High-precision measurement scheme for half-wave voltage

of Y-tap MIOC
Yuanhong Yang Weixu Zhang Jing Ma
The 5th Research Division, Beijing University of Aeronautics & Astronautics
P0 BOX 85, 100083, Beijing, China

ABSTRACT

A high-precision measure scheme for half-wave voltage VR of Y-tap MIOC is proposed. This scheme is based on Sagnac
interferometer and a saw-tooth wave with special period is used. With this scheme, the temperature characteristic of V of a
type of MIOC is studied experimentally. The result shows that its half-wave voltage is temperature-dependent. In the range of
-10 to +55 , the varying value is O.162V, the temperature coefficient is 662 PPM/°C and the variation is linear and
repetitive. Experimental study and detail discussion demonstrate that high accuracy can be achieved with this scheme and it is
very suitable to be used in studying and calibrating V ofthe modulator used in Close-loop FOG in-site.

Key words: half-wave voltage, temperature coefficient, integrated optical circuit; fiber-optic gyroscope

1 INTRODUCTION

Y-tape MIOC (multifunction integrated optical circuit, or Y-tape multifunction integrated electro-optical phase modulator) is
one ofthe key components of close-loop fiber optic gyroscope (FOG). It take the roles of mode filter (polarizer), splitting and
combining light (directional coupler) and phase modulator in FOG optical path. The basic configuration is shown in Figi,
where wave-guide is fabricated in LiNbO3 with annealing proton exchanged (APE) technology, while electrode with coating

PM fiber wave-guide electrode LiNbO3 substrate

Fig 1 Configuration ofY-tap modulator used in

FOG
and light inputloutput with polarization-maintain(PM) optical fiber. For an electro-optic component, V is a very important
parameter which variety will bring nonlinearity to the output of close-loop FOG directly. So, It is very significant to measure
it and defme its variation model accurately. There are a few measure schemes'121 reported now. But they are all based on
polarization interference or Mach-Zehnder interferometer and can not achieve high accuracy. To satisfy the demand of high-
precision measuring and modeling, a scheme based on Sagnac interferometer is proposed. The experimental study and detail
discussion was made meanwhile.

2 PRINCIPLE

Figure 2 is the measurement scheme diagram for V of MIOC, where SLD is superluminescent diode; DC is directional
coupler. As shown in figure 2, SLD, DC, detector, fiber coil , and the MIOC tested compose a Sagnac interferometer. When
the interferometer is working, the output of detector will be:
v0=K(1+cos) 1
where, K is the coefficient related with incident optical power on detector and gain of detector unit. p is nonreciprocal phase
shift and it is zero without nonreciprocal modulation or noise. The variation of V0 due to p is shown as wave-form (a) in
figure 3. When a saw-tooth wave-form with amplitude Vs and period 4T(Tis the time demanded for light transmit across the

Part of the SPIE Conference on Advanced Photonic Sensors and Applications

436 Singapore • November—December 1999
SPIE Vol. 3897 • 0277-786X/99/$1 0.00

Downloaded From: http://reviews.spiedigitallibrary.org/ on 04/04/2014 Terms of Use: http://spiedl.org/terms

 fiber coil) is added to one of the two modulation arms, the nonreciprocal phase shift will be occur because the modulation arm
is at nonreciprocal position and the clockwise and counter clockwise light go through the arm and be modulated at different
time due to the time delay T induced by fiber coil delay line. This can be equivalent to the phase modulation wave-form of
counter propagating lights are shifted in time axis with each other. The time difference shifted is T and the nonreciprocal
phase can be calculated using following equation131:

MIOC tested Fiber coil V2

(d)

detector

- —Lr--U-L
oscillograph Signal generator

(b)
Fig2 Measurement scheme diagram for V Fig3 Wave-form during measuring

m(t)4)m(tt) 2
where, 4m (t) is the modulation phase wave, which is saw-tooth wave here. In figure 3 waveform (b), the solid lines represent
m(t) , while the dash lines represent m(t_ T) . Because the half-wave voltage to be measured is V , the modulation
coefficient Kp and the slope Ks of saw-tooth wave added are respectively:

K=,54 (3)

Ks/' (4)
During a modulation period, when O<t<T, the nonreciprocal phase is:

1 = —K K . .
3t = — (5)

the detector output is

vi = I([1 + )] (6)
when i<t<4T the nonreciprocal phase is:

2=K.K.t=YL (7)

the detector output is

V2 =K[1+cos(42)] (8)

This result can be seen in figure 3 waveform (b) (c)and (d) directly. In order to make equation V1 =V2 satisfied, combining
equation (5)—{8), the first non-zero solution of V can be got:

v=vS 9
This means, when V1 = V2 , the value of V, can be calculated with equation (9) because Vs is set artificially and known. So,
the measurement procedure can be summarized as:
(1) Mounting the testing setup according to figure 2;
(2) Set the period of saw-tooth as 4T;
(3) Increasing the voltage Vs from low to high and monitor the variation of V and V2;
(4) When V1 = V2, the V is half of Vs.

437

Downloaded From: http://reviews.spiedigitallibrary.org/ on 04/04/2014 Terms of Use: http://spiedl.org/terms

 UqM 'A = A ' ' = '€ / 'Z = / 'Z ''A tA ' £IA AWSUS1TM TJ UO!!JA SAJO pu 1Tt4 DuO1JJTp si
AV = —
'A = (-----—)uis (-—-)uis oi
Ai7 AZ
1TJ fljA JO AV ST pOSfl O UuIUOp JjA uounb (6) si P!JSS utT/t 'AZ + = AZ O1qM SA OUIflSS SA SAV SAV
si ion -wp SAV ,/ AZ S £IA 'u'-" ti uoiunb i)uis , (A si pjsqs XPuIxo1dd uounb
,i SAV =
(çj) uo :q
(-_-_)uis)J=Av TI
AZ
siq UoUnb SAOqS MJ2 flJA J0 SAV 51 1UWLI /q UO!3Op 13E1noD JO AV • 5fltOOH ST £IA 'qq
AV si £IA Awsus sAV oo u 'pj sup UO!mIS Si £IA JElituis o
jgus rnssooid m dooi-sojo ft]DOJ pu
ornoo jo juis uouop rn dooj-soio oo uo q psn o uJus ip Aomoo JO sAV UI dooj-soio 'ooj t q,0 •

oinoo uo q pAHjO /JTSE pu p spuodsuoo o 901 pi srnd uoup iomoo Xjoruuxoidd uoinb i) '(1
sup suw SAV i (Az) = 9OI • A si Auo i MOJSJOA SAV jJIM q qonw •JIIUIS LL SEISUOWp 3J2 sap WJOS AE4
ij IEiuoodJO UIATt[OE qtj .IcoJnoo
iainiaxa
v jo
Io si "-ujj
jjj sq uq poinsw qi
dAi jo qd1UojiOso pU1 jUS 1OJUO
suj wqos ojj
OZZScIJ. jU1
dms Si uoqs rn
dH voz i(IAtOdS1
'z w' n quII jo iq
AtJJ 1E 'Tp
O6 N
a
a
• 1EP p,s1
. UiJ A-m'

.
9WE

f7•
E z8
o8.
8L
9L
a
—

rn- o ot oz oc: Øj? oc 09

1TuEIdu1L (3)

fIJ iUfl3JO AM-JEq 1jOA intidw SA

ozzsui si psn o ioituom &IIJ(IA JO 'A ' ZA IN" £dH yØJ IOJ OO-M5 AM Ut1JOU pui UTAI tJ fljEA JO
qj
'A osnoq si no-poi Aomoo si 'Awl ws/s uuxoinstw AoEJnooSt AWT OO T'' !4 dns UI iuwtidx it
'ompooid jyj,s st Ajuo punow uj g oinEJdui 'xoq A si pmsow 'wiod Ins1 ujjp inidw
mj
si uoqs ui j stqj OS A si intidw iupudop qj &1IA1A fltEA SI AZ9TO UT OTfl U1 ØJ-JO o cc+ PUB
op inidw
J
jrnuiouliod
uoijjooo si Z99 /JAdd •
quit' q2 E1E 'pOS TJ UOI21IJA si
nsai spuodsuoo
punoj irnj pu
qj
nsai
i.p iocbi
jo Auu.i smi
•[j7] iojy inbs
suuiudx AOid U AUtdOJ
qt iniq sj
.OO

I' NOISSfDSI

woJd(J) uouEnbo (6)—(c) pu j) '(j

tA = A AV
_ pu' j
ouj
jo oq
si
mooo m uoijnpom pouod - noq
pnoq s snoouiuisut 'o Au jEoo1d!oiuou sEqd jns ponprn Xq iro '1OPJ qOflS S1 JEUUIUO1IAU miodm
u sri qj uunsw uto q j
sjd oouaijjtp uoiq '4 pu tp )pUI uotEnb(6) 'p!Jsus

8E17

Downloaded From: http://reviews.spiedigitallibrary.org/ on 04/04/2014 Terms of Use: http://spiedl.org/terms

 varying, fiber coil sensing rotation, etc., will not affect the measuring because these sort of varying is low frequency, the
variation of 4 and 42 w'be equal during a modulation period. With the same reason, the measuring is insensitive with
the small variation of K induced by optical path loss change or source power change in equation (1). This is why high
precision can be achieved in above experimental setup, where the MIOC tested was set only in temperature change
environment.
(2)As discussion above, when the period T of saw-tooth wave is 41, the measuring will achieve high accuracy. But this is not
necessary for all measure, the period can be set as other value too. Assuming 1>1, we have:

K=
1 = —K •K .(Tt) = —(T—t) (12)
VT
(13)

(14)

when V = 2V , the phase difference of 4 and 2 is still 2rr and V1 = V2 . To the contrary, we can still use the voltage
difference z\V to determine whether the equation (9) is satisfied and calculate V with equation (9). But at T = t I n

'
(n1 ,2,3 . . .), this scheme can not operate because the phase modulation wave is superposition and the phase difference of
and 2
S always zero. When T<i and T r I n , the same resulr as equation (14) can be obtained with the same
derivation procedure too.
(3)In this scheme, the voltage difference between V1 and V2 is used to determine whether equation (9) is satisfied. Further, it
can be used as feedback error signal to control the amplitude of saw-tooth wave through a set of special electric circuit
completely, and then the auto measuring can be realized. The circuit is close-loop and can be illustrate as figure 5, AVis
feedback signal used to control the amplitude of saw-tooth modulation signal and make 42 —4 = 2ir and equation (9)
satisfied. For comparing, draw the signal circuit ioop of close-loop FOG in figure 6, where another square wave signal is
added to the modulator, V and V is the output correspond to the positive and negative half of a period. Their difference
LXV' feedback too to control the frequency of saw-tooth modulation signal to make the total phase T =
is keep
zero, where c1 is the sagnac phase duo to rotation, is the feedback phase. Comparing the two loops, we can find they
are very similar and the hardware will be same when this loop is realized with digital circuit. This demonstrate that the
scheme can be taken to realize the self-checking and calibrating on V in close-loop FOG easily.
LW
I detectorl V2-V1 04 Integral

MiOck —J signal generator

Sagnac interferometerj I (changing amplitude]

Fig 5 measuring loop for V,

Fig 6 signal proceeding loop of FOG

439

Downloaded From: http://reviews.spiedigitallibrary.org/ on 04/04/2014 Terms of Use: http://spiedl.org/terms

 (4)In figure 2, the MIOC tested take the role of splitting and combining the light of optical path too. Ifuse another directional
coupler to take place the function, we can measure V of other sort optical phase modulator (PM), such as single channel
integrated electro-optical phase modulator and bulk phase modulator etc. The setup is shown in figure 7, where PM is
phase modulator.

DC DC Fiber cod

ctor .

L1i
1 P?:1 to be tested
I
— —u_u-t_ +
— v- I

oscillograph i- ignaI generatorj

Fig 7 Measurement Scheme diagram for Vuof other type of phase modulator

5 CONCLUSION

Experimental study and detail discussion demonstrate that the scheme proposed can measure V of Y-tap MIOC and other
sort of phase modulator conveniently and accurately. The highest accuracy can be achieved with period 4i and other period
value except for -r I n can be used to do the measurement too. This scheme is very suitable to be used in studying and
calibrating V of MIOC used in Close-loop FOG in-site. Temperature experiment on V of MIOC tested demonstrates V is
temperature-dependent and the variation is linear and repetitive.

REFERENCES

[1] Geng Fan, Shong Qiang, Li Caifen. Half-wave voltage measure with polarization interference method. Proceeding on
FOG , (In Chinese) 1990.
[2] Yang Yuanhong, Ma Jmg, Zhang Wei-xu. The PZT optical fiber phase modulation testing with varying amplitude
modulation. CHINESE JOURNAL OF LASER, Vol.B3, No.6, Dec.1994.
[3] Herve Lefevre. The Fiber-Optic Gyroscope. BostonLondon, Artech House, 1993, 33;l 11-126.
[4] George A. Pavlath. Close-loop fiber optic gyros. Proc. Fiber Optic Gyroscope: 20th Anniversary Conference. SPIE Vol.
2837, P46-60.

440

Downloaded From: http://reviews.spiedigitallibrary.org/ on 04/04/2014 Terms of Use: http://spiedl.org/terms