Error detection/correction and fault detection/recovery – Pulse or data error handling – Digital logic testing
Reexamination Certificate
2001-05-04
2004-02-24
Decady, Albert (Department: 2133)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Digital logic testing
Reexamination Certificate
active
06697982
ABSTRACT:
COPYRIGHT STATEMENT
A portion of the disclosure of this patent document contains material which is subject to copyright or mask work protection. The copyright or mask work owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright or mask work rights whatsoever
BACKGROUND OF INVENTION
1. Field of Invention
This invention relates to improvements in testing integrated circuits, in general, and to improvements in methods for testing embedded core integrated circuits while preserving the intellectual property contained in the core circuits, that is, protecting the core circuits from being discovered or determined by inspection of the testing software, in which virtual or pseudo pins are used to represent at least a portion of the core circuits to the testing software.
2. Relevant Background
Recently, embedded integrated circuit cores have been increasingly popular. An embedded circuit core is an integrated circuit building block or module that is embedded into a system chip. Examples of such embedded cores include customizable digital signal processors (cDSPs), which may be embedded into a system that is integrated onto a single integrated circuit chip. The DSP may be of a proprietary design, and the peripheral and supporting circuitry may be customized for a particular use or customer.
Presently, some companies sell “embedded core” or circuit designs for use by customers who wish to design their own custom circuit content to produce their own customized or application specific integrated circuits (ASICs). As an example, a customer may wish to design its own peripheral circuits for use with a particular DSP core. In practice, the designer of the DSP provides a circuit description of the DSP in a format that will enable the customer to fabricate his desired circuit. The DSP description, however, will ideally be only sufficient to enable the customer to fabricate the custom DSP, together with peripheral circuits of its own design onto an integrated circuit, but does not disclose the specific design and circuitry of the embedded DSP. A chip in which the embedded core is included is sometimes referred to as a system on chip (SOC).
However, using this chip manufacturing technique gives rise to several problems. For example, from a business standpoint, to make the design of a core circuit economically feasible, the design often needs to be reusable in many diverse customer circuit designs. However, the delivery of a complete circuit design to a customer often enables the customer to reverse engineer the circuit, enabling the circuit configuration, and other aspects of the circuit, to be discovered. The customer then has the ability to reproduce the circuit at will, thereby shortening the lifetime during which the core designer can recoup its design and engineering costs. Therefore, unless steps are taken to protect against reverse engineering the core, the value of the intellectual property in the core may be compromised. Such embedded cores in which the owner has one or more intellectual property interests have recently been referred to as “IP cores”.
Therefore, when an embedded IP core is delivered to a customer, details of construction of the core are generally not disclosed. The information that is disclosed often is just sufficient to enable the user to interface its own circuitry to the IP core. Thus, previous IP core models often were treated as “black boxes”, in order to protect the IP content embodied in the models. No visibility into the internal structure or circuitry of the models was provided. Thus, for example, the core often may be delivered only as a circuit description, with only the connection pins and required signal parameters thereof being disclosed. However, this makes testing of the embedded core difficult or impossible. However, a properly designed core should be completely tested at least in a netlist version before being delivered to a customer for use with its custom designed circuitry. Thus, ideally, it should be unnecessary for the customer to repeat the core tests, except to the extent necessary to properly test the added circuitry and their interface and operation with the core.
Nevertheless, after an ASIC has been designed and manufactured, the chip should be tested for manufacturing defects. This typically involves first simulating a properly functioning ASIC, developing a variety of input test patterns and applying them to the simulated ASIC, and determining the expected outputs of the ASIC. The input test patterns are then applied to the actual ASIC. The actual outputs are then compared to the expected outputs. Deviations from the expected outputs indicate that the ASIC has a manufacturing defect.
Because of the complexity of the logic buried within the ASIC, an enormous number of test patterns may be required to properly test the integrated circuit device. One technique that has been used to test such integrated circuits is to add “scan cells” to the ASIC at strategic locations throughout the circuit. Scan cells are points at which logic values may be forced into the ASIC (scan writable gates) and/or observed within the ASIC (scan readable gates). Scan cells normally take the form of flip-flops which may be forced to desired logic values from outside the ASIC, or which may be read from outside the ASIC to determine particular behavior of the ASIC, for example after the test inputs have been applied. For instance, some scan cells enable an external tester to preset a counter within the ASIC to a particular count, and to capture values that verify that combinatorial logic derived from the counter outputs is working properly.
In general, in the design of such scan flip-flops, non-scan flip-flops are often converted into scan flip-flops by adding some additional logic gates, often including a multiplexer. The multiplexer selectively allows either the test values or the circuit values to be applied to the scan cell with which it is associated. The multiplexer is operated for example by a selection signal applied to a selected pin of the ASIC.
Strategic placement of scan cells within an ASIC allows the number of test patterns required to fully test an ASIC to be drastically reduced. Scan cells are normally connected in long chains, allowing all of the scan cells to be written or read from using only a few dedicated test I/O pins on the ASIC. This is important primarily because the computer simulations necessary to produce the test patterns and expected resulting outputs require a great deal of computer resources. Without scan cell techniques, simulation times to produce test patterns that adequately test an ASIC would be impractically long. Furthermore, without scan cells some gates within an ASIC may be simply not testable. However, using scan cells it is often possible to develop test patterns that will detect 99% or more of possible gate failures using a set of test patterns that is not impractically long. The fault coverage obtained is the percentage of possible gate failures that will be detected by a given set of test patterns.
Nevertheless, ASIC designs using embedded cores present a challenge to test. To assist the custom designers, embedded cores typically are supplied by the vendor along with a set of test patterns, which, if applied in isolation, will produce 99% fault coverage. However, these patterns cannot usually be applied to the embedded core in isolation, because the core is buried inside the ASIC with no direct access to the core”s primary inputs and primary outputs. Hence, it is not possible to apply the supplied set of test patterns to the core in isolation unless some mechanism is supplied for accessing the core.
Among the isolation schemes that have been proposed are the use of multiplexed isolation, boundary scan cells wrapped around the core, and core parallel module testing, which includes the use of multiplexers to allow direct access to the circuitry in question. All o
Butler Kenneth M.
Chakravarthy Srinivasa
Hernandez Julio C.
Parekhji Rubin A.
Bassuk Lawrence J.
Brady W. James
Chase Shelly A
De'cady Albert
Telecky , Jr. Frederick J.
LandOfFree
Generating netlist test vectors by stripping references to a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Generating netlist test vectors by stripping references to a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Generating netlist test vectors by stripping references to a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3336920