Browse Prior Art Database

ENCODING SCHEME TO REDUCE CAPACITIVE COUPLING EFFECT BETWEEN OVERLAPPING DIGITAL AND ANALOG METAL LAYERS IN A MIXED-SIGNAL INTEGRATED CIRCUIT

IP.com Disclosure Number: IPCOM000028328D
Original Publication Date: 2004-May-10
Included in the Prior Art Database: 2004-May-10
Document File: 5 page(s) / 809K

Publishing Venue

Motorola

Related People

Barry Herold: AUTHOR [+3]

Abstract

In this paper we present an encoding scheme, which we call the average encoding scheme to compensate for the detrimental effects of capacitive coupling between digital metal layers that overlap analog metal layers in a mixed signal integrated circuit. The scheme is not used for the purpose of reducing power consumption and is not limited for use in integrated circuits but can also be used on circuit boards. It does not need any additional bit lines and hence saves on expensive silicon real estate. In certain cases, as in our example implementation, namely, the digital pixel sensor, the encoding scheme was used without having to do a chip re-spin. The encoding scheme has the characteristic of averaging and constraining the rate of change of charge accumulation due to coupling capacitance hence avoiding a charge build up on the analog metal lines.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 27% of the total text.

ENCODING SCHEME TO REDUCE CAPACITIVE COUPLING EFFECT BETWEEN OVERLAPPING DIGITAL AND ANALOG METAL LAYERS IN A MIXED-SIGNAL INTEGRATED CIRCUIT

Barry Herold, Russell Mareachen, Mohamed Imtiaz Ahmed


ABSTRACT

In this paper we present an encoding scheme, which we call the average encoding scheme to compensate for the detrimental effects of capacitive coupling between digital metal layers that overlap analog metal layers in a mixed signal integrated circuit. The scheme is not used for the purpose of reducing power consumption and is not limited for use in integrated circuits but can also be used on circuit boards. It does not need any additional bit lines and hence saves on expensive silicon real estate. In certain cases, as in our example implementation, namely, the digital pixel sensor, the encoding scheme was used without having to do a chip re-spin. The encoding scheme has the characteristic of averaging and constraining the rate of change of charge accumulation due to coupling capacitance hence avoiding a charge build up on the analog metal lines.

1. INTRODUCTION

Power dissipation due to capacitive loading has been studied extensively and many encoding techniques to counter this power dissipation problem have been proposed [1-11]. However in many cases, data corruption due to capacitive coupling outweighs the need to save power. Such is the case in the digital pixel sensor (DPS). At this point, however, it is important to note that similar problems or scenarios like that of the DPS might exist and hence can make use of the proposed average encoding technique. The problem and the proposed solution are explained next.

2. CROSS-COUPLING CAPACITANCE AND ITS EFFECT

The digital pixel sensor contains analog voltage reference metal 2 (M2) lines that come into the chip from the top left corner of the chip and run horizontally across the pixel array. The digital data metal 3 (M3) lines enter from the bottom of the chip and run vertically across the array, overlapping the analog voltage reference M2 lines as shown in the simplified diagram of the pixel array in Figure 1.

The close proximity of these two metal layers, add to the capacitive coupling on the voltage reference lines. In the case of the DPS, ten bit data lines traverse vertically over each pixel in the array of pixels. The cross coupling between the metal layers can be modeled as shown in Figure 2. Each of the cross coupling capacitors on the data lines get charged or discharged with switching in the logic levels on the data lines. Since these data lines pass over each pixel, the total number of cross coupling capacitors quickly adds up with the size of the array. This has the effect of increasing the load on the voltage reference lines especially to the right of the pixel array and hence causing a shift in the voltage levels (noise) as well as increased settling times depending on the bit values being presented to the data lines. It is this dependency on the bit values that is being exploited...