INTRODUCTION
The reliability of a system is affected by the reliability of its components
and the way they are interconnected to serve its intended mission under
certain operating conditions. It is well known that the reliability of
discrete components, analog circuits and their associated interconnects
is very low. The reliability of a system can be improved if the reliability
of its constituents could be increased. This can either be done by using
more reliable components, for example, MIL SPEC parts or by redesign of
parts of the system using newer technologies with a higher reliability.
We have chosen the second approach in this research project and used integration
of several analog circuit boards by using an FPAA.
Filed Programmable Analog Arrays (FPAA) which are the analog counterparts
of Filed Programmable Gate Arrays (FPGA) for digital circuits can be used
to integrate various analog boards in existing electronic systems in order
to enhance their reliability. This can also result in a smaller, lighter,
cheaper and more robust system. The performance of the system could also
be improved. Modern FPAA`s enable one to translate complex analog circuits
to a set of low level functions and describe analog functions such as
filters and gain stages without worrying about components such as op amps,
capacitors, current mirrors, etc.
Analog signal conditioning and processing is used in a variety of applications
including sensor interfaces, industrial controls, medical monitoring,
seismic systems and laser control. In most applications the analog signal
conditioning is done with discrete components, analog ASSP/ASIC, or a
DSP following digitization of the signal. However, all signal conditioning
functions such as linearization, summation, rectification, phase detection,
threshold detection and integration can be integrated within a single
FPAA device (Fig. 1).
The new FPAA technology provides an elegant way of implementing these
designs with the added benefit of reconfiguration and improved reliability.
Looby and Lyden (2000) proposed a new continuoustime FPAA architecture
which simultaneously achieves bandwidth and repeatability comparable to
the accuracy tolerance of switchedcapacitor FPAAs and the bandwidth of
continuoustime FPAAs. Even neural network applications could be realized
with FPAAs. Lee and Gulak (1991) presented the design details and test
results of a Fieldprogrammable Analog Array (FPAA) prototype chip in
1.2 μm CMOS based on subthreshold circuit techniques which consists
of a collection of homogeneous Configurable Analog Blocks (CABs) and an
interconnection network. Interconnections between CABs and the analog
functions to be implemented in each block are defined by a set of configuration
bits loaded serially into an onboard shift register by the user. They
have even developed macromodels for the analog functions in order to simulate
various neural network applications on the fieldprogrammable analog array.

Fig. 1: 
The block diagram of an FPAA 
The key benefits of an FPAA over a fixedfunction solution are:
• 
The design process is simplified 
• 
There is a onecomponent solution for multiple designs, greatly
simplifying inventory management and repairability which also improves
reliability 
• 
The integrated design provides design specifications that are immune
to temperature, process and component aging 
• 
There is precision operation and increased system reliability 
Becker et al. (2008) presented a FPAA to implement a unique hexagonal
topology of 55 tunable OTAs as a prototyping environment for rapid reconfigurable
analog signal processing.
There are even possibilities for developing mixed signal solutions using
FPAAs where there is a mixture of both digital and analog signals. In
such cases the usual solution is the use of analog circuits for interfacing,
digital circuits for control and computation and DSPs for signal processing.
However, the disadvantage of this type of solution is that DSPs use a
lot of power. In such situations FPAAs can be used to implement the analog
part of the system while FPGAs could be used for the digital parts shown
in Fig. 2.
The analog computer of the gyrocompass in the navigational system of
a naval vessel is usually used to find the proper direction of the north
and corrections for the horizontal direction. It uses the information
it receives from the accelerometers and gyroscopes to calculate the necessary
settings for the motor drives for the stable platform. Reliability is
an important issue especially in sensitive applications like the system
under study.
The gyroscopic navigational system of the naval vessel under study is
exactly a mixed signal system. We proposed to integrate its analog computer
with FPAAs and its digital parts with FPGAs. The integration of the analog
computer using FPAAs was carried out and its effect on the reliability
improvement of the system was shown. Further work should be funded in
order to carry out the integration of the total gyroscopic navigation
system.

Fig. 2: 
An FPGA/FPAA solution for reliability improvement of
mixed systems 
Navigational information and control signals used in the navigation of
naval vessels include the vessels direction, range, roll, pitch, etc.
Each naval vessel should be equipped with navigational equipment. The
gyroscopic compass system on a naval vessel is used to provide such signals.
This system under study works based on the input information and a mathematical
model already programmed in its analog computer and provides analog and
digital outputs for position, roll, pitch, etc. in real time. The input
to this system is the ship`s speed, the speed of the currents, longitude
and latitude as well as the outputs of the accelerometers.
This provides a reference plane on board the ship which always follows
a certain direction relative to the Earth despite any movements by the
ship. These directions are usually the geographical north/south, east/west
and local horizontal direction. Therefore, we always have a coordinate
system on board the ship using which the measurement of roll, yaw or pitch
are easily measured. The analog computer calculates the position and establishes
the proper latitude and altitude using the initial latitude and altitude
as well as the inputs from the speedometers and the accelerometers.
THE VARIOUS TYPES OF FPAA
There are various techniques used in the manufacture of FPAAs including OTA,
current conveyor, bipolar and operational amplifiers. FPAAs usually operate
in one of two possible modes: continuous time and discrete time. There are several
manufacturers such as Anadigm and Zetex which produce FPAAs. We used Zetex TRAC020
series totally reconfigurable analog circuit in this research which is a BiCMOS
and Bipolar chip that works in continuous mode. It can be programmed using a
hardware programmer which can program an EPROM and 4 TRAC 020 chips (Zetex Semiconductors,
1994).
INTEGRATING ANALOG PARTS WITH FPAAS
An analysis of analog circuits shows that they are usually composed of
parts such as amplifiers, attenuators, log and antilog, inverters, summers,
buffers, differentiators and integrators. Zetex`s Totally Reconfiguarable
Analog Circuit (TRAC) offers facilities to implement the analog circuit
using a set of configurable analog blocks plus a programmable interconnection
network. Zetex`s TRAC chip has a 20 cell, continuoustimebased architecture
based on logarithmic amplifiers. There is a design tool with which a designer
can work at a high level and does not need to know anything about the
underlying analog circuits. It offers functions noninverting pass, negate,
add, log, auxiliary, alog, rectify, Log/alog and Log/rec with which one
can implement analog circuits. One can also use the “OFF”
function to not use a given cell. One can use TRAC and develop a solution
for a signal management problem at hand rapidly.
RELIABILITY OF THE CIRCUITS UNDER STUDY
Reliability of electronic circuits has been a major concern especially
in important subsystems of naval vessels and military apparatus. The reliability
is the probability of successful operation during the mission and under
prespecified conditions. Since naval vessels must operate at sea, the
general conditions for which electronic subsystems of naval vessels are
evaluated are the Naval Sheltered and the Naval Unsheltered as per MILHDBk217F
(1995). The failure rates for the parts which make up the electronic circuits
can also be either estimated based on the generic rates of MILHDBK217F
or other data sources. The reliability can be calculated using various
techniques including RBD, Markov state space, analytical, or Monte Carlo
Simulations. We used the RBD technique along with failure rate data from
MILHDBK217F. For example, the general failure rate for a resistor is
λ_{p} = λ_{b} π_{T }π_{p
}π_{S }π_{Q }π_{E} where,
λ_{b }is the base failure rate, π_{T} is the
temperature factor, π_{p} is the power factor, π_{S}
is the power stress factor, π_{Q} is the quality factor and
π_{E} is the environment factor. There is a similar relationship
for other devices with appropriate factors to include stresses and operating
environment.
EXPERIMENTAL RESULTS
The circuits in the analog computer mostly consisted of metal film resistors,
cermet trimmer potentiometers, polyester capacitors, glass diodes, transistors,
opamps, regulators plus the printed circuit and interconnections.

Fig. 3: 
The failure rate of the analog computer in the naval
unsheltered case before and after integration using FPAA (1 is the
case before and 2 is the case after integration) 

Fig. 4: 
The failure rate of the analog computer in the naval
sheltered case before and after integration using FPAA (1 is the case
before and 2 is the case after integration) 
The circuits were integrated into a new one using several discrete components
and TRAC020 FPAA. The system was first integrated using software to redesign
the I/O interface. Then it was implemented and tested. The new system
was much smaller in volume and weight and much more reliable than the
original system. The data from MILHDBK217F was used for the failure
rates of the electronic parts of the system under study. We obtained the
following results:
λ_{NU} 
= 
180.203242 FPMH 
λ_{NNS} 
= 
54.04694 FPMH 
MTTF_{NU} 
= 
5548.391349 HRS 
MTTF_{NS} 
= 
18502.43511 HRS 
These results are shown in Fig. 36 to compare the
reliability measures for the analog computer before and after integration
by using FPAAs.

Fig. 5: 
The MTTF of the analog computer in the naval sheltered
case before and after integration using FPAA (1 is the case before
and 2 is the case after integration) 

Fig. 6: 
The MTTF of the analog computer in the naval unsheltered
case before and after integration using FPAA (1 is the case before
and 2 is the case after integration) 
CONCLUSIONS
In this study we have presented the results of integration of the various
circuits of the analog computer of the gyrocompass of a naval vessel and
showed the improvement in reliability. We showed a 16.67% improvement
in the Naval Unsheltered case and a 18.2 % improvement in the Naval Sheltered
case. The comparison of the MTTF from Fig. 36 also
show an improvement in reliability. Since the gyrocompass of the naval
vessel is composed of several analog modules similar to the one presented,
plus several other digital modules, we can integrate the whole mixed system
using several FPAAs and FPGAs. We will gain a great improvement in its
reliability measures such as MTTF, MTTR and MTBF and achieve a much more
reliable system with a smaller weight, volume and cost plus better performance.
This has been proposed as the next step in our research.