Measuring and testing – Specimen stress or strain – or testing by stress or strain... – Specimen clamp – holder – or support
Reexamination Certificate
1998-12-09
2001-05-29
Noori, Max (Department: 2855)
Measuring and testing
Specimen stress or strain, or testing by stress or strain...
Specimen clamp, holder, or support
C073S827000
Reexamination Certificate
active
06237422
ABSTRACT:
This invention concerns a device and method for testing the integrity of a bond between an electrical device and an electrical conductor thereof.
Electrical semiconductor devices are very small. Typically a silicon wafer or chip comprising an electrical circuit is encapsulated in or affixed to a relatively rigid substrate; the substrate is necessary in order to give the wafer sufficient mechanical strength. Several silicon devices may be attached to a single substrate.
Individual semiconductor device comprising a number of silicon chips are usually fixed with other electrical components to a printed circuit board in order to make a complete control device, for example for a personal computer.
A typical prior art semiconductor device has peripheral connection sites whereby edge contacts can be connected to a circuit board or to an intermediate circuit assembly. Usually these connection sites are densely packed in a linear array around the semiconductor device. Electrical connections are by way of individual wires bonded to the semiconductor connection site at one end, and to the circuit board at the other end.
Reductions in size of the semiconductor device, due to advances in technology, have resulted in insufficient edge length for the necessary number of connection sites. Accordingly the connection sites may be arranged in a rectangular array over the face of the device. This requires a new method of making electrical connections, and it has been proposed to provide raised bumps of electrically conductive material at each connection site. In order to make connection to another component, such as a circuit board, the bumps are placed in register over a corresponding array of connection sites, and the bumps are bonded in order to form a permanent electrical connection. In the case of bumps of solder, the solder is reflowed in order to make the connection. In the case of bumps of gold, pressure may be sufficient to form a good bond. Other methods may also be suitable, and the bumps may be formed of other materials.
Such bumps are typically in the size range 0.05-1.0 mm in diameter, and a single control device may incorporate many hundreds of bumps of a similar shape and size.
Test methods are required in order to have confidence that the bumps are adequately bonded to the connection sites prior to connection to the corresponding connection sites; for example by solder reflow. Typically a shear test is used whereby a test head pushes the bump sideways to measure the breakaway force. Continual testing is required in order to identify statistical trends which indicate deviation from the desired quality standard.
A problem arises in the case of multi-layer substrates having a plurality of electrical printed circuits separated by intermediate insulating layers. Such substrate are also a result of miniaturization whereby sufficient electrical tracks cannot be provided in a single layer. Typically the outer layer of substrate will have apertures to permit electrical connection to the internal electrical tracks, and the bumps are thus partially within the printed circuit substrate instead of being wholly on the surface thereof.
The same problem can arise where an electrical track on the surface of the substrate has a solder bump bond site defined by a solder resist mask, which may for example be screen printed on the substrate. The mask confines the solder as it is applied to the substrate, and has the effect of raising the surface of the substrate with respect to the bond site.
Clearly a shear test is inappropriate for these kinds of partially buried bump since the wall of the aperture in the outer layer of substrate will support the bump and thus resist shear forces. Accordingly a tensile test is required.
It has been proposed to apply a test head to a solder bump, heat the head to reflow the solder whereby the bump and test head are bonded, and apply a tensile force after the bond has cooled. This has the disadvantage of being slow, requiring a new test head for each test, and of potentially affecting the bond site itself as the solder is reflowed to attach the test head. This later disadvantage is an especial problem since the arrangement of metal layers at the base of the bump may be complex, and the application of heat may affect the metallurgy thereof.
Other proposals have included gripping the bump between jaws. In the case of plain jaws the difficulty is to develop sufficient friction whilst avoiding crushing loads. Since the bump is often approximately hemispherical, the jaws typically make a point contact, and thus the local stress on the bump can be very high; this solution has proved impractical.
Another solution provides pointed jaws to indent the bump so as to improve traction, but in practice the bump is likely to shear at the plane of the jaws due to inward migration of the opposed indents. Furthermore, both of the mechanical gripping proposals may change the shape of the bump sufficiently to reduce confidence that the electrical bond breaks only as a result of the tensile test.
According to the present invention there is provided a test head of a device for tensile testing electrically conductive projections of an electrical device, the test head having jaws closable on a projection thereby to grip the projection, and adapted to apply a tensile load in the direction of projection thereof, wherein the jaws are adapted to closely engage the projection over a substantial part of the surface thereof, the distal edge portion of each jaw approaching the axis of the jaws.
Typical projections are of solder, gold or a suitable electrically conductive composite material.
Preferably the jaws are concave and substantially part spherical when intended for engagement with a projection such as an arcuate bump of solder. Alternatively the jaws may define a somewhat planar contact surface for bumps which are substantially rectangular.
In a preferred embodiment the jaws confine the bump sufficiently closely to obviate spreading or crushing of the bump under clamping forces. Thus the opposed jaws both clamp and support the bump such that the compressive clamping loads are resisted. The jaws may nevertheless not wholly confine the bump so long as the unconfined portions are not subjected to significant spreading or crushing loads.
In a preferred embodiment the jaws are concave and adapted to engage a bump over greater than 40% of the jaw surface area thereof, preferably over 60%, and most preferably over 75%. In the case of bumps of regular shape, at least 85% and preferably greater than 90% of the jaw surface area is adapted to engage the bump. The jaws may define a relief volume less than 10% of the surface of the bump intended to be gripped by the jaws, and most preferably less than 5% in the case of bumps of regular shape.
In the preferred embodiment the inner edges of the jaws, in the closed condition, are substantially further apart than the width of the electrical contact area of the bump.
Preferably two, three or at most four jaws are provided in order to minimize manufacturing cost.
In a preferred embodiment the jaws may engage the bump sufficiently closely to cause re-shaping forces whereby the bump can more closely conform to the shape of the jaws. The material of the bump, e.g. solder, is relatively soft and a re-shaping of the surface portion is useful in increasing the contact area, and thus reducing areas of high stress. Necessarily the re-shaping must not result in crushing or spreading loads, but can be determined according to the parameters of the bump to be tested. Relevant factors are size, shape and material specification of the bump.
The jaws may include relieved areas to permit re-shaping in a confined and deliberate manner. In this way the close fit of the jaws can support the bump whilst ensuring that excess or mis-shapen material is accommodated within the jaws in a controlled way.
In a preferred embodiment the jaws define a substantially cylindrical bump engaging cavity having sidewalls extending in the direction of the tensile load to be applied. In this embod
Dage Precision Industries Ltd.
Kilpatrick & Stockton LLP
Noori Max
LandOfFree
Apparatus and method for testing strength of electrical bond... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for testing strength of electrical bond..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for testing strength of electrical bond... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2552923