Apparatus and Method for Detecting Floating Nodes

In the manufacture of very large scale integrated circuit chips, the manufacturing steps include the initial step of creating a detailed circuit diagram of the system which is to be manufactured on the integrated circuit chip. After the circuit diagram has been designed, different photographic masks are prepared for the implementation of the circuit diagram into physical form on the chip. Different steps in the manufacturing process use these masks for effecting the doping of the silicon crystal forming the chip. The die size of a typical chip, including many thousands of components, is one-half inch by one half inch. Because of the large number of components, it is possible for errors to be made, either in the original circuit design or in the various masks which are employed to incorporate that circuit onto the chip through the manufacturing process.

To ensure that the manufactured chip performs in accordance with the original intended design, test operations of the chip typically are undertaken prior to the sale of the chip to customer. Even so, an occasional problem exists, in which chips tend to draw excess current when they are installed into an end-use system. Excessive current draw can be a significant problem for chips which are used in portable computers and the like, since the excessive current draw depletes the useful battery life of such portable devices sooner than intended. More generally, excessive current draw is often viewed as a potential reliability problem. One type of flaw which causes excessive current draw is the existence of one or more "floating nodes" on a chip when these nodes are created by design flaws in the original circuit design and/or in the manufacturing processes as a result of a flaw in one or more of the masks used during the manufacturing process. During operating tests of such chips, such "floating nodes" often are undetected, since the chips generally perform as designed, with the exception of the excessive current draw.

It has been discovered that "floating node" design flaws on a chip are often light sensitive (that is, the excessive current draw disappears in the presence or absence of light). Since integrated circuit chips typically are potted or shielded from light in their final packages, the floating node excessive current draw takes place. Even should the current draw not be inversely sensitive to light (that is, the excessive current draw disappears in the absence of light), the unstable nature of a floating node makes the future behavior of the activation unpredictable.

Accordingly, it is desirable to provide a method and apparatus for locating floating node circuit design flaws in an effective manner to pinpoint the specific areas of the chip having such floating nodes.

It is an object of this invention to provide an improved method and apparatus for locating circuit design flaws on an integrated circuit chip.

It is an additional object of this invention to provide an improved method for locating light sensitive design flaws on an integrated circuit chip.

It is another object of this invention to provide an apparatus for use in locating light sensitive design flaws on an integrated circuit chip.

It is a further object of this invention to provide a method and apparatus for pinpointing the area of floating node design flaws on an integrated circuit chip.

In accordance with a preferred embodiment of this invention, an apparatus and method has been developed for locating light sensitive circuit design flaws on an integrated circuit chip. The apparatus comprises a device for directing a spot of light, having an area which is less than the surface area of an integrated circuit chip, along a predetermined path scanning the surface of the chip. Simultaneously, the chip is stimulated; and a measurement of current supplied to the chip is made. Observation of the location of the spot of light on the surface of the chip is made simultaneously with observation of the current drawn by the chip, to detect changes in the current drain as the spot of light impinges on the chip.

In employing the method, once an initial scan of the chip locates one or more areas of the chip where the current drain changes, the spot of light is reduced in size; and the region of the chip causing the initial current drain change is scanned along a pre-established path on the surface of the chip, once again observing the location where current drain changes take place. If a further narrowing of the area on which the current drain changes take place is desired, the process is repeated with a still smaller spot of light in order to narrow down the area of chip where the circuit design flaw exists. This is done to permit subsequent examination of the actual circuit diagram for that area, along with an investigation of the various masks used to construct the circuit on the chip, to determine where the design flaws exist.

FIG. 1 is a diagrammatic representation (not to scale) of an apparatus used in accordance with a preferred embodiment of the invention;

FIG. 2 is a diagrammatic representation of the method of use of the apparatus of FIG. 1; and

FIGS. 3A through 3E are diagrams useful in explaining the method of operation of the apparatus of FIG. 1 in accordance with the representation of FIG. 2.

Reference now should be made to the drawing, in which the same reference numbers are used throughout the different figures to designate the same or similar components. FIG. 1 shows a stereoscopic microscope 17, which has been modified in order to practice the method of a preferred embodiment of the invention. The microscope 17 is a conventional microscope, having two oculars 18 and 19. The ocular 18 is fitted with a conventional eyepiece 20 for viewing by the eye of a user of the microscope to observe an integrated circuit chip 10 placed on, and electrically interconnected with, a test system load board 11. The load board 11, in turn, is electrically interconnected with an integrated circuit chip ester 14. The load board 11 and the integrated circuit chip tester 14 are of conventional types utilized in the integrated industry for supplying power to integrated circuit chips 10 and for performing various tests on the input/output I/O) leads of a chip 10 placed on the load board 11.

Typically, the integrated circuit chip 10 has a die size of approximately one-half inch by one-half inch, and includes many thousands of electrical components. In the operation of the device of FIG. 1, the chip 10 typically is a silicon chip; and the test forming the subject matter of the preferred embodiment of this invention is performed on a chip which has been packaged such that the entire die surface area is exposed. Contacts are made between the chip 10 and corresponding contacts on the load board 11 in a conventional manner, leaving the top surface of the chip 10 exposed.

A chip tester 14 is used for measuring the overall current draw of the chip 10. A test start is made after each movement of a scanned light spot. The light does not move until the test is complete and measurement taken. This technique is used to provide a continuous indication of the current being drawn or consumed by the chip 10. This constitutes the operating current for the chip 10. The device of FIG. 1 is used for the purpose of detecting floating node errors in the design of the chip 10 when consistent excess current draw of chips 10 of a particular design have been encountered after the chips 10 have been incorporated into integrated circuit packages for subsequent use in various systems.

It has been discovered that floating node design flaws on a chip 10 are often light sensitive, that is, the excessive current draw disappears in the presence or absence of light. Consequently, the device of FIG. 1 has the second ocular 19 of the stereoscopic microscope 17 fitted with a termination plug 22 for a fiber-optic light tube 26 used to supply light from a source 24 through the ocular 19 of the microscope 17 and through the internal optics of the microscope by way of a light path 30, illustrated in FIG. 1, to the upper or exposed surface of the chip 10. For situations where the design flaw of a chip is not light sensitive, a conventional liquid crystal analysis may be employed. For light sensitive design flaws, however, liquid crystal analysis is precluded. Even if liquid crystal is an option, the technique relies heavily on an extremely controlled ambient temperature, which is difficult to achieve in a test floor environment.

In using the device of FIG. 1, the light source 24 is turned on; and the ambient or surrounding light is removed (that is, the device is used in a darkened room or in a darkened box). The spot of light 36 from the light source 24 then is focused to a first size or diameter, which has an area which is substantially less than the area of the surface of the chip 10. The light beam or spot of light 36 from the beam 30 is scanned over the surface of the chip in a winding path 40, as illustrated in FIG. 2; so that the spot of light 36 impinges on all areas of the chip throughout the scan. The scan may be effected automatically by causing a relative movement between the spot 36 and the surface of the chip 10; or it may be effected manually under control of the knobs 32 and 33 on the microscope, which are used to effect relative X-Y adjustments of stereoscopic microscopes With respect to a specimen (here, the chip 10) placed at the focal point of the microscope.

During the scan of the spot of light 36 along the path 40 shown in FIG. 2, test starts are sent to the chip tester 14 and current measurements are made. The measurements are observed to determine where, if any, changes in the current drawn by the chip 10 occur. Once these changes in the current draw indicated by the tester 14 (at the indicator 15) take place, visual observation of the location on the surface of the chip 10 is effected through the eyepiece 20 to locate the specific quadrants in which the current changes take place.

FIG. 3A is representative of the first or gross scan of the chip 10 to determine which quadrants, A, B, C or D of the chip 10, include a light sensitive floating node error. For illustrative purposes, assume an error that is typified by an inverse relationship to light, that is the excessive current draw drops in the presence of light; so that for regions of the chip which do not include a floating node design error, the current draw remains high even though the spot of light 36 is impinging on the upper surface of the chip 10. When a quadrant or region of the chip 10, however, which has a floating node error in it, is illuminated by the spot of light 36, the current draw for the entire chip drops as a result of the light impinging on the chip. In FIG. 3A, this is indicated by the circled illustrations showing high current draw for quadrants A, C and D and a low current draw for the quadrant B of the chip 10, indicative of a floating node error somewhere in quadrant B of the chip 10.

For the purposes of illustration in FIGS. 3A through 3D, the current levels are indicated as "spikes". It should be noted, however, that the current levels encountered, particularly for quadrants A, C and D, are a steady state, relatively high or undesirable current level, and that when the light spot 36 impinges upon the floating node error portion of the chip, the current level drops to a lower steady state current so long as the light spot is on the defective region. For purposes of illustration, however, the effect is considered to be better illustrated by showing "spikes" of current in the circled portions, since it is a relative current difference which is detected on the tester 14 for isolating the different quadrants of the chip 10 which may contain a floating node error in them.

Once the scan of FIG. 3A has taken place, and it is determined that only the quadrant B of the chip 10 has a potential floating node error in it, the procedure is repeated for the quadrant B only. When this is done, the size of the spot 36 is reduced by affecting the zoom of the microscope to create a smaller spot of light from the light source 24. The process is repeated for the four subquadrants B1 through B4 of quadrant B, shown in FIG. 3A. This subdivision of quadrant B, with the same scan pattern taking place as shown in FIG. 2, is illustrated in FIG. 3B. As shown in FIG. 3B, it is determined that subquadrants B1, B3 and B4 are correct or normal, but that a floating node error occurs in subquadrant B2. Once again, further isolation of subquadrant B2 is effected; and a scanning of that subquadrant is effected by employing the method described above. The spot of light 36 is focused to an even smaller spot of light; and the process is repeated. As illustrated in FIG. 3C, a determination is made that the floating node error occurs in the sub-subquadrant B2-1 of the four sub-subquadrants B2-1 through B2-4.

If possible, given the die area, the process may be repeated one more time, as illustrated in FIG. 3D, breaking the sub-subquadrant B2-1 into four sub-sub-subquadrants B2-1A through B2-1D. As illustrated in FIG. 3D, the floating node error is isolated into the sub-sub-subquadrant B2-1A, which significantly limits the area of the original chip circuit and mask set which must be re-examined in order to determine the location of the design flaw creating the floating node on the chip 10.

FIG. 3E illustrates the final breakdown, which is effected by means of the steps shown through FIGS. 3A through 3D, pinpointing the location of the floating node in the sub-subsubquadrant B2-1A. It is apparent from an examination of FIG. 3E that the percentage of the area of the chip 10 which must be examined in the circuit diagrams and mask sets to determine the cause of the floating node error is pinpointed to a relatively small portion of the total surface area of the chip 10. This greatly expedites the redesign steps necessary to locate and correct the design error. If the entire circuit design and mask set areas for the chip 10 needed to be examined to locate the design area of sub-sub-subquadrant B2-1A, a considerable additional expense of time and effort beyond that required for just the examination of the small area B2-1A would occur. As a consequence, the method which is employed in the preferred embodiment of this invention significantly reduces the time and effort required to locate and then correct floating node system design errors on a chip.

The foregoing description of the preferred embodiment of the invention is to be considered as illustrative and not as limiting. The technique which is used to simultaneously observe the location of the spot of light and to scan the spot of light over the surface of the chip 10 may be varied to perform substantially the same function, in substantially the same way, to achieve substantially the same result, without departing from the true scope of the invention as defined in the appended claims.