This application relates to integrated circuit package substrates, and more particularly to a low-profile package with a passive device.
In a passive-on-glass (PoG) package, passive components such as inductors and capacitors are integrated onto a glass substrate. The PoG package may then be coupled to a circuit board along with semiconductor packages to form a complete working device such as a radio frequency (RF) front end. As compared to a conventional coupling of discrete passive devices to a circuit board, the use of a PoG package is much more compact. In addition, a PoG package is less expensive than integrating the passive devices into the dies containing the active devices for an electronic system because glass substrates are relatively inexpensive in comparison to crystalline semiconductor substrates.
Although a PoG package is thus an attractive alternative for providing passive components for an electronic system, PoG design faces a number of challenges. In particular, there is an ever-increasing need to reduce the dimensions of the electronics incorporated into mobile devices. As users demand more compact devices, the electronics contained within the devices must shrink in size accordingly. One of the dimensions that must shrink for a PoG package is its height with regard to the underlying circuit board. A straightforward way to reduce the PoG package height is to reduce the thickness of its glass substrate. But glass is inherently brittle. A glass substrate is thus prone to cracking if its thickness is reduced excessively such as less than 150 or 100 microns. The problem does not go away if the passive components are instead integrated onto a semiconductor substrate because such substrates also are brittle and become too fragile if excessively thinned. Since the issues are largely the same regardless of the type of substrate used to support passive components, the term “passive-on-package” is used herein to denote a package containing passive components integrated onto a glass, semiconductor, or organic substrate.
Another issue with reducing the glass substrate thickness is the inductance for embedded inductors formed by through-substrate vias within the glass substrate. The coil or loop for each embedded inductor is formed by a pair (or more) of the through-substrate vias. For example, a first through-substrate via in an embedded inductor may extend from a first surface of a substrate to a lead or conductor formed on an opposing second surface for the substrate. The conductor also couples to a second through-substrate via in the embedded inductor that extends from the second surface back to the first surface. Current driven into the first through-substrate via from the first surface will thus flow through the conductor on the second surface and loop back down to the first surface in the second through-substrate via. This current loop provides the inductance for the resulting embedded inductor. The inductance depends upon (among other factors) the area encompassed by the current loop. If the through-substrate via lengths are decreased by thinning the substrate, the resulting inductance for the embedded inductor will also shrink. As the thickness for the substrate is reduced, the height or length of through-substrate vias through such a reduced-thickness substrate is of course reduced accordingly. For example, a substrate that is 200 microns thick may have through-substrate vias that extend through such a thickness and thus also have a corresponding length of 200 microns. But if the substrate is just 100 microns thick, the through-substrate vias would then have a length of just 100 microns. Reducing the package height for a PoG package will thus tend to reduce the inductances for its inductors. The necessary inductance is thus also a barrier to reducing PoG package heights.
The solder balls or other types of interconnects that couple a passive-on-package to the underlying circuit board are another factor that limit passive-on-package height reduction. To better illustrate these challenges in passive-on-package design, a conventional passive-on-package 100 is shown in
Accordingly, there is a need in the art for more compact package designs with passive devices.
To provide a low-profile package substrate including a passive device, a first side of a substrate includes a plurality of recesses. As used herein, a low-profile package substrate including a passive device may also be denoted as passive-on-package. Each recess receives a corresponding interconnect such as a solder ball or metal pillar. A redistribution layer on the first side of the substrate electrically couples to at least a subset of the interconnects. The substrate includes a plurality of through-substrate vias. In one embodiment, a pair of the through-substrate vias forms an embedded inductor. The redistribution layer may include a lead or conductor that extends from a first one of the recesses to one of the through-substrate vias forming the inductor. In this fashion, the interconnect received in the first recess electrically couples through the conductor in the redistribution layer to the first through-substrate via in the embedded inductor. The substrate may include additional embedded inductors having through-substrate vias coupled to corresponding interconnects through the redistribution layer in this fashion.
Embodiments of the present disclosure and their advantages are best understood by referring to the detailed description that follows. It should be appreciated that like reference numerals are used to identify like elements illustrated in one or more of the figures.
A low-profile passive-on-package is provided that includes a first side having plurality of recesses. Each recess may receive a corresponding interconnect such as a solder ball, a metal post, or a metal cylinder. The following discussion will be directed to a solder ball interconnect embodiment but it will be appreciated that other suitable types of interconnects may be used in alternative embodiments. The substrate also includes a plurality of through substrate vias that extend from the first surface to an opposing second surface of the substrate. A redistribution layer on the first side of the substrate electrically couples to one or more of solder balls in the recesses. For example, the redistribution layer may comprise a patterned metal layer that forms leads or conductors coupled to corresponding ones of the solder balls received in the recesses. A redistribution layer conductor couples between a corresponding solder ball to an end of a corresponding through-substrate via. Since the redistribution layer is adjacent the first surface of the substrate, the end of a through-substrate via that the redistribution layer conductor couples to is also adjacent the first surface.
A pair (or more) of the through substrate vias may be coupled together through a conductor on the second surface of the substrate to form an embedded inductor. For example, the redistribution layer may include a first conductor extending from an interconnect in a first one of the recesses to a through-substrate via in the embedded inductor. Similarly, the redistribution layer may include a second conductor extending from an interconnect in a second one of the recesses to another through-substrate via in the embedded inductor. The interconnect in the first recess thus electrically couples through the embedded inductor to the interconnect in the second recess. In this fashion, a current driven from the interconnect such as a solder ball in the first recess conducts through the embedded inductor to, for example, a solder ball in the second recess. This is quite advantageous because the embedded inductor may have a relatively robust inductance as each of its through-substrate vias is relatively long in that they extend from the first side for the substrate to the opposing second side. Yet the resulting passive-on-package has an advantageously low profile because the solder balls are received in the recesses. The portion of each solder ball that is received in the corresponding recess makes no contribution to the package height.
In addition, the substrate may include through-substrate vias that extend from a corresponding ones of the recesses to the opposing second surface of the substrate. To distinguish between the various through-substrate vias, a through-substrate via extending from the first side of the substrate to the opposing second side is denoted herein as a first through-substrate via. In contrast, a through-substrate via extending from a recess to the opposing second side of the substrate is also denoted herein as second through-substrate via. A second through-substrate via is shorter than a first through-substrate via by the depth of the corresponding recess. This reduced length is advantageous when driving an integrated capacitor on the second surface of the substrate such as a metal-insulator-metal (MIM) capacitor because the reduced length of the second through-substrate via coupling to the capacitor has less parasitic resistance and inductance as compared to a coupling from a first through substrate via. This is quite advantageous because a substrate may be relatively thick so as to be robust against breakage and warpage and so as to support relatively-long first through-substrate vias that provide increased inductance to embedded inductors yet the same substrate supports second through-substrate vias that may drive integrated capacitors with reduced parasitic resistance and inductance.
Given the receipt of the interconnects such as solder balls in the substrate recesses, the substrate need not be excessively thinned and the solder balls may still have a sufficiently robust diameter to resist cracking yet the resulting passive-on-package has a reduced thickness or height because the solder balls are received in the blind vias or recesses. Since the substrate need not be excessively thinned, the substrate may have a thickness that is sufficiently large so as to be robust to breakage and warpage. In addition, note that embedded inductors formed using a pair of through-substrate vias extending through the substrate benefit from the relatively robust substrate thickness despite the resulting passive-on-package having a reduced height due to the solder-ball-receiving recesses. As discussed earlier, an inductor's inductance is a function of the loop area enclosed by the winding or coil forming the inductor. With regard to the embedded inductors disclosed herein, the inductor coil may be formed by a pair (or more) of first through-substrate vias. The substrate may then have a thickness of a sufficient magnitude to achieve a robust inductance from the inductor yet the package height is reduced because the solder balls are received in the corresponding recesses.
In addition, the thickness for the substrate may be sufficiently robust so as to reduce substrate fragility, warpage, and breakage yet the package height is reduced because the solder balls are received in the corresponding recesses. Similarly, the solder balls may each have a sufficiently robust diameter so as to reduce cracking and increase board level reliability. Although the solder balls may have such a robust diameter, these diameters only partially contribute to the package height due to the solder balls being received within the recesses. These and other advantages may be better appreciated through the following discussion of example embodiments.
Passive-on-package 200 may include one or more first through-substrate vias such as first through-substrates 202a, 202b, 202c, and 202d extending from first surface 208 of substrate 204 to an opposing second surface 206 of substrate 208. First through-substrate via 202a couples through a lead or conductor 203a on second surface 206 of substrate 204 to first through-substrate via 202a to form an embedded inductor 215. Similarly, first through-substrate via 202d couples through a conductor 203b to first through-substrate via 202c to form an embedded inductor 217. Each embedded inductor 215 and 217 has an advantageously robust inductance since a thickness T for substrate 204 is not excessively thinned. For example, the current loop area encompassed by inductor 215 is a function of the length (among other factors) of each first through-substrate via 202a and 202b. In turn, the first through-substrate via lengths are a function of the thickness T for substrate 204. Since the thickness T need not be excessively reduced to achieve an advantageously low package height H2 for passive-on-package 200, first through-substrate vias such as vias 202a and 202b may be relatively long to provide enhanced inductance for inductor 215.
The coupling to inductors 215 and 217 may occur through a redistribution layer 220. For example, a solder ball 212 received in a recess 214a couples to first through-substrate via 202b in inductor 215 through a redistribution layer conductor 216a and recess pad 210 formed from redistribution layer 220. Another solder ball may couple through an analogous redistribution layer conductor and pad (not illustrated) to first through-substrate via 202a to complete the coupling to inductor 215. An analogous coupling may be provided with regard to embedded inductor 217. For example, a solder ball 212 received in a recess 214c couples to first through-substrate via 216b in inductor 217 through a redistribution layer conductor 216b and a recess pad 210. In one embodiment, redistribution layer 220 may be deemed to comprise a means for electrically coupling certain ones of the recess-received interconnects to corresponding ones of the first through-substrate vias.
In contrast to the first through-substrate vias, second through-substrate vias have a reduced length. For example, a second through-substrate via 202e extends from a recess 214b to second surface 206 of substrate 204. As compared to a length for the first through-substrate vias that substantially equals the thickness T for substrate 204, second through-substrate via 202e has a length that is shortened by the depth or height of recess 214b. This reduced length reduces the parasitic inductance and resistance in a coupling of second through-substrate via 202e to a capacitor 207 integrated onto surface 206 of substrate 204. In one embodiment, capacitor 207 may comprise a metal-insulator-metal (MIM) capacitor.
Recesses 214 may also include an adhesive (not illustrated) to aid in retaining solder balls 212. The first and second through-substrate vias 202 may serve both an electrical coupling function as well as a heat transfer role. Second through-substrate vias are particularly useful for heat transfer from second surface 206 to solder balls received in the corresponding recesses due to their reduced length as compared to first through-substrate vias. A passivation layer or solder resist layer 230 may cover second surface 206. Similarly, a passivation layer or solder resist layer 225 may cover first surface 208 of substrate 204. Passivation layers 230 and 225 may comprise a wide variety of suitable materials such as silicon nitride, dielectric polymers such as polymide, or organic polymers.
A passive-on-package 300 shown in
A plan view of surface 208 of a substrate 360 for an exemplary passive-on-package is shown in
The enhanced thickness T for the disclosed substrates such as substrate 204 shown in
The following discussion will be directed to wafer-level-process (WLP) embodiments in which the substrate used to support passive components in a passive-on-package is processed as part of a wafer (or panel) before being diced into individual packages. But it will be appreciated that the processes discussed herein may also be applied individually to substrates that have been diced from the wafer as compared to processing the wafer (or panel) as a unit. Regardless of whether a WLP process is used to manufacture passive-on-packages to achieve a reduced height, the reduced-height passive-on-packages disclosed herein receive interconnects such as solder balls within corresponding blind vias or recesses.
An example manufacturing process flow is shown in
As shown in
Surface 208 may then be etched or drilled to form blind vias or recesses 214 as shown in
As illustrated in
A flowchart for an example method of manufacture is shown in
The passive-on-packages disclosed herein may be incorporated into a wide variety of electronic systems. For example, as shown in
As those of some skill in this art will by now appreciate and depending on the particular application at hand, many modifications, substitutions and variations can be made in and to the materials, apparatus, configurations and methods of use of the devices of the present disclosure without departing from the spirit and scope thereof. In light of this, the scope of the present disclosure should not be limited to that of the particular embodiments illustrated and described herein, as they are merely by way of some examples thereof, but rather, should be fully commensurate with that of the claims appended hereafter and their functional equivalents.
This application claims the benefit of U.S. Provisional Application No. 61/941,308, filed on Feb. 18, 2014, the contents of which are incorporated by reference in their entirety.
Number | Date | Country | |
---|---|---|---|
61941308 | Feb 2014 | US |