JPH07198766A - Calibrating method for circuit constant measuring instrument - Google Patents

Abstract

PURPOSE:To improve the measurement accuracy of phase angles without deteriorating the measurement accuracy of absolute values of impedances by using an air capacitor as a phase standard and making up the uncertainty in phase angle of a limited-impedance standard with the phase standard. CONSTITUTION:A limited-impedance standard and phase angle standard are respectively represented as LOAD1 and LOAD2. The absolute value ZIs1 of the impedance of the LOAD1 is evaluated, but the phase angle thetaIS of the LOAD1 is unknown. In addition, the absolute value ZIs2 of the impedance of the LOAD2 is unknown, but the phase angle of the thetaIs2 is evaluated. Correction factors A and B must become the same even when either one of the LOAD1 and LOAD2 is used. From such a condition, the unknown phase angle thetaIs1 or absolute value ¦ZIs2¦ is found based on a prescribed formula. Then the correction factors A and B are obtained by substituting a complex impedance found from the absolute value of the impedance of the LOAD1 or the phase angle of the LOAD2 for the ZIs of the formula.

Description

Detailed Description of the Invention

[0001]

FIELD OF THE INVENTION The present invention relates generally to apparatus for measuring circuit elements and networks, and more particularly to a measuring apparatus for measuring characteristics of high Q circuit elements or networks.

[0002]

2. Description of the Related Art A three-point calibration method, which allows complete correction, is used for error correction of circuit elements and network measuring devices. This is a method of obtaining the values of the three correction coefficients included in the correction equation by calibration using three different impedance standards. Before pointing out the problems of the prior art, the correction formula will be described. FIG. 3 is a generalized model diagram of a circuit element and a circuit network measuring device. In the figure, the signal source 31, the device under test 32, and the measuring devices 33 and 34 are connected to an intermediary network 35. Impedance is a function of only the ratio of two measurement values (voltage or current, but a complex number with a phase) at the points inside the intermediary network (bilinear transformation)
Given by. That is, if the true value of the impedance of the device under test 32 is Zx and the ratio of the indicated values of the measuring devices 33 and 34, that is, the measured value is Zm, the relationship between Zx and Zm is expressed by the following bilinear transformation equation. become.

[0003]

[Equation 1]

A, B, and C are unique coefficients determined by the intermediary network and do not change if the network is linear. The equation (1) is a correction equation for obtaining the true value from the raw measurement value, and the coefficients A, B and C are the correction coefficients. Since the correction coefficient is three complex numbers, A, B, and C can be obtained by calibrating with three different impedances as standards. That is, the three standard true values of impedance are respectively substituted into Zx of the equation (1) and the measured value thereof is substituted into Zm to make simultaneous equations, and A, B and C can be solved as unknowns. Since the correction coefficient is obtained from the standard values of the three impedances as described above, it is desirable that the three standard impedances be significantly different from each other. Therefore, normally, a short circuit, which is the limit value of impedance, an open circuit, which is the limit value of admittance, and a finite value standard of impedance are used.
The standard of finite impedance is hereinafter referred to as a finite impedance standard. At high frequencies, a resistor of 50Ω is usually used. Here, the open of the standard device of three impedance is OPEN, the short circuit is SHOR
T, a finite impedance standard is referred to as LOAD. As mentioned above, solve the simultaneous equations created by substituting the three standard true values and raw measured values into the equation (1),
The correction coefficients A, B, and C of the three-point calibration method are calculated as follows.

[0005]

[Equation 2]

The correction coefficient thus obtained is substituted into the equation (1) to complete the correction equation. If the accuracy of the complex values Yos, Zss, and Zls including the parasitic components for each of these short-circuit, open-circuit, and finite impedance standards is better than the value required from the measurement accuracy of the circuit element and the network measurement device, (1 ) It is possible to make a correction that satisfies the required accuracy by the equation (2). At low frequencies where residual and parasitic components of the three standards can be ignored, there is no practical problem even if the short-circuit impedance is 0, the open admittance is 0, and the finite impedance standard is a pure resistance value. The method is being used effectively.

The problem with the prior art is that, in the field of high frequency or high precision measurement, particularly in the field of impedance measurement of low loss components in the high frequency band, it is not possible to set the price of a standard device with sufficient precision to realize high accuracy measurement. Is. Since the element to be measured has a finite impedance, the accuracy of the phase of the measured value depends on the accuracy of the phase of the finite impedance standard used for calibration. As described above, when the phase of the finite impedance standard is calibrated as 0, that is, pure resistance over the entire range of the measurement frequency, the difference between the true frequency characteristic of the phase of the standard and the assumption that the phase is 0 over all frequencies is calibrated. It remains as a later phase error. For example, if the phase error after calibration is 1000 μrad, the Q measurement error of the device under test with Q≈100 is 10%.

FIG. 2 shows the configuration of a coaxial 50Ω type finite impedance standard device which is often used in a high frequency range. A plate-shaped 50Ω resistance element 22 is housed in the outer conductor 23, and the outer terminal of the resistance element is connected to the inner wall of the outer conductor. This is the coaxial 50 consisting of the connector section 21.
It is connected so as to terminate at the Ω line. Sleeve 24
Is a protective cover for covering the outer conductor 23. Factors that affect the phase characteristics of the impedance include the residual reactance component of the resistance element 22 itself, the dimensional error of the coaxial line portion of the connector section 21 and the variation of the characteristic impedance due to the skin effect, and the coaxial relationship between the resistance element 22 and the connector section 21. There are discontinuities in the connection with the track. Simply looking at the case where the skin effect brings the frequency characteristic to the coaxial inductance of the connector part, in the frequency range of about 1 GHz, 10
A phase variation of about 00 μrad can be estimated. Furthermore, considering the other factors mentioned above, the frequency characteristic of the phase angle is very complicated and the estimation itself is difficult. When mass-produced,
The phase variation is large and it is difficult to manage the variation. In this way, when calibrating by considering the phase angle of a finite impedance standard device having a complicated phase characteristic as 0, Q ≈ 1 GHz even if the error factor is only the skin effect.
When the Q measurement of 100 DUTs is made, the error is 10
It also becomes%. Therefore, as another method, it is considered that the frequency characteristic of the phase of the finite impedance standard device is correctly valued and used for calibration. However, as described above, it is not easy to develop a method for correctly setting the phase angle of a finite impedance standard device having a complicated frequency characteristic over the entire measurement frequency range. Even if such a method is developed, it is not easy to guarantee stability against changes in time and environment. Therefore, it is practically impossible to Q-measure Q≈100 at a high frequency of 1 GHz with an accuracy of 10% or less.

[0009]

The object of the present invention is to measure the phase angle without impairing the accuracy of measuring the absolute value of the impedance, as compared with the conventional three-point calibration method, mainly in the field of impedance measurement of high Q components in the high frequency band. The purpose is to present a simple calibration method that improves accuracy and a newly added standard.

[0010]

SUMMARY OF THE INVENTION In addition to the conventional three-point calibration standard, one embodiment of the present invention uses an air capacitor as a standard with very low loss, ie, a phase angle of impedance very close to -π / 2. This is a method in which a standard device is used and the uncertainty of the phase angle of the finite impedance standard device is compensated by the phase standard device. In this method, the measurement error of the absolute value of the impedance of the device under test is almost determined by the uncertainty of the absolute value of the finite impedance standard, as in the conventional three-point calibration.
The phase error is determined by the uncertainty of the phase angle of the newly added phase standard, and the uncertainty of the phase angle of the phase standard is smaller than the uncertainty of the phase angle of the finite impedance standard. Errors in phase measurement are reduced compared to point calibration.

[0011]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention provides a phase angle standard whose absolute value of impedance is unknown but whose phase angle is less than or equal to the required uncertainty. Compensate for the uncertainty of the above and calibrate the circuit element and the network measuring device. Here, the finite impedance standard is represented as LOAD1 and the phase angle standard is represented as LOAD2. The absolute value of impedance LOAD1 | Zls1 | is given, but the phase angle θls1 is unknown. Also L
The absolute value | Zls2 | of the impedance of OAD2 is unknown, but the phase angle θls2 is valued. LOAD1
The correction factors A, B, and C must be the same regardless of which of the two LOADs, LOAD and LOAD2, is used. From this condition, the unknown number θls1 or | Zls2 | is obtained using any one of the expressions (2). Then, in Zls of equation (2), L
Substituting the absolute value of the impedance of OAD1 or LOAD2 and the complex impedance obtained from the phase angle,
Finally, the correction coefficient is obtained.

Although this method is an exact solution for calculating the correction coefficient of the circuit network exactly, it can be replaced with an approximate solution by the iterative method. Next, the procedure of the iterative method is shown.
The unknown phase angle θls1 of LOAD1 is set to, for example, θls1 = 0, and the temporary value of Zls1 is substituted into the equation (2) to obtain the temporary correction coefficient. Next, using this correction coefficient, the correction value Zls2 'of LOAD2 is obtained according to the equation (2). Next, the difference between the phase angle θls2 'of Zls2' and the valued phase angle θls2 of LOAD2

[0013]

[Equation 3]

Only the phase angle of LOAD1 is corrected. That is, the phase angle of LOAD1 is newly added.

[0015]

[Equation 4]

It is modified as follows. Next, using this new phase angle, change the value of LOAD1

[0017]

[Equation 5]

Substituting this value into equation (2) again,
Calculate the correction coefficient. The correction coefficient converges to an exact solution by repeating this procedure. This iterative method is practically 1
One time is enough.

Equation (2) is an equation in which residual and parasitic components of OPEN and SHORT are not ignored, but ideal O
In case of PEN and SHORT

[0020]

[Equation 6]

[0021] Therefore, the equation (2) is simplified as follows.

[0022]

[Equation 7]

When OPEN and SHORT are ideal, only the coefficient A depends on the value of LOAD as shown in the equation (7). Therefore, even if the exact solution is applied, θls1 and Zl
s2 is easily found. I.e.

[0024]

[Equation 8]

From this θls1 and the known | Zls1 |, or from | Zls2 | and the known θls2, Zls is obtained, and Z in the equation (7) is obtained.
Substituting for ls, the correction coefficient is obtained.

Next, an embodiment of the phase standard will be described. What is required of the phase standard in calibration is that the uncertainty of the phase, that is, the loss coefficient D is small, and the accuracy and stability of the capacitance are not required. To keep the uncertainty of D small,
It is sufficient to reduce D itself. To reduce D, it is necessary to reduce the shape of the capacitor. For example, as a capacitor having a coaxial structure, by making the total length a little less than 10 mm, D can be set to 500E-6 or less at 1 GHz. FIG. 1 shows an embodiment of a capacitor having a coaxial structure that realizes this. The figure is a vertical sectional view passing through the coaxial center axis.
Details not relevant to the present invention have been omitted. In the figure, 11 is a coaxial inner conductor, 12 is an outer conductor, and these are sleeve 1
It is stored within 3. In the figure, a thin portion on the left side of the inner conductor 11 is a coaxial line portion that forms a characteristic impedance of 50Ω and is connected to a 50Ω connector on the measuring instrument side. The thickened portion 11 and 12 constitute a concentric cylinder air capacitor. The insulator 15 is press-fitted into the inner conductor 11, and the insulator 15 holds the inner conductor 11. The insulator itself is fixed to the outer conductor 12 with a push screw 18. The recess 14 on the right side of the portion where the outer conductor 12 forms the electrostatic capacitance is provided so that the electric field is not concentrated on the right end surface of the inner conductor 11. Reference numeral 16 is a knob for holding and fixing the outer conductor 12 when the sleeve 13 is rotated to attach / detach the standard device to / from the measuring device. The knob 16 is fixed to the outer conductor 12 with a push screw 17.

In order to reduce the loss coefficient D of the standard device,
It is natural to use a material having a small shape and a small loss for the insulator 15 that holds the internal conductor. The capacitance is determined by the ratio of the outer diameter of the inner conductor to the inner diameter of the outer conductor and the length of the inner conductor, but the accuracy of the capacitance value does not matter with this standard. With this structure, as described above, at 1 GHz, D <5
00E-6 can be realized. By applying the calibration method of the present invention using this, the Q of the device under test with Q≈100 can be measured at 1 GHz with an uncertainty of 5% or less. As described above, it is easy to estimate D of the air capacitor, as compared with the difficulty of estimating the phase characteristic of 50Ω. Therefore, it goes without saying that if the value of LOAD2 is carefully set, the Q measurement accuracy will be further improved.

[0028]

As described above, it is easy to estimate the loss coefficient D of the capacitance, while it is difficult to estimate the phase characteristics of the resistance standard device. By applying the calibration method of the present invention using this, it is possible to measure a device under test with Q≈100 at 1 GHz with an uncertainty of 5% or less. This makes it possible to measure the impedance characteristics of various parts and networks with particularly high Q at higher frequencies with better accuracy, which is useful for practical use. It should be noted that the exemplified calibration method and standard are not limited to the format and others, and can take a wider range of values, and can be modified in structure and components as necessary without losing the gist of the present invention. Is also acceptable.

[Brief description of drawings]

FIG. 1 is a diagram showing an embodiment of a phase standard device of the present invention.

FIG. 2 is a diagram showing an example of a coaxial finite impedance standard device.

FIG. 3 is a model diagram of a circuit element and a circuit network measuring device.

[Explanation of symbols]

 11: inner conductor 12: outer conductor 13: sleeve 14: recess of outer conductor 15: insulator 16: knob 17: push screw 18: push screw 21: connector portion 22: resistance element 23: outer conductor 24: sleeve 31: signal Source 32: Device under test 33: Measuring instrument 34: Measuring instrument 35: Mediation network

Claims (2)

[Claims] 1. A method of calibrating an error of a measurement system using a standard of open, short, and finite impedance, and correcting a measured value, wherein a finite impedance standard device has an absolute value of impedance but a phase angle. A method for calibrating a circuit constant measuring device characterized by performing high-precision correction by dividing into a standard device with uncertain values and a standard device with uncertain absolute values for the phase angle. 2. The calibration method according to claim 1, wherein a low-loss capacitor having a coaxial structure is used as the standard device whose phase angle is valued.