|
|
|
A 16 x 16 cells CNN Chip for CCD with Optical Input
Team: Servando Espejo, Rafael Domínguez-Castro
and Ángel Rodríguez-Vázquez.
Date: 1993
| Physical Data |
- Technology: 1.6µ m CMOS n-well,
single-poly, double metal
- Cell density: 85 cells/mm2
- 47 MOS plus 2 vertical bipolar transistors/cell
- Cell dimensions: 118.0µ m
X
95.6µ m
- Die size: 2479.2µ m
X
2500.0µ m
|
Electrical Data |
- 2.7 to 5.0 V Supply
- Optical input, voltage mode output
- 105µ W/cell @ 5V
- Total dissipation: 42mW @ 5V
|
Design Technique |
- Analog Current-Mode Continuous Time
- Darlington stage as optical sensor
- Conventional CMOS logic
- Full-custom
|
Features and Applications |
- Optical input 16 x 16 CNN cells for CCD
- Automatic adaptation to environmental
illumination.
- Boundary cells set to -1 (white)
- Independent 16-lines-wide output bus
- Processing time: 4µ s / image
|
This 16 x 16 cells prototype is a variation
of a similar one (See "A 16 x
16 cells Cellular Neural Network Chip for Connected Component Detection.")
incorporating a novel optical interface [Espe94a] with automatic adaptation
to environmental illumination conditions. Input images are projected over
the surface of the chip. A self-adjusting circuitry centers the photogenerated
current levels at each pixel, resulting in automatic adaptation to the
average illumination. The cell circuitry [Rodr93a],
[Rodr93b], [Espe94b],
[Espe96], is basically identical to that of
the previous one (See "A 16 x
16 cells Cellular Neural Network Chip for Connected Component Detection."),
except for the optical sensor and adaptation circuitry. Dedicated output
stages are included to interface the analog current-mode circuitry with
the external voltage-mode world.
The design was planned as a feasibility demonstrator
of a novel optical interface with automatic signal-range centering, and
its use as front end interface to Cellular Neural Networks realizations.
Results confirmed the viability of the idea.
References:
- [Espe94a] S. Espejo,
R. Domínguez-Castro, R. Carmona, and A. Rodríguez-Vázquez:
"CMOS Optical-Sensor Array with High Output Current Levels and
Automatic Signal-Range Centering". Electronics Letters, Vol. 30,
No. 22, pp. 1847-1849. 27th October 1994.
- [Rodr93a] A. Rodríguez-Vázquez,
S. Espejo, R. Domínguez-Castro, J.L. Huertas, and E. Sánchez-Sinencio:
"Current-Mode Techniques for the Implementation of Continuous-
and Discrete-Time Cellular Neural Networks" IEEE Transactions on
Circuits and Systems II: Analog and Digital Signal Processing.Vol 40,
No. 3, pp 132-146, March 1993.
- [Rodr93b] A. Rodríguez-Vázquez,
S. Espejo, J.L. Huertas, and R. Domínguez-Castro: "Switched-Current
Cellular Neural Networks for Image Processing", Chapter 17, pp.
459-486 in C. Toumazou, J.B. Hughes and N.C. Battersby (Editors): "SWITCHED-CURRENTS
an analogue technique for digital technology", IEE Circuits and
Systems Series 5, Peter Peregrinus Ltd. 1993. ISBN: 0 86341 294 7.
- [Espe94b] S. Espejo, A. Rodríguez-Vázquez,
R. Domínguez-Castro, J.L. Huertas, and E. Sánchez-Sinencio:
"Smart-Pixel Cellular Neural Networks in Analog Current-Mode CMOS
Technology" IEEE Journal of Solid-State Circuits.Vol 29, No. 8,
pp 895-905, August 1994.
- [Espe96] S. Espejo,
R. Carmona, R. Domínguez-Castro and A. Rodríguez-Vázquez:
"A VLSI-Oriented Continuous-Time CNN Model". International
Journal of Circuit Theory and Applications. Vol 24, No. 3, pp 341-356,
May-June 1996.
Images List:
|
|
