Patent Application
20090152546
The present invention relates to the domain of integrated circuits, and more precisely to the test of integrated circuits (or dies) defined in wafers.
As it is known by one skilled in the art dies must be tested before being integrated with electronic equipments. For practical reasons they are tested when they still belong to their wafer, i.e. before being separated one from the other by a cut-off process along scribe lanes (or lines) defined therebetween.
Dies are tested alone or in parallel by means of a probe card controlled by automated test equipment (ATE). To be tested, a die must be provided with internal pads connected to some of its integrated components, generally through internal test circuits that are only used during the test.
When a die comprises first and second complementary signal processing parts (or dual parts), each part must be tested. This applies to any type of dual circuit, such as transmitting and receiving paths (in case of radio communication equipment), digital to analog converter (DAC) and analog to digital converter (ADC), or else demodulator and modulator, for instance.
Most often, dual parts (such as receiving and transmitting paths) are tested independently. Sometimes, a loop back configuration is used. In this case either a radiofrequency (RF) transmit signal is connected to the RF receive input, or the receive base-band output signal of the receiving path is fed back to the input of the transmitting path.
An advantage is that both receiving path and transmitting path are tested at the same time, which reduces test time and, therefore, cost. Another advantage is that there is no more need for example to generate an RF signal from the ATE, which reduces cost of the ATE. The connections run through the needles and through some adaptive circuits located on the probe card to which the needles are connected.
Such a loop back configuration may require definition of some internal test circuits within the dies, to ease wafer testing. The die cost in terms of area for adding these internal test circuits might be too high.
So, the object of this invention is to improve the situation.
For this purposes, it provides a wafer comprising at least one die comprising first and second complementary signal processing parts, this first part having an input to receive first input signals and an output to deliver first output signals and this second part having an input to receive second input signals and an output to deliver second output signals, and scribe lanes defined between and around each die.
By “first and second complementary signal processing parts” is meant a first signal processing part of a die delivering output signals which can feed the input of a second signal processing part of the same die or of another die (possibly through an intermediate signal processing means), which can itself deliver output signals which can feed the input of a first signal processing part of the same die or of another die (possibly through an intermediate signal processing means). So, they might be a transmitter path and a receiver path or a transceiver, or a digital to analog converter (DAC) and an analog to digital converter (ADC), or else a demodulator and a modulator, for instance.
This wafer is characterized in that it comprises at least a coupling means defined in at least a part of the scribe lanes and connecting:
The wafer according to the invention may have additional characteristics considered separately or combined, and notably:
Other features and advantages of the invention will become apparent on examining the detailed specifications hereafter and the appended drawings, wherein:
The appended drawings may not only serve to complete the invention, but also to contribute to its definition, if need be.
The invention aims at reducing the number of integrated components that are defined in dies exclusively for test purposes, when they still belong to their wafer and when they comprise first and second complementary signal processing parts.
As is schematically illustrated in
In the following description it will be considered that the wafer comprises several dies that are intended to be integrated with the transceiver of communication equipment such as a mobile telephone adapted to radio communication, for instance in a GSM or UMTS network. But it is important to notice that the invention is not limited to this type of electronic equipment and to wafers comprising several dies. It applies to any circuit comprising complementary signal processing parts implementing complementary functions, such as transmitter path/receiver path, DAC/ADC and demodulator/modulator, for instance. Moreover, it applies to any type of signal to be processed, such as radiofrequency (RF) signals, analog signals and digital signals, for instance.
Moreover, in the following description it will be considered that each die D (defined in the wafer W) comprises first and second complementary RF signal processing parts which are respectively a transmitting path, having an input to receive first input signals (for instance baseband signals) and an output to deliver first output signals (for instance RF signals), and a receiving path having an input to receive second input signals (for instance RF signals) and an output to deliver second output signals (for instance baseband signals).
The invention offers to integrate at least some of the integrated circuits, which are required for testing dies, with some of the scribe lanes SL.
More precisely and as illustrated in the first example of embodiment sketched in
The coupling means CM is arranged to connect:
As illustrated in
The following description of the first example illustrated in
A first conductive track T1 is connected to the input of the first part (transmitting path) P1 of each or some dies D to be tested in order to feed it (or them) with first input signals (for instance baseband signals).
In the example illustrated in
For instance and as illustrated in
As mentioned before, the (or each) coupling means CM is arranged to couple the output of the first part P1 of at least one die D1 and/or D3 to the input of the second part P2 of at least one other die D2 and/or D4.
In the first example illustrated in
So, when a die D1 (or D3) is configured through the (first) bus B1 to work with its first part P1 and a second die D2 (or D4), coupled to this die D1 (or D3) through the coupling means CM, is configured through the (second) bus B2 to work with its second part P2, the die D1 (or D3) receives first input signals (for instance baseband signals) on the input of its first part P1 and delivers first output signals (for instance RF signals) on the output of its first part P1, and the die D2 (or D4) receives these first output signals on the input of its second part P2, through the coupling means CM, and delivers second output signals (for instance baseband signals) on the output of its second part P2. Therefore, the result of the test of the first part P1 of a die D1 (or D3) is used to test the second part P2 of another die D2 (or D4).
A second conductive track T2 is connected to the output of the second part (receiving path) P2 of each or some dies D to be tested in order to collect the second output signals (for instance baseband signals) it delivers (they deliver).
By defining sequences of states for the different switch means SWi and corresponding sequences of configuration signals for the different dies D it is thus possible to test automatically the first and second parts P1, P2 of every die D, without mechanically moving the probe card.
Reference is now made to
In the following description of the second example illustrated in
This second example aims at overcoming a drawback of the first example. Indeed, in the first example, if the test of a transmitting part (P1) of a first die with the receiving part (P2) of a second die fails, then it is not possible to know whether the first die and/or the second die is/are faulty. So the second example adds to the first example the capability to connect the first part P1 of a die to the second parts P2 of several other dies and conversely, to connect the second part P2 of a die to the first parts P1 of several other dies, to determine whether it is the first part or the second part of a die that is faulty.
For this purposespurposess, the wafer W comprises at least two groups Gj (here j=1 and 2) of at least three dies coupled by means of a coupling means CM comprising at least one switch means SW3.
Such a switch means SW3 comprises at least two inputs respectively connected to the output of the first part P1 of at least two dies D1 and D3 (in group G1), or D5 and D7 (in group G2), and at least one output connected to the input of the second part P2 of at least one other die D2 and D4 (in the same group G1), or D6 and D8 (in the same group G2).
Each switch means SW3 is arranged to selectively feed the second part input of one die of its corresponding group Gj with the first output signals delivered by the first part output of another selected die of this group Gj.
In the second example illustrated in
With such an arrangement four combinations of dies may be tested in each group Gj depending on the chosen states of the corresponding switch means SW3. Indeed, the output of the first part P1 of the first die D1 (or D5) may be coupled to the input of the second part P2 of either the second die D2 (or D6) or the fourth die D4 (or D8) to feed it with the first output signals it delivers, and the output of the first part P1 of the third die D3 (or D7) may be coupled to the input of the second part P2 of either the second die D2 (or D6) or the fourth die D4 (or D8) to feed it with the first output signals it delivers.
As in the first example, one uses first T1 and second T2 conductive tracks. A first conductive track T1 feeds the input of the first part P1 of the odd dies D1 and D3 (or D5 and D7) of each group Gj with the first input signals (for instance baseband signals). A second conductive track T2 collects the second output signals (for instance baseband signals) delivered by the output of the second part P2 of the even dies D2 and D6 (or D4 and D8).
Although this is not illustrated in
Moreover, as in the first example, one preferably uses first and second buses B1, B2. But this is not mandatory. Two buses are preferred when the number of dies D to be tested is important. For instance the first and second buses B1, B2 are coupled to a control input of each odd die D1 and D3 (or D5 and D7) and even die D2 and D4 (or D6 and D8) respectively to feed it with configuration signals, when required.
For instance and as illustrated in
By defining sequences of states for the different switch means SWi and SW3 and corresponding sequences of configuration signals for the different dies D it is thus possible to test automatically the first P1 and second P2 parts of every die D, without mechanically moving the probe card.
This second example is of great interest because it facilitates the discovery of a faulty part P1 or P2 in each tested die D, and reduces notably the yield loss during die tests.
Reference is now made to
In the following description of the third example illustrated in
In this third example, the wafer W comprises a first group of at least two odd dies D1, D3, . . . and a second group of at least two even dies D2, D4, D6, . . . . The output of the first part P1 of each odd die D1 (or D3) is coupled to the input of the second part of each even die D2 (or D4 or D6) to be tested, by means of a common coupling means CM comprising one pair of switch means SW4 associated to each odd die D1 (or D3).
Each switch means SW4 comprises two bidirectional inputs/outputs respectively connected, on the one hand, to the output of the first part P1 of one odd die D1 (or D3) and to an input/output of the neighboring switch means SW4 (except for the first and the last one), and on the other hand, to the input of the second part P2 of one even die D2 (or D4 or D6) and to an input/output of another neighboring switch means SW4 (except for the first and the last one). So, each switch means SW4 is arranged to selectively feed either the second part input of one even die D2 (or D4 or D6) or a neighboring switch means SW4 with the first output signals delivered by the first part output of the odd die D1 (or D3) or of a distant odd die.
With such an arrangement any combination of odd and even dies may be tested depending on the chosen states of each switch means SW4.
As in the first example, one uses first and second conductive tracks T1, T2. A first conductive track T1 feeds the input of the first part P1 of the odd dies D1 and D3 (or D5 and D7) with the first input signals (for instance baseband signals). A second conductive track T2 collects the second output signals (for instance baseband signals) delivered by the output of the second part P2 of the even dies D2 and D6 (or D4 and D8).
Although this is not illustrated in
Moreover, as in the first example, one preferably uses first and second buses B1, B2. But this is not mandatory. Two buses are preferred when the number of dies D to be tested is important. For instance the first and second buses B1, B2 are respectively coupled to a control input of each odd die D1 and D3 and even die D2, D4, and D6 to feed it with configuration signals, when required.
For instance and as illustrated in
By defining sequences of states for the different switch means SWi and SW4 and corresponding sequences of configuration signals for the different dies D it is thus possible to test automatically the first and second parts P1, P2 of every die D, without mechanically moving the probe card.
This third example is also of interest because it facilitates the discovery of a faulty part P1 or P2 in each tested die D, and minimizes yield loss during die tests.
In each example of embodiment a coupling means comprises conductive tracks and possibly at least one active switch means. But for test purposes a coupling means may also comprise one or more components dedicated to signal processing, such as a load or a voltage level controller in case of RF signals, or a frequency converter, or buffers or digital logic in case of digital signals, or else amplification circuit or filters in case of analog signals. These signal processing components may allow to apply at least one chosen processing to the first output signals delivered by the output of a first part P1, before they feed the input of a second part P2, and/or to the second output signals delivered by the output of a second part P2, before they feed the input of a first part P1.
Moreover, it is important to notice that in each example of embodiment:
The dies may be realized in the wafer in CMOS or BiCMOS technology or in any technology used in chip industry fabrication.
The invention is not limited to the embodiments of the wafer described above, only used as examples, but it encompasses all alternative embodiments which may be considered by one skilled in the art to be within the scope of the claims hereafter.
So, in the preceding description parts of wafers have been described in the scribe lanes of which buses and conductive tracks were defined either for feeding the first part of some dies with first input signals or for collecting the second output signals delivered by the second part of some other dies, for clarity. But, for the first and second parts of each die to be testable, some other conductive tracks may be used.
| Number | Date | Country | Kind |
|---|---|---|---|
| 05300779.5 | Sep 2005 | EP | regional |
| PCT/IB2006/053477 | Sep 2006 | IB | international |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/IB2006/053477 | 9/25/2006 | WO | 00 | 10/24/2008 |
