=Paper=
{{Paper
|id=Vol-2300/Paper4
|storemode=property
|title=Software for Tolerance Design of Electronic Devices
|pdfUrl=https://ceur-ws.org/Vol-2300/Paper4.pdf
|volume=Vol-2300
|authors=Galina Shilo,Nataliia Furmanova,Tetyana Kulyaba-Kharitonova
|dblpUrl=https://dblp.org/rec/conf/acit4/ShiloFK18
}}
==Software for Tolerance Design of Electronic Devices==
https://ceur-ws.org/Vol-2300/Paper4.pdf
14
Software for Tolerance Design of Electronic Devices
Galina Shilo, Nataliia Furmanova, Tetyana Kulyaba-Kharitonova
Department of Information Technologies of Electronic Devices, Zaporizhzhia National Technical University, UKRAINE,
Zaporizhzhia, Zhukovskogo str., 64, E-mail: shilo.gn@gmail.com, nfurmanova@gmail.com, tikylyaba@gmail.com
Abstract: The algorithms of tolerance design which are However, to ensure sufficient accuracy in the synthesis of
based on the geometric representation of the tolerable tolerances, the Monte Carlo method requires a large number of
domain are proposed. The shape of the domain is defined tests. The method of moments is used only in the normal law
by law of distributions and optimization criterion. of the parameters distribution and does not provide sufficient
Software for tolerance design is developed. The software accuracy due to the limited number of parameters of the
can be used as a separate system for mathematical distribution law (mathematical expectation and dispersion).
modelling of electronic devices and can be integrated as a The accuracy of the procedure for assigning tolerances was
component in different electronic design automation increased when the method of tangent [6] was introduced. In
systems or computer-aided design systems, for example in this method, standard deviations of the parameters are formed
SPICE or HFSS. at the tangent point of tolerance domain and domain of
Keywords – tolerable domain, tolerance design, operational capability of device. The interval models of output
optimization criterion, electronic devices, computer-aided functions are also used. This approach is successfully used in
design. the procedures for assigning tolerances with a uniform [7],
normal law of the parameters distribution [8] or if such a
I. INTRODUCTION distribution law is given by statistical series [9]. There is a
The assignment of tolerances is one of the most important need to implement these approaches in modern computer-
stages in the design of various devices, because standard aided design systems.
deviations values of the elements parameters set the accuracy When dealing with issues, the peculiarities of the
of the output functions and affect the equipment cost. The manufacturing and operation of electronic equipment should
problem of ensuring the accuracy of output characteristics and be taken into account. It is necessary to provide the possibility
parameters of the electronic equipment is solved at the stage to consider the law of distribution of parameters, the
of circuit design by tolerance design. Two aspects are to be correlation between the parameters of the components, and
considered. The deviations of the output functions are to be possible compensation of the external factors.
obtained if the deviations of components parameters are given The tolerance synthesis is the inverse problem and therefore
(tolerance analysis). The deviations of parameters are to be it is inaccurate and resolved ambiguously. Thereby, it is
defined if the boundary deviations of the output functions are necessary to apply methods of optimization according to
given (tolerance synthesis). different criteria for solving this problem. In this case it is
In modern systems of computer-aided design of electronic convenient to use different algorithms for each design strategy:
devices the procedures of the analysis of tolerances are strategy for equal tolerances ( δ -strategy), maximum volume
implemented. These procedures enable to calculate the relative tolerance domain (V- strategy), the minimum cost (P-
sensitivity of the characteristics of the circuit to change the strategy), optimal price / quality ratio (P/V- strategy). Thus, for
parameters of the selected element, to perform the Monte the software implementation, the following algorithms were
Carlo statistical analysis, the simulation of the worst case, to developed:
research the circuits taking into account the technological • algorithms for the synthesis of interval tolerances
parameters spread of the electronic elements, and temperature using 4 design strategies;
dependences. Unfortunately, these systems do not include • algorithms for the synthesis of tolerances on
tools for solving the problems of the tolerances synthesis. component parameters, taking into account the normal
The purpose of the work is to develop software for the distribution law for 4 design strategies;
tolerance design of electronic devices and integrate it with • algorithms for the synthesis of tolerances on
EDA or ECAD systems. component parameters taking into account the distribution law
II. ALGORITHMS OF THE TOLERANCE DESIGN FOR given by statistical series for 4 design strategies;
THE ELECTRONIC EQUIPMENT • algorithms for the synthesis of tolerances on the
parameters of components, taking into account the correlation
Knowledge in the field of tolerance design has been between the parameters;
accumulated for many years [1-4]. During this time, it has been • algorithms for the synthesis of tolerances, taking into
established that the value of the tolerances assignment greatly account the coefficients of external influences, which are given
depends on the law of distribution of parameters, which is in the form of interval structures.
formed in the process of manufacturing components for The algorithm that enables to assign the same symmetric
different devices. It has led to the widespread using the method interval tolerances to the component parameters is following:
of moments, statistical tests of Monte Carlo in tolerances
assignment [5].
ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic
15
Step 1. The coefficients of the model of the output function factors is based on the mapping method and has the following
are determined at the point of the parameters nominal values form:
and the initial values of the parameters deviations of the Step 1. x di
( 0)
boundary values of the parameters of the
elements are assigned:
elements are determined based on y boundary operating
e
δ y
, a ri > 0 value of the output function and xri rating values of elements
δi = n
( 0)
, (1) parameters. The algorithm for assigning interval tolerances
δ
− y otherwise with a given optimization criterion is used.
n
Step 2. In the vicinity of B (d0) boundary point d i relative
where δ i is the relative deviation of component parameters; changes in the elements parameters are determined regarding
( )
δ y = y − y r / y r is the lower standard deviation of the output the most unfavorable combination of external factors.
( 0)
Step 3. The coordinates of x di boundary points are
function; y r is the rating value of the output function; y is
mapped onto normal environmental conditions, and the
the lower value of output function.
nominal boundary value of the output function is determined
Step 2. The initial coordinates of the point of tangency
in the first approximation:
between the top of the tolerance domain and the domain
boundaries of working operation are determined: x (r0i ) = x (di0) / d i ; y (1) = y ( X (r0) ) , (6)
r
where X r = {x r1 , , x r n } is the set of coordinates of
( 0) ( 0) ( 0)
x i(0) = xri (1 + δ i(0) ) . (2)
Step 3. The coefficients of the model of the output function B (r0) boundary point.
are determined at the point of tangency between the top of the Step 4. The possibility of implementing the algorithm for
tolerance domain and the boundaries of the capability region. a given boundary operational value of the output function and
Standard deviations are assigned by the formula given coefficients of external influences is tested. To do this,
n the condition is checked:
δ = bw / ∑ a i + a i xri , (3)
i =1
y (1) < yr , (7)
r
where bw = b − b ; xri are the rating values of components where yr = y ( X r ) is the rating value of the output function.
parameters; ai are the model coefficients of the output If the condition is satisfied, the algorithm ends and the
function at the point of the rating values of parameters. message about the impossibility of implementation is
Step 4. The coordinates of the point of tangency between displayed.
the top of the tolerance domain and the boundaries of the area Step 5. Nominal interval tolerances are assigned at the
(k )
of working operation are determined: boundary values of y (k ) output function and x r i rating
r
(
x i( k ) = xri 1 + δ i( k ) ), (4) boundary values of the elements parameters are determined.
The algorithm for assigning interval tolerances is used.
Step 6. The boundary values of the elements parameters
where δ i(k ) is the standard deviation of parameters on k
and the output function are determined by the effect of external
iteration. factors:
Step 5. The value of the output function is determined at
x (dik ) = x (rki ) d i ; y ( k ) = y ( X (dk ) ) . (8)
the tangency point of the top of the tolerance domain and the d
boundaries of the operational capability region. The Step 7. The relative change in the boundary value of the
completion condition of the algorithm is checked: output function is determined and its boundary nominal value
is specified:
y (k ) − y
≤ε , (5) d (yk ) = y ( k ) / y ( k ) ;
d r
y
y ( k ) = y ( k −1) − ( y ( k ) − y ) / d (yk ) . (9)
where ε is accuracy of calculations. r r d e
Step 8. The condition for the completion of the algorithm
When the condition is satisfied, the algorithm is over,
is checked:
otherwise, step 3 follows.
The algorithms for V-, P- and P/V- strategies include the y (k) − y
same steps but deviations of parameters are defined by
d e
≤ε, (10)
y
formulae given in [7]-[10]. e
In order to take into account external influences on where ε is the accuracy of calculation.
electronic devices, interval structures are used in tolerance If the condition is satisfied, the algorithm ends, otherwise,
synthesis [11]. It allows to store information about a range of go to step 5.
changes in external factors and to estimate compensation for
their effect. The algorithm for taking into account external These algorithms are used to develop the software for
tolerance design.
ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic
16
III. TOLERANCE DESIGN SOFTWARE The rating values of parameters are the following
R2 = 110 kOhm; R3 = 110 kOhm; C1 = 0,47 mkF;
Tolerance design software are allowed to analyze and
synthesize deviations of geometric and electric parameters of C2 = 0,47 mkF. The rating value of pulse duration is
electronic devices and select components. The main features τ r = 35,84 ms.
are
- analysis of deviations of output function by given standard If the deviation of pulse duration is ∆τ = ±1 ms, then standard
deviation of parameters of components; deviations for the elements parameters are δ R = δ C = ±1,43 .
- synthesis of equal deviations of the parameters; If the deviation of the pulse duration is ∆τ = ±2 ms, then
- synthesis of deviations of the parameters for case of standard deviations for the parameters are
maximal volume of tolerance domain; δ R = δ C = ±2.86 %. For the boundary values of the delay
- selection of components by optimal price/quality ratio.
Software consists of the modules. They are the forming of time, the values of tolerances on the parameters of the scheme
the models of output characteristics, tolerance analysis, components are given in Table 1 and Table 2.
tolerance synthesis, component selection, report generation. TABLE 1. BOUNDARY VALUES OF DELAY TIME FOR THE
The structure of the software is shown in Fig. 1.
DEVIATION OF ∆τ = ±1 MS PULSE DURATION.
Boundary deviations of parameters, % Deviations
Delay
of delay
time
δ C1 δ R3 δ R2 δC2 time, ms
τ 1,43 1,43 1,43 1,43 1,031
τ –1,43 –1,43 –1,43 –1,43 –1,017
TABLE 2. BOUNDARY VALUES OF DELAY TIME FOR THE
DEVIATION OF ∆τ = ±2 MS PULSE DURATION.
Boundary deviations of parameters, % Deviations of
Delay
delay time,
time
δ C1 δ R3 δ R2 δC2 ms
Fig. 1. The structure of the software for the tolerance design
τ 2,86 2,86 2,86 2,86 2,077
The Database is used for component selection. Tolerance
synthesis module includes τ –2,86 –2,86 –2,86 –2,86 –2,019
- selection of optimality criterion; If the component parameters are distributed according to
- procedures for synthesis of the tolerances of electronic the normal distribution law, then for the case ∆τ = ±1 ms;
device parameters for the uniform law of parameter
tolerances are δ R = δ C = ±2.74 %. Delay time values are
distribution, the normal law, and distribution given by
statistical series; τ = 1,031 ms; τ = −1,017 ms. If ∆τ = ±2 ms, then
- procedures for synthesis of the tolerances taking account δ R = δ C = ±5.47 % and boundary values of delay time are
correlation between parameters;
- procedures for synthesis of the operating tolerances; τ = 2,077 ms; τ = −2,019 ms.
Input data are deviations of output function, the law of Integration of the developed automated system with
parameter distribution, rating values of parameters, modern CAD is possible in three ways:
coefficients of external factors, accuracy of calculation. As a • by developing the interface software module, which
result, the tolerances of parameters are defined. In the software uses specialized macros to calculate the output characteristics
it is necessary to choose the mode of output function forming of the radio electronic device in CAD environment;
in the program system. There are the symbolic form mode and • by developing an internal application in CAD, which
the mode of forming models by external ECAD system. implements the calculation of output characteristics of radio-
The module for the formation of output characteristics electronic devices using API functions;
enables to describe the mathematical models in a symbolic • by inputing the value of the output characteristic,
form. The formulae are created in two modes: command mode calculated using a special CAD system in a dialogue mode.
and visual ones. MathML library is used for visualization. The first integration method is implemented in the ANSYS
An example of the application of tolerance design software HFSS system for the low frequency coaxial filter. The example
for synthesize the deviations of the parameters of the power of program interface is shown in Fig. 2. LPF is designed to
pulse generator is given below. Pulse duration is defined by reduce the level of side-effects in the spectrum of probe signals
τ τ and suppression of off-band radiation in the spectrum of output
e R3C1
= 1− e R2C 2
. signals of amplifier modules of a distributed transmitting
device in the antenna.
ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic
17
IV. CONCLUSION
The software of tolerance design has been developed. It
provides an opportunity to analyze and synthesize tolerances
if the tolerance domain is represented by various geometric
objects (boxes, ellipsoids, and their combination). Individual
procedures are implemented to assign tolerances of parameters
taking into account the correlation between the parameters, the
effect of external factors during the operation stage. The
possibility to calculate tolerances for different design
strategies is provided. These strategies ensure either the best
quality or minimum cost, or taking into account the features of
the technological process (strategy of equal tolerances).
REFERENCES
[1] B. Mandziy, B. Volochiy, L. Ozirkovskyy “Program
models for interactive design of fault-tolerant system with
mixed structural redundancy taking account maintenance
service strategies,” Computing, Vol. 7, Issue 1, 2008,
pp.161-163.
[2] R. Spence and R.S. Soin, Tolerance Design of Electronic
Fig. 2. Program interface of tolerance design software Circuits, World Scientific, 2002, 232 p.
[3] J.B. Grimbleby, Computer-aided analysis and design of
For an example of the applying of the tolerance assignment electronic networks, London: Pitman, 1990, 289 p.
system, a coaxial low pass filter (LPF) with following [4] L. Kolev, “Worst-case tolerance analysis of linear DC and
parameters was designed: AC electric circuits”, IEEE Trans.Circuits Systems, vol.
• number of links -19, cut-off frequency of the LPF ~ 47, 2002, pp. 1 – 9.
3300 MHz; [5] J. Gajda and T. Sidor, “Using Monte Carlo Analysis for
• bandwidth from 2800 to 3100 MHz; Practical Investigation of Sensitivity of Electronic
• maximum losses in the bandwidth are not more than Converters in Respect to Component Tolerances,”
0,3 dB; Electrical And Electronic Engineering Vol. 2 (5), 2012,
• the static wave constant at the voltage in the pp. 297-302.
bandwidth K ≤ 1,2; [6] G.Shilo Geometric Methods of Tolerances Setting
• effective attenuation in the barrier band (for the 2nd //Journal of Automation and information sciences –2007.
and 3rd harmonics band) - not less than 60 dB: from 5600 to – Vol. 39. – № 3. – P. 51-60.
6200 MHz; from 8400 to 9300 MHz. [7] G.M.Shilo, “Strategies for Assigning Interval Tolerances”
As result the three-dimensional model of filter is generated. Cybernetics and Systems Analysis, vol. 51, issue 4, 2015,
This model is shown in Fig.3. pp. 657-666.
[8] G.N.Shilo, “Normal tolerance assigning by given price
characteristics of radio components” Radioelectronics
and Communications SystemsVolume 55, Issue 3, March
2012, Pages 140-148
[9] G.Shilo, “Assigning Tolerances by Method of Smoothed
Vertices,” Journal of Automation and Information
Sciences, Begell House, Inc, vol. 45 , issue 10, 2013,
pp.36-48.
[10] G.Shilo, N.Furmanova “Statistically Oriented Tolerance
Design with Correlation between Parameters of
Components” Proceedings of the 9th IEEE International
Fig.3. The three-dimensional model of coaxial low pass filter. Conference on Intelligent Data Acquisition and Advanced
Analysis of the influence of the parameters tolerances of the Computing Systems: Technology and Applications 21-23
step transition to the standing wave coefficient by the interval September, 2017, Bucharest, Romania. – p.1082 – 1087.
methods was carried out. [11] G.M.Shilo, “Interval structures and their applications”
Computing,vol 6, issue 1,2007, pp. 80-87.
ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic