Laser-trimmable digital resistor

Electric heating – Metal heating – By arc

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

C219S121680

Reexamination Certificate

active

06664500

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to a laser system for trimming or severing thin-film resistors fabricated on an undoped gallium arsenide substrate. The present invention also relates to a laser-trimmable resistor network suitable for use in bias circuits for power amplifiers.
BACKGROUND OF THE INVENTION
Many integrated circuits, such as bias circuits for power amplifiers, include individual resistors that are required to have a specific value to achieve a desired circuit performance level. According to U.S. Pat. No. 4,782,320 issued to Shier, in applications of these types, the accuracies of individual resistors prior to trimming are typically on the order of only 15-20%, because there are wide manufacturing variations in the sheet resistance of the integrated circuit. These single resistors must then be trimmed using off-chip resistors to accurately reach a predetermined resistance value.
Other circuits do not require that an absolute value of a resistor be obtained, but rather require that two resistors be accurately matched in value, one relative to the other. Creating a closely-matched pair of resistors is easier than creating an individual resistor with a certain resistance value, because process variations in the former affect both matched resistors equally. Thus, the level of matching of integrated circuit resistors that is achievable by controlling parameters of the manufacturing process is approximately 0.1-0.3%, as described by Shier. But for some circuits, such as analog-to-digital converters, even this degree of precision is inadequate.
In order to achieve a higher level of precision than that achievable by fabrication processes, it is known in the art to use a laser to trim a thin-film resistor fabricated on a silicon substrate. The laser alters the shape of a resistor and thereby brings its resistance to a desired value. Alternatively, the resistor may be severed altogether, if used as part of a resistor trimming network.
A conventional laser-trimming system is shown in FIG.
1
. An integrated circuit
10
includes a trimmable resistor
20
and other circuit elements (not shown) that are fabricated on a silicon substrate. Resistor
20
is typically made from a resistive thin-film material, such as nickel chromide, tantalum nitride, cesium silicide, disilicide, and polycide.
Integrated circuit
10
is coupled to an automatic test system
50
that measures the electrical properties of integrated circuit
10
in its untrimmed state. In response to the measured electrical properties, automatic test system
50
computes a desired trimming resistance. A predetermined trim pattern is obtained from memory
60
and provided to the laser drive and control
40
. In response, laser drive and control
40
positions laser
30
at desired positions over integrated circuit
10
and actuates the laser to produce a radiation beam that is focused on a predetermined area of the trimmable resistor
20
.
The wavelength of this radiation beam is typically selected based upon the light absorption characteristics of the silicon substrate and the resistive material. At certain wavelengths (for example, 1.32 &mgr;m), the silicon substrate is almost transparent to the beam, while resistor
20
absorbs it. Thus, at these wavelengths, portions of resistor
20
may be vaporized without causing damage to the silicon substrate.
The laser-trimming technique described in the above paragraphs has been successfully employed to trim resistors fabricated on a silicon substrate. Clearly, it would be very desirable to employ the technique to trim resistors fabricated on a gallium arsenide substrate as well. But to the inventors' knowledge, no one had successfully done so at the time of the present invention.
In fact, only one reference has been found that mentions laser trimming of resistors fabricated on gallium arsenide: U.S. Pat. No. 5,569,398 issued to Sun et al. This reference suggests that resistors fabricated on gallium arsenide may be trimmed using a laser with an output wavelength within the range from 1.0 to 3.0 &mgr;m. Sun et al. derives this wavelength range (1.0 to 3.0 &mgr;m) by: (1) identifying wavelengths at which gallium arsenide is believed not to absorb laser light (those wavelengths above 1.0 &mgr;m); (2) identifying the wavelengths at which metallic thin-film resistive materials (such as platinum, nickel, tungsten, and aluminum) are known to absorb laser light (about 0.0 to 3.0 &mgr;m); and (3) comparing the former and the latter wavelength ranges to obtain a range in which the gallium arsenide substrate does not absorb laser light, while the metallic resistive material does: 1.0 to 3.0 &mgr;m.
Sun et al.'s suggestion that gallium arsenide does not absorb laser light having a wavelength from 1.0 &mgr;m to 3.0 &mgr;m is also supported by the experimental results of W. G. Spitzer and J. M. Whelan, as published in
Physics Review,
114, 1 (1959) 59-63 and reproduced herein as FIG.
2
. Spitzer and Whelan studied the relationship between the optical absorption characteristics of gallium arsenide and the doping level in gallium arsenide, for various wavelengths. Specifically, they showed that a strong correlation exists between the conduction electron concentrations induced by various levels of doping and the optical absorption coefficient. (The term “optical absorption coefficient” is defined as a unit of measure of the attenuation caused by the absorption of energy that results from its passage through a medium. Absorption coefficients are usually expressed in units of reciprocal distance. See Terms and Definitions, MIL-STD-2196 (SH), Glossary, Fiber Optics (1989).)
For every doping level tested by Spitzer and Whelan, the optical absorption coefficient was found to be at a local minimum over a wavelength range from about 1.0 to about 4.0 &mgr;m. For example, the absorption coefficient for a conduction electron concentration of 4.9·10
17
cm
−3
is fairly constant at about 3.0 cm
−1
within the wavelength range from 1.0 to 4.0 um. Similarly, the absorption coefficient for a conduction electron concentration of 1.3·10
17
cm
−3
is fairly constant at about 0.5 cm
−1
within the wavelength range from 1.2 to 4.1 um. Thus, based on the results of Spitzer and Whelan, one would expect that as the doping level is decreased, the optical absorption coefficient correspondingly decreases, and the lowest optical absorption coefficient would be obtained when the gallium arsenide was completely undoped.
One would also expect from the results of Spitzer and Whelan that for undoped gallium arsenide, the absorption coefficient would remain at a local minimum throughout the range from about 1.0 to 4.0 &mgr;m. One would expect, accordingly, that any wavelength within the range suggested by Sun et al. (1.0 to 3.0 &mgr;m) would, in fact, be suitable for trimming a resistor fabricated on undoped gallium arsenide. But the inventors of the present invention have found that this is not the case. Through a series of experiments carried out under their direction, they have found, rather, that the critical range of laser wavelengths at which undoped gallium arsenide does not absorb laser light is much narrower: about 0.9 &mgr;m to about 1.5 &mgr;m. The inventors found that trimming with laser light having a wavelength shorter than 0.9 &mgr;m or longer than 1.6 &mgr;m caused damage to the gallium arsenide substrate. The inventors also found that trimming with laser light having a wavelength of 1.047 &mgr;m produced the best results. These results were wholly unexpected in light of the teachings of Sun et al. and Spitzer and Whelan.
(The inventors note that their experiments were, in fact, carried out using gallium arsenide that was carbon-doped to a carbon concentration of between 1.0·10
15
cm
−3
and 5.0·10
15
cm
−3
. Without this weak n-type doping, deep-level donors (impurities) in the gallium arsenide substrate would have caused the substrate to be slightly p-type. The substrate would thus have been slightly conductive, rather than semi-insul

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Laser-trimmable digital resistor does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Laser-trimmable digital resistor, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Laser-trimmable digital resistor will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-3184007

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.