-
[show abstract]
[hide abstract]
ABSTRACT: This paper deals with the implementation of a CMOS analog neural network (NN) that has to be integrated in a new kind of optoelectronic measurement system. The aim is to achieve real-time surface recognition using a phase-shift rangefinder and a neural network. NN architecture is a multilayer perceptron (MLP) with two analog input signals provided by the rangefinder, three processing neurons in the hidden layer, and one output neuron whose output voltage indicates the detected surface. As the complete structure is analog, no analog-to-digital conversions or signal processing between the rangefinder and the network is necessary. Furthermore, the 3.3-V voltage supply, relative to the chosen CMOS 0.35-mum technology, allows to reduce the system power consumption. This paper focuses on the implementation in an ASIC of an elementary part of the NN, called neuron, and on the achievement of the complete NN from the integration of three ASICs in a printed circuit board. Comparisons between ideal case, simulations and tests are detailed in order to validate the design and the good functioning of the complete structure.
IEEE Sensors Journal 09/2008; · 1.52 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: An analog neural network (NN) was developed for real-time surface recognition by using two photoelectrical signals issued from a phase-shift rangefinder. The NN architecture consists of a multilayer perceptron (MLP) with two inputs, three neurons in the hidden layer, and one output. The NN output is compared with threshold voltages in order to classify the tested surfaces. In this type of application, analog NN implementation has many advantages, especially the small silicon area used, a low-power consumption, and no analog-to-digital conversions. This recognition system has been successfully tested for four types of surfaces (a plastic surface, a glossy paper, a painted wall, and a porous surface), at a remote distance between the rangefinder and the target varying from 0.5 m up to 1.25 m and with a laser beam incidence angle varying between and . This paper presents the NN training and the experimental tests of surface discrimination.
IEEE Sensors Journal 11/2007; · 1.52 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The linearity of CMOS buffer amplifiers is discussed. It is shown that enhancing the amplifier gain to reduce non-linearity presents limitations due to small signal gain mismatches between the two amplifier inputs. An amplifier structure and associated design precautions are proposed to overcome these limitations to some extent. As a result, a fully differential buffer used in a 12 bit 10 MSamples/s video chain has been successfully designed to achieve 16 bits linearity in a low-voltage CMOS 0.35 mum process.
Instrumentation and Measurement Technology Conference Proceedings, 2007. IMTC 2007. IEEE; 06/2007
-
[show abstract]
[hide abstract]
ABSTRACT: This paper deals with the design of an analog neural network (NN) that has to be implemented in a new kind of optoelectronic measurement system. The aim is to achieve real-time recognition of different kinds of surfaces using a phase-shift rangefinder and a neural network, at a remote distance between the rangefinder and the target varying from 0.5 m up to 2 m and with a laser beam incidence angle with regards to the target varying between -pi/5 and pi/5. The NN architecture is a multi-layer perceptron (MLP) with two inputs, three processing neurons in the hidden layer and one output neuron. Both inputs receive analog signals directly issued from the rangefinder. The output neuron gives a voltage that indicates which surface has been detected. The ANN design presented in this paper is completely analog in order to avoid signal pre-processing before the network and to limit power consumption. The 3.3 V voltage supply, relative to AMS CMOS 0.35 mum chosen technology, allows also a low system power consumption but important optimization of each circuit design is required. Each cell transfer is achieved with differential currents. Sigmoid function bias current is set outside the chip in order to control the tangent hyperbolic cell linearity range. NN global consumption is equal to 15 mA, with 4.4 mA per neuron. The neuron size is about 276 mumx 148 mum.
Instrumentation and Measurement Technology Conference Proceedings, 2007. IMTC 2007. IEEE; 06/2007
-
[show abstract]
[hide abstract]
ABSTRACT: The feedforward multi-layer perceptron (MLP) type neural network (NN) presented in this paper has been developed for on-board applications of high-speed signal processing (240 MHz). It is fully analog in order to avoid analog-digital conversions and to limit chip size and power consumption. It is constituted by a single input, ten neurons in the hidden layer and a single output. The MLP-NN has been implemented in a 84 pins -0.6mum CMOS ASIC. The NN layout size is 1.8mmx0.7mm and the consumption is intended less than 600mW. This paper presents the design and simulations of each implemented cell and the first experimental tests achieved on the implemented ASIC.
Electronics, Circuits and Systems, 2006. ICECS '06. 13th IEEE International Conference on; 01/2007
-
[show abstract]
[hide abstract]
ABSTRACT: By analyzing the classic clamping circuit, it is shown that the analog switches limit the precision of such a circuit, to 12 bits for 10Msamples/s signals, using a low-voltage CMOS 0.35mum technology. This paper presents an improved clamping architecture that overcomes these limitations. Precisions on the order of 16 bits and more can then be reached, the operational amplifiers being the new limiting factor
Instrumentation and Measurement Technology Conference, 2006. IMTC 2006. Proceedings of the IEEE; 05/2006
-
[show abstract]
[hide abstract]
ABSTRACT: This paper presents a digital capacitor error-averaging calibration technique for pipelined analog-to-digital converters (ADCs). The scheme utilizes a 1.5b/stage digital calibration algorithm and extends it to assess capacitor ratio-mismatches. This is accomplished by using an algorithm similar to analogue capacitor error-averaging (Song, 1988) which is performed in the digital domain. This calibration scheme is demonstrated for a 12-bit pipelined ADC, with 1.5b/stage architecture and 14 identical stages. Gain errors induced by capacitor mismatches (a equiv 3%) are successfully corrected by the proposed algorithm
Instrumentation and Measurement Technology Conference, 2006. IMTC 2006. Proceedings of the IEEE; 05/2006
-
[show abstract]
[hide abstract]
ABSTRACT: A fully differential high-speed high-resolution CMOS track & hold amplifier (THA) is designed for CCD analogue front-end (AFE) circuits. First THA structure is described and theorically analyzed in order to extract sub-circuit requirements. Then, low-voltage, high-frequency, low-power, nearly rail-to-rail fully differential 3.3V CMOS op-amp solution, used as THA sub circuit, is described. Finally, conception and implementation of an expected 14-bit, 30-MS/s CMOS THA in BiCMOS 0.35μm 3.3V is developed.
Signals, Circuits and Systems, 2005. ISSCS 2005. International Symposium on; 08/2005
-
[show abstract]
[hide abstract]
ABSTRACT: This paper presents a current memory cell (CMC) which can be used as basic elements of current pipeline analog-to-digital converter stages. This CMC is based on a fully differential structure which uses the Miller effect to reduce charge-injection errors. Using a 0.35μm 3.3V CMOS process, results show that the signal-dependent charge-injection error is less than 20nA for [-200μA;200μA] dynamic input current range. The acquisition time for a 200μ input step transition to achieve a 14 bit settling accuracy is 22ns. The active chip area and the power consumption of the proposed CMC are about 0.042mm<sup>2</sup> and 6mW, respectively.
Signals, Circuits and Systems, 2005. ISSCS 2005. International Symposium on; 08/2005
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes a new differential current memory cell which is based on the Miller effect and an interleaved ADC architecture. As far as the CMC is concerned, it was pointed out that the opamp design is not an issue to achieve high performances. Simulation results show good performances for sampling rates up to 20MS/s (assuming 22ns for both sampling and holding mode) and relatively large current input signals of mnplus200muA. The originality of the proposed interleaved ADC architecture consists in improving the ADC sampling rate with minimum penalty to SNR performances
Instrumentation and Measurement Technology Conference, 2005. IMTC 2005. Proceedings of the IEEE; 06/2005
-
[show abstract]
[hide abstract]
ABSTRACT: In this paper, wideband S/H amplifier design for CCD signal processor application is shown. First, asymmetric common closed-loop S/H amplifier architectures are analysed in order to define internal limitations for high frequency applications. Then, low voltage, high frequency, low power, rail-to-rail 3.3V CMOS OpAmp, used as S/H building block, is described. Finally, full conception of a S/H amplifier is developed.
Instrumentation and Measurement Technology Conference, 2004. IMTC 04. Proceedings of the 21st IEEE; 06/2004
-
[show abstract]
[hide abstract]
ABSTRACT: The paper focus on the behavior of the NN MLP predistorter when real adders or multipliers generate impairments integrated in CMOS technology. Simulations show that the analog NN predistorter reaches always very good performances even with severe voltage offsets at the input of the different MLP integrated elements. For satellite communications, the on board high power amplifier is a highly nonlinear amplifier. Nonlinearities of CMOS implemented multipliers adders and impairments of integrated hyperbolic tangent are taken into account by the learning algorithm and, also, no significant peformance degradation is obtained. The microelectronics aspects of the MLP integration are also presented.
Information and Communication Technologies: From Theory to Applications, 2004. Proceedings. 2004 International Conference on; 05/2004
-
[show abstract]
[hide abstract]
ABSTRACT: A feedback laser driver circuit has been developed to control the average optical output power of a double heterostructure CW laser diode and also to operate with a wideband amplitude modulation. The average light power monitoring is realized by measuring the photoelectric current of a photodiode integrated inside the package of the laser diode. Safety features, including transient signal suppression, protect the laser against excessive light power. To obtain responses that are not out of the absolute maximum ratings of the laser diode light power, a study of the feedback loop stability is necessary. Two transconductance structures, using inverting and noninverting circuits are compared using the dominant pole compensation method. Then, the more stable driver circuit is analyzed and integrated using CMOS technology.
IEEE Transactions on Instrumentation and Measurement 03/2004; · 1.21 Impact Factor
-
IEEE T. Instrumentation and Measurement. 01/2004; 53:102-108.
-
[show abstract]
[hide abstract]
ABSTRACT: A laser range-finder has been designed for distances from 1 to 20 m with diffusing targets. The measurement principle using a frequency modulation continuous wave technique is first described. Improvement for short distance measurements is then proposed with a new set-up using a delay line and an avalanche photodiode working as an optoelectronic mixer used to obtain a wide-band and a low-noise photoreception.
IEEE Transactions on Instrumentation and Measurement 11/2002; · 1.21 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: A laser range-finder has been designed for distances from 1 m to
20 m with diffusing targets. The measurement principle using a frequency
modulation continuous wave technique is firstly described. Triangular
and sine frequency modulation types are both analyzed. Improvements for
short distance measurements and for low noise photoreception are then
proposed. The designed set-up is also theoretically compared to a
phase-shift laser range-finder in order to point up its main advantages
Instrumentation and Measurement Technology Conference, 2001. IMTC 2001. Proceedings of the 18th IEEE; 02/2001
-
[show abstract]
[hide abstract]
ABSTRACT: A laser sensor using optical feedback interferometry has been
designed to measure displacements up to several micrometers. For this
purpose, a lithium niobate crystal generates a phase shift in order to
modulate at a frequency of 580 kHz the optical path difference between
the sensor and the target. This setup has been calibrated by means of a
commercial optical fiber sensor. The overall accuracy resulting is
better than 65 nm for a 1-kHz sinusoidal motion of the target at 10 cm
away from the sensor
IEEE Journal of Selected Topics in Quantum Electronics 10/2000; · 3.78 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The authors review the recent progress in dynamic interferometry applied to optical distance measurement devices. These techniques are based on modulating laser diodes to obtain a continuous wavelength-shift. Measurements can be done by a fringe counting technique or by detecting the beat frequency that is proportional to the time-of-flight. This paper shows too the importance of the spectral analysis of the photodetected signal. A signal processing allows the use of multiple-beam interferometers with multiple reflectors. It's also possible to determine absolute distances with rough targets from the ratio of the amplitudes of two harmonics.
Journal of Optics 07/2000; 23(3):117.
-
[show abstract]
[hide abstract]
ABSTRACT: A phase-shift expander has been designed for distance measurements with a low-frequency phase-shift laser rangefinder using two coherent indirect frequency synthesizers. Using such a set-up, a theoretical analysis shows that the resolution of the phase-shift measurements can be improved. The main error source due to the output voltage, which is a frequency-modulated signal, is then analysed. Other error sources, such as the output frequency stability and the cross talk between the two channels, are evaluated. The intrinsic resolution is estimated as a function of the frequency drift of oscillators and an improved circuit is proposed. Experimental results are presented for a 16 and 32 expansion factor of the phase shift.
Journal of Optics 12/1998; 29(3):229.
-
[show abstract]
[hide abstract]
ABSTRACT: A self-mixing type scanning range finder has been designed for the
acquisition of three-dimensional (3-D) image data and object
construction. The optical beam backscattered by a rough target into the
laser diode cavity causes strong variations of the optical output power,
the pseudo period between two of these variations corresponding to a
distance of a half wavelength. With such a device, images have been
obtained with a resolution of 0.1% over ranges up to 3 m, for 3-D object
construction
IEEE Transactions on Instrumentation and Measurement 11/1998; · 1.21 Impact Factor
