent from the above description that light may be transmitted and received oppositely.
[0073]
【The invention's effect】
Since the synchronous detection method of Document 1 is modulated and demodulated with a high-frequency sine wave, a filter having a frequency twice that of the carrier signal is required, whereas the analog insulation circuit of the present invention has a sinusoidal shape. Modulates at a high frequency but demodulates using a sampling circuit. Thus, a low-pass filter that removes the carrier signal is not always necessary. The sinusoidal frequency as the carrier signal does not need to be sufficiently high with respect to the band of the insulating circuit. Therefore, each circuit has an advantage that a high-band configuration is not required. The pulse transformer has the advantage that it can be used for a commercially available high-speed LAN.
[0074]
  Furthermore, the mainstream of oscilloscopes has recently shifted to a digital storage oscilloscope (DSO), which incorporates an A / D converter and an oscillator for the sampling clock of the A / D converter. When the present invention is used for DSO input isolation, the sampling circuit and the oscillator for outputting a sine wave according to the present invention can use a DSO A / D converter and an oscillator, respectively. By adding one multiplication circuit, there is an advantage that DSO input isolation is possible.In literature 1, two broadband multiplication circuits are required, whereas in the present invention, only one is required, and the broadband property as in literature 1 is not required.Therefore, the effect of the present invention is extremely great.
[Brief description of the drawings]
FIG. 1 is a circuit configuration diagram showing an embodiment of the present invention.
2 is an input and output waveform diagram of the clock circuit of FIG. 1. FIG.
FIG. 3 is a signal input and output waveform diagram to the multiplication circuit of FIG. 1;
FIG. 4 is a diagram showing input and output waveforms of other signals to the multiplication circuit of FIG. 1;
5 is an enlarged waveform diagram of part A in FIG.
FIG. 6 is a waveform diagram of still another signal input and output to the multiplication circuit of FIG. 1;
7 is an enlarged waveform diagram of part B in FIG. 6;
FIG. 8 is a circuit configuration diagram showing an embodiment of the present invention.
FIG. 9 is a circuit configuration diagram showing a conventional example.
10 is a circuit diagram of a Gilbert type multiplication circuit used in the multiplication circuit of FIG. 9;
11 is a waveform diagram of each part of FIGS. 9 and 10. FIG.
[Explanation of symbols]
1a, 1b, 3a, 3b, 4a, 4b terminals
5a, 5b input terminal
7 Buffer amplifier
9 terminals
10 Output terminal
11, 12 Insulation circuit
15, 15B clock circuit
18 Oscillator
20 Multiplication circuit
31-38 transistor
41-47 resistance
51 A / D converter
52 D / A converter
53 SH (sample hold) circuit
54 Low-pass filter
60 multiplication circuit
67 Low-pass filter
69 Buffer Amplifier

Claims (7)

A floating-side multiplication means (7, 20) for multiplying the floating-side analog input signal and the floating-side high-frequency sine wave signal to obtain a floating-side multiplication result;
An insulating means for input signals (11) for insulating the floating side from the non-floating side and making the result of multiplication on the floating side an input signal on the non-floating side;
Non-floating side sampling means (51, 52, 53, 54) for sampling the non-floating side input signal with a clock to obtain a non-floating side sampling output (10);
Oscillating means (18) for oscillating a high-frequency sine wave signal;
High-frequency sine wave insulation means (12) for insulating the floating side from the non-floating side in order to obtain the high-frequency sine wave signal from the oscillation means (18) as the high-frequency sine wave signal on the floating side;
An analog insulation circuit comprising: clock means (15) for obtaining the clock synchronized with a peak time of the high-frequency sine wave signal oscillated by the oscillating means (18). The floating side multiplication means (20)
The analog insulation circuit according to claim 1, comprising a Gilbert-type multiplication circuit for multiplying the analog input signal and the high-frequency sine wave signal . The analog insulation circuit according to claim 1, wherein the input signal insulation means (11) and the high-frequency sine wave insulation means (12) are each constituted by an insulation transformer. The analog insulation circuit according to claim 1, wherein each of the input signal insulation means (11) and the high-frequency sine wave insulation means (12) includes a capacitor. The sampling means;
The analog insulation circuit according to claim 1, further comprising an A / D converter (51). The sampling means;
The analog insulation circuit according to claim 1, comprising an A / D converter (51) and a D / A converter (52). The sampling means;
The analog isolation circuit according to claim 1, comprising an A / D converter (51), a D / A converter (52), and a low-pass filter.

Images