approach is used for separating and selecting the needed signal. The pulse
width is selected to be equel to the group transition time
τ
through the SSR
and the modulation frequency is selected to be
2
/
(
N
+1)
of the SSR eigen
frequency.
In the passive SSR, the CW and CCW light beams can only propagate
2–4 round trips, after that, they will be shadowed by the shot noise of the
PD and can not be detected. Therefore, in the case of passive SSR,
N
= 4
is chosen. In the setup, the length of SSR is 200 m, and the pulsed phase
modulation frequency is 200 kHz. The time interval between the two pulses
is
Nτ
.
The phase biasing is realized by the MIOC. The peak voltage is
designed to be equel to the half-wave voltage of MIOC for alternately
realizing the optimal phase bias of
±
π/
2
.
In order to improve the Signal-Noise-Ratio (SNR), the pulsed phase
modulation has been modified by using a pair of positive and negative
modulation pulses with the same amplitude and width
τ
. In this case, a
subtrator should be used in the ADCL for processing the paired signals.
As mentioned before, the small couplng ratio
α
will lead to waste
the SLD output light power and cause the saturation of the pre-amplifier
circuits after the PD due to the increment of the DC component. In this
case, the Sagnac effect signals of light beams after 3–4 round trips will be
difficult to be detected. Therefore, in the setup the pre-amplifier has been
re-designed by adding a special cascade for completely cutting off the DC
component before amplification.
In order to investigate the performance of the proposed system for
MOG, a specially designed ADCL electronics has been developed, in which
the above mentioned phase modulation approach is realized. The setup with
the ADCL electronics provides the digital readout as a gyro.
In the drift rate test, the samples of the Sagnac phase shift signals
after 2 round-trips are selected with the sampling frequency equal to
the modulation frequency (200 kHz). In every eight sampling periods,
one digital data processing cycle has to be completed for providing the
digital readout data. The data processing cycle includes signal sampling,
demodulation, integration and digital ramp signal generation.
The random drift rate of the setup was tested with the integration time
of 10 s and the sampling length of 0.5 h. The standard deviation of the drift
rate is 1.25 deg/h.
The same drift rate test has been carried out on the setup, but with the
additional monitoring PD (M-PD in Fig. 1). As shown by the test results,
with the same integration time and the same sampling length, the standard
deviation of the drift rate has been reduced to 0.75 deg/h.
ISSN 0236-3933. Вестник МГТУ им. Н.Э. Баумана. Сер. “Приборостроение”. 2005. № 4 113
