Electricity: measuring and testing – Measuring – testing – or sensing electricity – per se – With rotor
Reexamination Certificate
1999-01-14
2001-01-16
Metjahic, Safet (Department: 2858)
Electricity: measuring and testing
Measuring, testing, or sensing electricity, per se
With rotor
C324S073100, C714S734000
Reexamination Certificate
active
06175230
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention is directed to automatic circuit-board testing. It particularly concerns digital in-circuit testing.
There is usually a stage in printed-circuit-board manufacture at which the boards are subjected to so-called in-circuit tests. This type of test is performed before any overall functional testing, i.e., before any testing of the board's overall functional characteristics, with which in-circuit testing does not concern itself. In-circuit testing instead focuses on the performance of individual components after they have been placed on the circuit board.
In an in-circuit test, probes contact the board primarily at internal board nodes, typically circuit-board traces, rather than at board terminals. Digital drivers apply signals to some of these traces, and digital sensors determine whether proper signal levels result at these or other similarly probed traces. The ensemble of the digital levels to be applied and/or sensed at all of the test nodes at a given clock time is known as a vector, and the tester includes fast pin memories that together contain codes for a large number of such vectors. A sequencer causes a burst of those codes to be fetched and the resultant levels to be applied to the drivers and sensors at a rapid rate, and the sensors' results are simultaneously loaded back into the pin memories to record the results of the test. Each component's test requires one or more such bursts, so a board containing many components requires many bursts to test it.
Between bursts, the circuit board's components are allowed to cool down and the pin memories are reloaded with new vectors. Tester multiplexers known as “scanners” may also be operated during the inter-burst interval to change the connections between board nodes and digital instruments. (There are typically are many fewer digital instruments than probed nodes, because only a subset of the probed nodes needs to be driven and/or sensed during any individual burst.)
If a circuit board and its components are very simple, then the time required for an in-circuit test may be so small that the most-significant factor affecting tester throughput is the time required to load the board into the tester and remove it again. For most of the more-complex circuit boards, though, the time required for the actual electrical testing can be quite long, so test length has a significant effect on the testing volume of which a given number of board testers are capable.
Unfortunately, the very nature of in-circuit testing tends to make it a slow process. This is a result of the need to isolate individual components during an in-circuit test. Specifically, the board is powered up during an in-circuit test, so components other than the one currently being tested may be attempting to drive a board trace to a level different from the one that the test specifies. Various techniques are employed to minimize other component's tendencies to interfere in this manner; it is often possible to apply signals that will disable those neighbor components, for instance. But not all components can be disabled. Circuit testers typically deal with such components by backdriving, i.e., by driving the node with enough current to overpower the conflicting component.
Although it is possible in principle to make backdrive-capable digital drivers very fast, doing so is usually too expensive to be practical, so backdrive-capable digital drivers ordinarily can operate only at speeds considerably slower than, say, the speeds at which the boards being tested will operate in their intended environment. So it typically is not practical to increase test-vector-application speed significantly. Moreover, the high backdrive currents heat board components up during a burst, so cool-down time must be allowed between bursts.
SUMMARY OF THE INVENTION
We have recognized that a significant reduction in overall test time can be afforded if the digital drivers are provided with current sensors whose outputs are functions of the load currents that those drivers source or sink. The current sensors' outputs indicate whether any backdriving actually occurred during the previous burst. In accordance with our invention, the minimum time intervals between bursts depend on those current sensors' outputs. If the sensors' outputs during a given burst indicate that no backdriving occurred, for instance, the inter-burst interval may be permitted to be as short as the time required by other necessary inter-burst operations, such as pin-memory reloading or scanner reconfiguration. But a minimum cool-down delay will be imposed if the backdriving observed during the previous burst reached a certain level.
Without current-sensor use, a backdrive-dependent cool-down period would not be advisable, because backdriving cannot be predicted reliably during all parts of the test. Differences in component supplier and production lot can cause different boards of the same design to exhibit different backdrive behavior, even if those boards are all good. And a bad board is particularly likely to exhibit unpredictable backdrive behavior. By using current sensors, though, the tester can reliably match the cool-down interval to the actual need for it.
REFERENCES:
patent: 4520416 (1985-05-01), Karash
patent: 4554630 (1985-11-01), Sargent et al.
patent: 4569048 (1986-02-01), Sargent
patent: 4588945 (1986-05-01), Groves et al.
patent: 4740895 (1988-04-01), Sargent et al.
patent: 4827208 (1989-05-01), Oliver et al.
patent: 4856001 (1989-08-01), McElfresh et al.
patent: 4862069 (1989-08-01), Albee
patent: 4931742 (1990-06-01), Karash et al.
patent: 4973904 (1990-11-01), Sonnek
patent: 4994732 (1991-02-01), Jeffrey et al.
patent: 5101150 (1992-03-01), Sullivan et al.
patent: 5101153 (1992-03-01), Morong, III
patent: 5124636 (1992-06-01), Pincus et al.
patent: 5127009 (1992-06-01), Swanson
patent: 5144229 (1992-09-01), Johnsrud et al.
patent: 5146159 (1992-09-01), Lau et al.
patent: 5235273 (1993-08-01), Akar et al.
patent: 5477139 (1995-12-01), West et al.
patent: 5519331 (1996-05-01), Cowart et al.
patent: 5550844 (1996-08-01), Lucas
patent: 5602855 (1997-02-01), Whetsel, Jr.
Ministry of Defence—“Safe Operating Limits for Backdriving”—Interim Defence Standard Improvement Proposal, Defence Standard No. 00-53/1, Nov. 15, 1991.
D.S. Morris “The MOD Investigation into Backdriving or Nodeforcing,” UK Ministry of Defence (Navy) London, 1987, (no month available).
Hill et al., “Safe Operating Limits for Backdriving in ATE,” ERA Technology Ltd., Cleeve Road, Leatherhead, Surrey, KT22 7SA UK, IEEE 1987, (no month available).
Safeguard In-Circuit—“A description of HP's safe implementation of Digital In-circuit test,” Hewlett Packard, Product No. 3065-2, Sep., 1985.
Eskici Jak
Hamblin Michael W.
Suto Anthony J.
Cesari and McKenna LLP
GenRad Inc.
Metjahic Safet
Tang Minh
LandOfFree
Circuit-board tester with backdrive-based burst timing does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Circuit-board tester with backdrive-based burst timing, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Circuit-board tester with backdrive-based burst timing will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2528552