Method of Chip Burn-In
Original Publication Date: 1995-Nov-01
Included in the Prior Art Database: 2005-Mar-31
Greschner, J: AUTHOR [+3]
Disclosed is a method of chip burn-in, whereby the contacting of the connexions of the chip is done by contact probes having contact wires functioning as buckling beams.
Method of Chip Burn-In
a method of chip burn-in, whereby the contacting
of the connexions of the chip is done by contact probes having
contact wires functioning as buckling beams.
of semiconductor chips is, on the one hand,
determined by the fulfillment of special functions like speed of
signal processing, completeness of the electrical networks, etc. On
the other hand quality is measured in view of reliability when in use
(live time). Live time is tested before delivery by simulation of
application conditions using a time-lapse method. This, however, is
very time-consuming and costly. Soldering a chip to a carrier,
applying different temperatures while introducing electrical voltage
and subsequent removal of the chip from the carrier will damage the
connections in such a way that they will have to be prepared again
for attachment to the ultimate carrier.
the contacting is carried out using contact probes
having buckling beams, whereby the guiding plates are made of
silicon. This is important since the burn-in is carried out at
temperatures between about -65 degree C and +300 degrees C, where the
silicon chip (wafer) will expand. Since the guiding plates are also
made of silicon, they will expand equally. Thus, a slipping of the
beams off the connections due to different coefficients of expansion
themselves are made of a composite material, whereby
the core is formed of Cu-Be, being the actual buckling beam. Within
this core there are arranged several fibres (about 7) made of Ag-Pd
30. These fibres increase the contact characteristics of the beam.
The beam is formed at its front side such that the Ag-Pd fibres
protrude only slightly from the Cu-Be core. The fibres (similar to
dendrites) will penetrate the oxide of the contact points of the chip
more easily than a surface of Cu-Be due to their hardness and shape
(about 5-10 mu.m diameter/fibre).
core of the contact beam there is a sheath of pure
copper, which serves to increase the cross-section thereby reducing
the ohmic resistance of the whole beam. The copper sheath is
separated from the Cu-Be core by means of a thin nickel layer, which