lnvited Paper
This paper describes threeextremdy simple electronic circuitsin Hands-on experience with those circuits tells us that
which chaotic phenomena have been observed. The simplicity of
the circuits allowsone to there are low-order deterministic (Newtonian)
i) build them easily, systems which are ”unpredictable” (1.3)
ii) confirm the observedphenomena by digital computer sim- in the sense that even anextremely small changeof the ini-
ulation, and in some cases
iii) rigorously prove the circuit is indeed chaotic. tial condition eventually gives rise to an entirely different
trajectory. The periodic oscillatorsare “predictable” in that
A consequence of i) is that the interested reader can build, and
then see and even listen to chaos. every trajectory eventuallyconverges to the same periodic
It is to be emphasized that these circuits are not analog com- orbit irrespective of the initial condition. Experience also
puters. They arereal physical systems. shows that
o 1987IEEE
001a9219/~7/~1033501.00
IVR
-
(b)
Fig. 2. A realization of the circuit in Fig. 1. (a) Circuitry. Q,
Q2 = 2SC1815, D,,D2 = 1S1588. (b) Measured v-i charac-
teristic of N. Horizontal scale: 5 V/div. Vertical scale: 1 mA/
div.
C. Confirmation
The dynamics of the circuit inFig. 1 is governed by
(b)
Fig. 1. A simple autonomous circuit with a chaotic attrac-
tor. (a) The circuitry. (b) v-i characteristic of the nonlinear
resistor.
8.2mH D. Analysis
Because of the simplicityof (2.2), one can perform arig-
orous analysis. In order to simplify the analysis, we trans-
form (2.2) into
-15V t 15V p x - y + z
Fig. 6. Another realization of the circuit in Fig. 1.
I$ = -By
these form the "double scroll." would stay on €YO)and asymptotically approach 0, because
Let uslook at a cross section of theattractor. Fig. 11 gives EC(0)is invariant. Such a trajectoryis called homoclinic and
the cross section at vc, = 0, where the double-scroll struc- it is related to a very complicated behavior of solutions to
ture is clearly seen. differential equations.A rigorous statementis given bythe
following theorem of Shilnikov ([3]-[5l):
Theorem (Shilnikov)
Consider
dx
-
dr
= fQ
Let us briefly describe how one can prove this. Recall that whc .e Q1(U,(resp., CoIU1)denotes the restriction of P1
what onewants to prove is (2.7).This, however, is extremely (resp., q0)
to U1.These maps can be explicitly givenin terms
difficult, for onehas to compute the return time, tl, at which of the eigenvalues.
a trajectoryhits the plane U1.In general, it is impossible to Consider the negative half return map (Fig. 14) in the Dl-
compute tl analytically, because thetrajectory cp"(C) unit defined by
involves sin, cos, and exp, andtherefore, tl i s defined only rdx) = cpY'(X), x E Vl (2.11)
implicitly by transcendental
a equation.In order to overcome
this difficulty, wewill make the following change of coor- where 9'; is the flow in the &-unit and
dinate systems (Fig. 14): T = inf { t > o:cp;'(x) E v1}. (2.12)
a) Take a map Pl: ?13 + Fi3 such that
Now the homoclinicity condition(2.7)can be expressedas
*.,(P+)= 0
P1(U,) = v, = { ( x , y, z ) : x +z = I}
F l (2.13)
rul-I O- where
x (2.8) CT = Pl(C) A1 = *1(A) €1 = PI(€).
Although the transformed flow (p;'has a simplerexpres-
L
sion (recall (2.8))than the original flow (or, the half return
where u1 = 6,/Gl and y1 = +,/G1 and D denotes a deriva- time, T, defined by (2.12)is still a solution to a transcen-
tive. dental equation.The following proposition, however, pro-
vides us with a breakthrough.
z D,-unit
(2.14)
where h = (1,0,1).
1.0
A1 = (1, p1) CI = (XC, yc)
and
U.”
El = (X€, YE)
-1.r where
-2.0
--
-3.u
-4.0
-9.9 -2.0 -1.0 0.0 2.)
* { k l ~ d u l ( u-~71) + 11 + 2uo71(~1- 71))
-
Fig. 15. V,-plane. ~ D E
~ , ( F , B J e. , ~ l w 2 ~=l *,(elal). p1 = + kl(4 + lY71, kl
Ul = -7Jro
e 3 , = X,(=). -, - r1(e2a2),and fl = ril(Fl).The position
€4,
off, is exaggerated in this figure for clarity. The actual Posi- Q1 = (01 - 71)’ + 1
tion of fl is very close to a,.
XE = YI(Y~ - 01 - PJQ1
Y /p‘
YE = YI[~ - (01 - rl)YQ1.
The eigenvalues,in turn, are afunction of 8. The real eigen-
value Ti,i = 0, 1, is a real solution
to thecharacteristic equa-
tion
T; + (ac; + l)$ + (aci - a + 8 ) +~a8ci~ = o
where co = a,c1 = b. A simple calculation shows that the
complex conjugate pairsatisfy
5; = -(aci + 1 + Ti)/2
5 ; = -(aci - 1 - ~;)’/4 - a’c;/(~; + aci).
This means that given a8, one can compute Al, C,,and El
by finding zeros of polynomials of degree atmost 3 and by
performingtheoperations +, -, x,and +./nprinciple,this
can bedone byhand. However,it would be formidably tedi-
ous. Thecomputer-assistedproof given in accuratelyesti- m
mates the errors incurred by
- 7 1 I I i) finding a zero of a polynomial,
Fig. 16. The annulus region bounded by Sa and SL,.
ii) +,
-, x, t
iii) conversion of a real number to and from the cor-
responding machine represented number.
Finally, if
[ Cl
and
Cl
(8 = 6.5)is outside of Sa
(8 = 10.5) is inside of S b
(2.15)
The last error needs to be takencare of, sincea givendec-
imal number may not be machine-representable. Thepro-
gram in m
accurately gives a lower bound and an upper
bound for every value involved. In particular
the attractor disa pears: (2.4) diverges with any initial con- We will give onlytwo pictures at two different values of
6
dition (see Box 11 )! This disappearing act provokes the
interesting questionas to how the attractor dies. A careful
Cl. Fig. 21(a) shows a 2-torus, while Fig. 21(b) indicates a
ary crisis.
Box shows the attractorat the parametervalue where
thehomoclinicityof (2.4) occurrs. Box depicts the
homoclinicity. Note that the symmetry(2.4) of implies that
homoclinic trajectoriesare present in a pair. Finally, on the
curve”Hopf at 0,” the eigenspace EC(0)changes its stability
type, while €70) is always unstable.
Looking at this bifurcationdiagram, one sees that chaos
can be quenched by makinga sufficiently small, i.e., mak-
ingCl sufficientlylarge, or makinga sufficiently large,when
B is fixed. In the formercase, the trajectory converges to
P , while in thelatter case, the trajectory converges to the
(b)
large periodicattractor [I], [9]. Similarly, chaoscan be
quenched by adjustingfi appropriately when a is fixed. Fig. 19. A simple third-order autonomous circuit which
exhibits a folded torus. (a) Circuitry. (b) Nonlinear resistor
In closing this section, there has been an interesting
v-i characteristic.
recent discoveryof the fact atthat certain parameter values
the saddletype periodic orbit is stabilized into a periodic
attractor [14].
Fig. 23 shows the corresponding simulation results. The rescaled parameters which correspond to the orig-
inal circuit are
D. Analysis
a = 0.07 b = 0.1 p = 1 (3.6)
Let us transform (3.1) into the following dimensionless
form: and Fig. 23(a) (resp., Fig. 23(b)) corresponds to
*dt
= -f(y - x) - z ccl =0
(resp., p1 = 0.027
/LZ
p2 =0
0 / ~ 3z
p3
-0.00675
= -0.1134).
(3.7)
(3.8)
dz
-
dt
= By (3.3) Because no Lyapunov exponent in (3.7) is positive, the sys-
tem is not chaotic. However, sinceonly oneLyapunov expo-
0 1 x) 20 30 ma
' II
L
P
Fig. 28. Driven R-L-Diode circuit. R = 107 Q, L = 2.5 mH,
f = 150 kHz, Diode: 3CC13.
b
B. Experimental Observations
Fig.29showsthetwo-dimensionalPoincar6sectiontaken
at each period T = l / f i nthe (voltage, current)-plane ofthe
diode.
7F (b)
i2 (C)
C. Confirmation -dQ
=I
dr
Although the circuit in Fig. 28 contains only three ele-
ments, its dynamics is rather involved in view of the non-
ifQrO
linearities of the p n junction diode, which are not purely
resistive at frequencies above 100 kHz. A reasonably accu-
rate circuit model of the diode [21] is given byFig. 30, where ifQ<O
both theresistor and the capacitor (Fig. 30(b)) are
nonlinear.
From extensivelaboratory measurements and digital com-
puter simulations, it has been observed [22] that in order
to reproduce thesame qualitative behavior, the nonlinear
ifnTsr< (n+:)T )
resistor in the above model is not essential. Moreover, the
nonlinear q-vcharacteristic of capacitor
the can be replaced
by the drastically simpler two-segment piecewise-linear
curve shownin Fig. 30(c),without changing the bifurcation
pictures.
Fig. 31 showsthe simulation correspondingto Fig. 29.The
cross section, however, is taken on the (charge, current)- where we use Q, I, and r to denote the original circuit
plane insteadof the(voltage, current)-plane, dueto alack variables. Defining the followingnormalized variables:
of time to prepare the material.
D. Analysis
Toanalyzethecircuit,wewillfurthersimplifythedynam- k=-
R a=-
1
p=- 1
(4.2)
ics, and then observe several key
properties of the Poincark Lf LC,f2 LC2f2'
1.
Fig. 32. Simplified circuitwhich capturesessentiallyall the
Using theabove simplified circuit model and solution com-
ponents, we can uncover the essential features of the cir-
cuit dynamics with the helpof the followingobservations:
i) The areacontraction rate is constant and is strictly less
than 1. This stems from the fact that the area contraction
experimentally observed phenomena. rate is determined by the divergence of(4.3), namely
area contraction rate = exp (divergence)
Equation (4.1) can be transformed into where divegence = - k = - R / L f . (4.4)
ii) 0 5 t < 1/2.
cp: cp:
Fig. 33 shows the flows and with a = 0.1, /3 = 10.0.
Each trajectory correspondsto a different initial condition.
I ',
First observethat any solution of(4.3)is made up of com-
I, \ L,
E. Bifurcations
Fig.34. Deformation of the set C along a trajectory for Fig. 37 gives an experimental observation showing the
1/2 s t 5 1. ondimensional bifurcation diagram of thecurrent iof the
circuit of Fig. 28 when the amplitude E of the applied
for 112 I t < 1, the set &C) never hits thei-axis if the initial sinusoidal voltage source is increased periodically from0
rectangle A in Fig. 33 is chosen appropriately. to7.7VIEismodulated byasawtoothwaveform).Eachpoint
Combining the above three observations, we see that in this "bifurcation tree" represents a one-dimensional
during the period0 It < 1, rectangle A stretches, folds, Poincark section taken at each fundamental period T =
and eventually returnsto the originalregion D. Extensive l l f o f the sinusoidalsource. There are
two striking features
numerical observations show that we can choose appro- in this bifurcation tree:
priateA andD such that A 3 D. During thistransformation i) A succession of large periodic windows the periods
process,theareaofAiscontinuallybeingcontracted.Ifthis of which increase exactybyone as we move from any
mechanism is repeated many times,it can give rise to a very window to thenext window to the right.
complicated behavior, such as chaos. Fig. 35 gives a global ii) A succession of chaotic bands sandwiched between
picture of this transformation over one period of the flow the large periodic windows.
Ot.
2) TweDimensional Map Model: Based upon the pre- The crosssection in Fig. 29 corresponds to E = 6.2 V, i.e.,
ceding observations, we propose a surprisingly simple two- thefivechaoticbandsofFig.37correspondtothefive"legs"
dimensional mapmodel whichmimics the transformation of Fig. 29.
described in Fig. 35. Fig. 36gives a more precise description Let us examine how the simple map (4.5) captures the
of the transformation mechanism. A simple two-dimen- essential featuresof the bifurcation phenomena observed
sional map which transforms thesquare STUV in Fig. 36(a) experimentally from the R-L-Diode circuit. Fig. 38 shows
into the lambda shaped set in Fig. 36(d) is described by the one-parameter bifurcation diagram of xfor (4.5) where
a,x, if x, 2 0) a, = 0.7 b = -0.13
x,+, = Yn - 1+ and a2 is varied over the range
-+x,,, if x, < 0
0 5 a2 5 20.
a2 X
38.
Fig. One-parameterbifurcationdiagram ofxforthetwo- Fig. 39. Attractor observedfrom the two-dimensionalmap
dimensional map model where 0 5 a, I20. model at a, = 18.0.
Fig. 39 shows the attractorin the( x , y)-plane corresponding larger valueat the topposition. The’knapshots” i n column
to A show the corresponding experimental observations taken
from the original R-L-Diode circuit as E increases from the
a2 = 18.0.
bottom. The four insets in column Care enlarged pictures
Note that the attractor is qualitatively identical to the one in a small neighborhood of the periodic pointP4A (of the
obtained experimentally in Fig. 29. two-dimensional map) identified by the solidtriangles A.
A detailed analysis of (4.5) can be performedbecause of Wecan nowgiveacomplete pictureofwhatis happening
its simplicity. Based upon the bifurcation analysis of (4.5) in the original circuit.
one can understand the bifurcations of the original circuit. i) Let us begin with the picture at the bottomi n column
Fig. 40 shows the detailed bifurcation mechanism associ- Band lookat the folded object. *
The symbol identifies the
atedwith theperiod4window.Bifurcationsassociatedwith location ofthe fixed-point Q o f (4.5) which is a saddle point
otherperiodicwindows have similarstructures. The for the present parameter range. As we increase the value
sequence of drawingsin column B of Fig. 40 shows how the of a2( E in the originalcircuit), a saddlenode bifurcation of
attractor of the two-dimensional map model is deformed period4 takes placeoutside the region wherethe attractor
as a2 is increased from its value at the lowest positionto a lives. This period4 orbit has a strong influence on the
ACKNOWLEDGMENT
!I’ 2
[15] W. F. Langford, Numerical Studies o f Torus Bifurcations
. (International Seriesof Numerical Mathematics, vol.70) Hei-
delberglNew York: Springer-Verlag, pp. 285-295.
[I61 J. Stavans, F. Heslot, and A. Libchaber, “Fixed winding num-
ber and the quasi-periodic route to chaos in a convective
fluid,” Phys. Rev. Lett.,vol. 55, no. 6, pp. 596-599, Aug. 5,1985.
[ I 7 T. Bohr, P. Bak, and M. Hogh Jensen,“Transition to chaos by
interaction ofresonances in dissipative system II.Josephson
junctions, chargdensity waves, and standard maps,” Phys.
Rev. A, vol. 30, no. 4, pp. 1960-1969, Oct. 1984.
[I81 M. Sano and Y. Sawada, ‘Transition from quasi-periodicity
to chaos in a system of coupled nonlinear oscillator,” Phys.
Fig. 41. Chuang Tsu’s story of chaos [29]. Lett., vol. 97A, no. 3, pp. 73-76, Aug. 15, 1983.