The present invention generally relates to the field of integrated circuits. In particular, the present invention is directed to a system and method for testing integrated circuits.
Mixed-signal integrated circuits are becoming increasingly prevalent with the continually increasing scale of circuit integration that has spurred the development of system on chip (SOC) design and other types of design integration. Concurrent with increasing integration scale, the operating speeds of circuits has also increased. The increases in both integration scale and speed has raised the complexity and difficulty of testing these circuits and their component circuitry. As a result, design for testability (DFT), built-in self test (BIST) and embedded test core testing schemes have been employed to facilitate testing of these circuits. However, each of these schemes has one or more drawbacks, especially as integration scale and operating speeds continue to increase.
Generally, DFT is a design approach in which integrated circuit designs are partially dictated by testing considerations. Typically, DFT includes providing ready access by automated testing equipment (ATE) to certain portions of the integrated circuits, e.g., by providing externally accessible test pads that allow ATE to communicate with the portion of the circuit being tested. One drawback of DFT is that with increasing speeds, the relatively slow speeds of conventional ATE and the relatively long signal propagation delay attendant the physical distance between the ATE and the circuit under test (CUT) hamper testing. Another drawback is that the manner in which conventional ATE conduct testing reduces testing speeds. For example, conventional ATE may perform a particular test by storing a large amount of measurement data from a particular test and then processing the data after the measurement data has been collected. This sort of testing/processing can result in relatively long testing times, which are undesirable, particularly with high-production circuits.
BIST and embedded testing core schemes are similar to one another in that they require dedicated testing circuits to be added to the integrated circuits. Drawbacks of BIST and embedded testing core schemes include the need to add often-complex testing circuits to the integrated circuits. As levels of integration of SOCs and similar devices increase, the number of additional testing circuits needed likewise increase. Generally, if given a choice, circuit designers would prefer not to utilize valuable chip space for testing circuits.
In view of the foregoing, what are needed, among other things, are a system and method for testing digital/analog integrated circuits, and/or portions of the circuitry thereof, that minimize impact on chip space and minimize testing time, yet are capable of testing high-speed circuits. It is also desirable that, in some embodiments, such a system and method be “backward-compatible” for use with conventional and legacy ATE. The system and method of the present invention satisfy these and other needs.
In one embodiment, the present invention is directed to an integrated testing system for testing circuitry of a DUT via an interface board having a first socket for receiving the integrated testing system and a second socket for receiving the DUT. The integrated testing system comprises a module operatively configured to engage the first socket of the interface board. The module includes at least one measurement engine operatively configured to electrically excite the circuitry, measure a response of the circuitry and generate measurement data when the integrated testing system is engaged with the first socket and the DUT is engaged with the second socket. The module further includes a compute engine operatively configured to perform at least one computation on said measurement data.
In another embodiment, the present invention is directed to a method of testing circuitry of a DUT. The method comprises the steps of engaging a module containing an integrated testing system with an interface, engaging a DUT with said interface and causing said integrated testing system to conduct a test of circuitry of said DUT.
For the purpose of illustrating the invention, the drawings show a form of the invention that is presently preferred. However, it should be understood that the present invention is not limited to the precise arrangements and instrumentalities shown in the drawings, wherein:
Referring to the drawings,
Interface board 112, e.g., may be a conventional device interface board (DIB) used in conjunction with ATE 116. Interface board 112 may include one or more DUT sockets 120 for receiving a corresponding number of DUTs (104), as well as one or more tester-support sockets 124 for receiving various testing modules or testing boards, one or more of which may include integrated testing system 108. As used herein and in the appended claims, the term “socket” and like terms are used in a broad sense to denote any structure(s) that are part of, and/or engage, interface board 112 so as to substantially fix integrated testing system 108 relative to the board and electrically connect the integrated testing system to the board. On conventional DIBs, tester-support sockets 124 are generally provided for receiving various modules and/or boards (not shown), such as signal conditioning boards, among others, that support testing via ATE 116. Each DUT socket 120 may include a plurality of electrical connectors/contacts 128, e.g., pogo pins, among others, for electrically connecting the corresponding DUT 104 to interface board 112. Similarly, each tester-support socket 124 may include electrical connectors (not shown), e.g., pin contacts, among others, for electrically connecting the corresponding test module, in this case integrated testing system 108, to interface board 112. Interface board 112 may also include other conventional electronics (not shown) for controlling the board, and communications ports 132 for electrically connecting ATE 116 to the board. In other embodiments of testing setup 100 of the present invention, the interface may be standalone device characterization board (not shown), e.g., a board similar to interface board 112, but not including the electronics and communications ports 132 needed for communicating with ATE 116. ATE 116, if present, may be a conventional ATE, e.g., a legacy ATE, or may be specially adapted for use with integrated testing system 108 of the present invention.
Testing setup 100 may optionally include a host computer 136 containing a user interface 140, e.g., a graphical user interface, operatively configured for programming and/or controlling the operation of integrated testing system 108. Host computer 136 may be integrated into integrated testing system 108 or may alternatively be located remotely from the integrated testing system, e.g., across one or more computer networks, such as a local area network (LAN) 144 and wide area network (WAN) 148, including the Internet. Implementation of user interface 140 could allow a user to operate integrated testing system 108 from virtually any suitable user device 152 from virtually any location. Implementation of user interface 140 is described below in more detail.
As shown in
At a high level, important features of integrated testing system 108 include, among others: 1) its ability to be configured to process measurement data “on-the-fly,” i.e., while one or more measurement engines are proceeding with testing; 2) its ability to be configured to interface with a wide variety of ATE 116, including legacy and conventional ATE; and 3) its ability to be configured to interface with dedicated user interface 140 outside the context of ATE. As discussed below, each measurement engine 200, 202, communications engine 204 and compute engine 208 may be implemented using various hardware and software schemes.
In general, primary tasks of measurement engines 200, 202 include providing stimulus (or stimuli) to DUT 104 and measuring the DUT's response(s) to that stimulus (stimuli). Both of these functions may performed via an interface, e.g., interface board 112, that provides one or more electrical communications pathways (not shown) between DUT 104 and measurement engines 200, 202. To perform their respective functions, each measurement engine 200, 202 may comprise one or more stimulus instruments 212 for providing a corresponding number of stimulus signals to DUT 104. As discussed below, stimulus instruments 212 may be controlled separately or in one or more groups and in seriatim or in parallel to provide the desired stimulus (stimuli) to DUT 104, depending upon the nature of the testing being performed. Similarly, measurement engine 200, 202 may additionally or alternatively include one or more measurement instruments 216 for measuring the response(s) of DUT 104 to the one or more stimulus signals. Similar to stimulus instruments 212, measurement instruments 216 may be controlled separately or in one or more groups and in seriatim or in parallel to measure the desired response(s) of DUT 104, again depending upon the nature of the testing being performed.
In addition, depending on the nature of the testing, the number of measurement instruments 216 may or may not be the same as the number of stimulus instruments 212. For example, when testing an 8:1 serializer, measurement engine 200 may utilize eight stimulus instruments 212, i.e., one for each of the eight parallel inputs, and only one measurement instrument 216 for measuring the response signal of the serializer on its single serial output. Conversely, when testing a 1:8 deserializer, measurement engine 202 may utilize only one stimulus instrument 212 for exciting the serial input of the deserializer and eight measurement instruments 216 for measuring the deserializer's response on the eight parallel outputs. Of course, measurement engines 200, 202 may have greater than the number of stimulus instruments and/or measurement instruments needed for a particular test. This may be the case when a “general purpose” or highly flexible version of integrated testing system 108 is used for testing circuitry having fewer inputs and/or outputs than the number of stimulus and measurement instruments 212, 216 aboard the integrated testing system. In such cases, integrated testing system 108 can be controlled in a manner that only those of instruments 212, 216 needed for a particular test are utilized.
Stimulus measurement instruments 212 may be implemented using any suitable hardware or hardware/software scheme. For example, the present inventors have found that certain types of stimulus instruments 212, e.g., arbitrary waveform generators (AWGs) and timing generation units (TGUs), among others, and certain types of measurement instruments 216, Vernier delay lines and digitizers (DIGs), are effectively implemented in custom microcircuitry, e.g., application specific integrated circuits (ASICs). Implementing such instruments 212, 216 in discrete custom circuitry can provide a good compromise in size, cost and performance. Of course, even these instruments may be implemented in other ways, if desired, such as off-the-shelf circuitry or programmable or reprogrammable logic devices (PLDs or RLDs), among others. Those skilled in the art will readily appreciate the variety of ways stimulus and measurement instruments 212, 216 may be implemented, such that neither an exhaustive list nor detailed description of each is necessary for an appreciation of the broad scope of the present invention and to enable skilled artisans to make and use the present invention to its fullest scope.
Compute engine 208 may be configured to function as a central processor and central controller of integrated testing system 108. That is, compute engine 208 may be utilized, among other things, to: 1) setup and control the operation of stimulus instruments 212; 2) process and/or provide input, e.g., stimulus parameters, to the stimulus instruments; 3) initialize and control the operation of measurement instruments 216; 4) receive output, e.g., digital measurement data, from the measurement engines and process this output into desired results; 5) provide digital input directly to DUT 104; 6) initialize, communicate with and control communications engine 204; 7) perform self-testing; and 8) configure any RLDs utilized in integrated testing system 108, e.g., in the measurement engines or the compute engine itself. Of course, compute engine 208 need not provide all of this functionality if a certain implementation of integrated testing system 108 does not require it. In such cases, only the functionality(ies) desired for that implementation need be provided to integrated testing system 108.
Compute engine 208 may be implemented in any suitable hardware or hardware/software scheme, including, but not limited to, single and multiple chip solutions. Exemplary implementations include full digital signal processor (DSP) implementations, RLD implementations using, e.g., one or more field programmable gate arrays (FPGAs), and implementations comprising both DSP and RLD components, among others. Utilizing RLD technology for compute engine 208 can provide a flexible solution, particularly when it is desirable to customize integrated testing system 108 to each of a variety of testing applications, while enjoying the economies of manufacturing generic hardware. As those skilled in the art will appreciate, any portion(s) of measurement engines 200, 202, e.g., measurement instruments 216, and any portion(s) of compute engine 208, e.g., computation logic 800A-F (
For certain applications of integrated testing system 108, e.g., the testing of high-speed SerDes circuitry, among many others, the speed of the system is paramount. The integration of compute engine 208 into integrated test system 108 allows testing and processing speeds to be readily optimized and maximized, if needed, due to the close physical proximity of the compute engine to measurement engine(s) 200, 202 and DUT 104. This close physical proximity, or coupling, allows signal propagation delays between compute engine 208 and measurement engine(s) 200, 202 to be minimized. For example, compute engine 208 is preferably, but not necessarily within about 6 inches (15.24 cm), more preferably about 3 inches (7.62 cm) of each measurement engine 200, 202, and even closer spacing may be advantageous. In addition, with the ability, discussed below, of compute engine 208 to process measurement data on-the-fly, data storage requirements and processing delays associated therewith are minimized. Therefore, integrated testing system 108 is not only amenable to high-speed testing, but also minimizes overall test time, which includes the time needed to process raw measurement data into desired results.
As mentioned above, communications engine 204 provides integrated testing system 108 with an ATE communications link 228 for communicating with ATE 116, and/or a user-interface communications link 232 for communicating with user interface 140 supported by host computer 136. ATE communications link 228, if provided, may be established on any one or more of the channels, parallel or serial, that link interface board 112, e.g., a DIB, to ATE 116. Generally, integrated testing system 108 may be configured to appear to ATE 116 as a DUT. Such a configuration can be exploited to allow the integrated testing system 108 to interface with ATE 116 via the ATE's serial channels conventionally used to write and read digital data vectors to a DUT during conventional ATE testing. Many types of legacy and conventional ATE equipment include such serial channels. This interface to ATE 116 can allow for initializing, setting up, controlling, programming (including (re)programming RLD 220, if present) and reading and displaying, or otherwise presenting, output from integrated testing system 108. A benefit to this interface is that manufacturers of conventional and legacy ATE would not have to divulge any proprietary software to a manufacturer of integrated testing system 108 that might want the ATE manufacturers to provide their ATE with custom user interfaces. Of course, in alternative embodiments ATE 116 can be provided with custom interfaces for interfacing the ATE with integrated testing system 108. An example of a programming/control interface compatible with utilizing data vector channels of ATE 116 is presented below in connection with
User-interface communications link 232, if provided, allows integrated testing system 108 to communicate with user interface 140 using any one of a number of communications protocols, such as an Ethernet protocol, among many others. User interface 140 may be implemented on host computer 136 in any suitable manner, such as in a platform-independent programming environment, e.g., the JAVA® programming environment developed by Sun Microsystems, Santa Clara, Calif. In this example, host computer 136 could be a network server, such as a Web server, or the like. In an embodiment in which host computer 136 includes a Web server, user interface 140 could be accessible via a Web browser from a user device 152 (
Referring to
As is well-known in the art, SerDes circuitry, such as SerDes circuitry 300, may be used to send and receive parallel data over one or more serial links in a wide variety of applications, such as data communication applications complying with any one or more of a large array of communication standards. These standards include chip-to-chip and board-to-board standards, such as XAUI, and long-distance telecommunication standards, such as SONET. Presently, SerDes circuitry is being designed to operate in a GBPS regime and will likely operate faster in the future. Conventional ATE, standing alone, are often unsuitable for testing GBPS SerDes devices due to their internal clocks being too slow to test these devices at their rated speeds.
In
In order to perform jitter testing of serializer 304 and deserializer 308, integrated testing system 108 may include two measurement engines, measurement engine 200 for testing the serializer and measurement engine 202 for testing deserializer 308. Measurement engine 200 may include four stimulus instruments 212A, 212B, 212C, 212D providing four independent channels SS1, SS2, SS3, SS4 of stimulus (“SS” denoting “serializer stimulus”), each for stimulating a corresponding one of parallel data inputs PDI1, PDI2, PDI3, PDI4. In the context of the jitter testing example being described, each stimulus instrument 212A-D is conveniently called “test generation unit” or “TGU” for short. An exemplary embodiment of each TGU 212A-D is described below in connection with
Similarly, for deserializer 308, measurement engine 202 may include two TGUs 212E, 212F providing two independent stimulus channels DS1, DS2 (“DS” denoting “deserializer stimulus”), each for stimulating a corresponding one of serial data input SDI and clock input ClkI. Like TGUs 212A-D, TGUs 212E-F may use any suitable signaling technology appropriate for the test, such as CML. On the measurement side, measurement engine 202 may include four TMUs 216C, 216D, 216E, 216F that provide four separate channels DM1, DM2, DM3, DM4 (“DM” denoting “deserializer measurement”) for receiving signals, respectively, from parallel data outputs PDO1, PDO2, PDO3, PDO4. Like TMUs 216A-B, TMUs 216C-F may use any suitable signaling technology appropriate for the test, such as LVDS. If deserializer 308 includes the slow clock (not shown) mentioned above, measurement engine 202 may include an additional TMU (not shown) for measuring the output of this clock. Although integrated testing system 108 has been described as comprising two measurement engines 200, 202, those skilled in the art will readily appreciate that all of TGUs 212A-F and TMUs 216A-F may be contained in a single measurement engine or, alternatively, arranged so as to provide more than two measurement engines. An extreme example would be the case in which each of TMUs 216A-F and TGUs 212A-F is considered to be in its own measurement engine.
Each TGU 212A-F may comprise a linear programmable delay circuit such as circuit 400 as shown in
Referring to
Because of the offset-frequency relationship between oscillators 710, 712, their respective phases will eventually align. A simple evaluation of the time it takes oscillators 710, 712 to phase align yields an accurate representation of the original phase difference between the edges of input signals s1, s2 detected at the inputs of flip-flops 704, 706, respectively. A benefit of circuit 700 is its compactness and its inherent linearity over a large measurement range. For example, the present inventors have found that with a 2.0 nanosecond range and a 25 picosecond step size, the non-linearity is less than 0.5 LSB. Circuit 700 shown in the '750 publication mentioned above is generally intended for use within a DUT. However, to execute this circuit outside a DUT, additional low noise receiver circuits (not shown) it will typically be necessary to sense the correct timing information related to the signals in question. Those skilled in the art will understand how to implement such low noise circuits.
In the context of the present jitter testing scenario, the output from any of TMUs 216A-F containing component-invariant vernier delay circuit 700 represents phase relationships of input signals s1, s2. Consequently, various algorithms for computing typical jitter performance parameters, e.g., root-mean-square (RMS), peak-to-peak, mean and median parameters, among others, may be implemented in compute engine 208, e.g., in custom computation logic circuitry 800A-F, as shown in
For example, a peak-to-peak measurement for which the computation is
J=max(X[n])−min(X[n]) for n=1 . . . N {1}
can be computed on-the-fly by using a pair of registers and a comparator (not shown). If compute engine 208 comprises RLD 220 as discussed above, these registers and comparator may be implemented in the RLD. Similarly, a moving average filter can be expressed recursively as
Average[n]=Average[n−1]+X[n] {2}
which may be implemented in hardware as an accumulator (not shown). It is important to note that the moving average filter just discussed is just one example of a filter that could be implemented recursively. As those skilled in the art will appreciate, many others exist.
As yet another example, it is noted that RMS jitter is related to the autocorrelation function of an input sequence at a lag of zero. Consequently, RMS jitter may be expressed recursively as
R[0]=Sum X[n]X[−n] {3}
which is simply a convolution summation of the incoming TMU measurements. A common feature of these examples is that compute engine 208 could complete the necessary computations in little or no time in addition to the time it takes for the samples to be acquired via a sample cycle of corresponding respective ones of TGUs 212A-F and TMUs 216A-F. The inventors have experienced sample acquisition rates of 2 Msamples/sec/channel. Consequently, collecting thousands of samples and computing their jitter statistics could be performed in a few milliseconds.
Other algorithms involving fast Fourier transforms (FFT) and discrete Fourier transforms can also be implemented in an on-the-fly manner. Such transforms enable more sophisticated signal processing operations to be performed on the measurement data, such as auto-correlation and cross-correlation operations. Those skilled in the art will also appreciate that other transforms and methods, including wavelet transform methods, may be implemented for separating measurement signals from noise. Skilled artisans will further appreciate that since the measurements are processed locally in pipeline fashion, the present invention may identify failed parts sooner than the prior art approach of taking all the measurements, storing all the measurements and transferring all the measurements to ATE 116 prior to processing the results.
Although measurement engines 200, 202 and compute engine 208 have just been described in detail in connection with jitter testing of SerDes circuitry 300 for convenience of illustrating various features of these components of integrated testing system 108, as mentioned above the present invention is by no means limited to jitter testing or SerDes testing. Rather, various aspects of the present invention may be implemented across a wide variety of testing types and tested-circuitry types. For example, each measurement engine 200, 202 may include in addition to, or in place of, TGUs 212A-F and TMUs 216A-F, one or more other types of stimulus and measurement instruments (not shown), such as AWGs, DIGs, precision measurement units (used to measure a resulting voltage when a specified current is forced into a circuit node and parametric measurement units used to measure a resulting current when a specified voltage is forced across a pair of circuit nodes, among others. AWGs, DIGs and PMUs can be particularly useful in conducting mixed signal testing.
One example of mixed signal testing utilizing AWGs and DIGs, albeit in the context of embedded core testing, is discussed in U.S. Patent Application Publication No. 20020019962 entitled “Integrated Excitation/Extraction System For Test And Measurement,” published Feb. 14, 2002 in the names of Roberts et al., which is incorporated by reference herein in its entirety. Although the AWGs, DIGs and associated testing scheme described in the '675 application are described in the context of embedded core testing, the general concepts and circuits described in that application are applicable to implementing a similar mixed signal testing scheme in the context of integrated testing system 108 of the present invention. Those skilled in the art will understand the modifications to the AWGs, DIGs and testing scheme of the '675 application needed to implement them externally relative to a DUT, such as DUT 104 of
In order to make integrated testing system 108 more easily programmed by a user, e.g., a test engineer, processor 804 may optionally utilize a set of operation codes, commonly called “opcodes,” that obviate the need for a user to generate programs in micro-code. Opcodes in general are well-known in the art. Relative to integrated testing system 108, those skilled in the art would certainly be able to develop and utilize an appropriate set of opcodes that efficiently utilize the micro-code of processor 804 to perform the necessary functions. However, for the sake of illustration, each opcode may be implemented in a short word, i.e., two bytes, in which the bits are assigned as shown in Table 1.
Again, for the sake of illustration, exemplary opcodes that may be implemented in integrated testing system 108 of the present invention appear in Table 2.
The opcodes may be used to create programs for conducting tests and performing various other tasks, such as calibrating stimulus and measurement instruments 212A-F, 216A-F. Each program may have any suitable format. For the sake of illustration, each program may consist of a plurality of lines, e.g., however many lines needed for the particular task(s) to be performed, each line consisting of a short word. Depending upon the program, configuration of integrated testing system 108, and/or the functions of the various opcodes, among other things, some or all of the lines may be opcodes. The short words of the lines not containing opcodes may contain, e.g., input data for a set of stimulus drivers 808A-F, if used, or stimulus instruments 212A-F. A simple ten-line program for loading stimulus data to memories (not shown) of stimulus instruments 212A-F is illustrated in Table 3.
In Table 3, lines 1-5 include opcodes, e.g., for initializing stimulus and/or measurement instruments 212A-F, 216A-F, memory and state. Line 6 contains the opcode for writing the stimulus data and lines 7-9 contain corresponding short words of the stimulus data to be written to the memory. Line 10 contains the opcode for ending the program. Although not shown, a higher-level programming language, e.g., an assembly language, may be implemented to further simplify the programming process. Such a language could express opcodes in more convenient terms and/or contain functions that each execute multiple opcodes. Those skilled in the art will readily understand how to implement a higher-level language suitable for a particular application.
As mentioned, compute engine 208 may optionally include stimulus drivers 808A-F that communicate with corresponding ones of stimulus instruments 212A-F to load the stimulus instruments with the proper stimulus parameters desired for a particular test. Each stimulus 808A-F driver may comprise a memory (not shown) that stores a number of sets of stimulus parameters. Consequently, when stimulus instruments 212A-F are being set up for a particular test, one or more setup instructions placed in the opcode program may direct stimulus drivers 808A-F to load the corresponding respective stimulus instruments 212A-F with the desired set of stimulus parameters. If stimulus drivers 808A-F are not provided, stimulus parameters may be passed directly to stimulus instruments 212A-F from processor 804 as described above in connection with Table 3.
Briefly mentioned above is the fact that compute engine 208 may optionally be provided with one or more digital links 812 for connecting it directly to DUT 104 via an interface, e.g., interface board 112. In this manner, digital circuitry 816 aboard DUT 104 may be tested using conventional digital testing techniques. Since such techniques are well known in the art, it is not necessary to list them or describe them herein for those skilled in the art to understand how to make and use the present invention to its fullest scope. Like measurement data from measurement instruments 216A-F, the digital testing response of DUT 104 may be analyzed by compute engine 208 or passed out of integrated testing system 108 via an interface, such as interface board 112, e.g., to ATE 116 or user interface 140 for analysis. If the digital testing response is analyzed by compute engine 208, the compute engine may send only a pass/fail flag or analysis summary out of integrated testing system 108.
When AWGs, DIGs and/or PMUs are used, stimulus drivers 808A-F and computation logic circuits 800A-F would need to be augmented or replaced in order to provide the proper stimulus input and algorithms necessary to support these instruments. Of course, those skilled in the art will appreciate that the stimulus drivers 808A-F and computation logic 800A-F for supporting these instruments will be application specific and that it is neither practical nor necessary to exhaustively list and describe all possible alternative for those skilled in the art to understand how to implement the present invention in these and other applications.
As mentioned above, if any one or more of compute engine 208 and measurement engines 200, 202 includes a PLD or an RLD such as RLD 220 and RLD 820, processor 804 may communicate with microcontroller 224 for the purpose of (re)programming each PLD or RLD as needed. Again, depending upon the design of integrated testing system 108, various components of compute engine 208, e.g., computation logic circuits 800A-F and stimulus drivers 808A-F, and measurement engines 200, 202, e.g., stimulus and measurement instruments 212A-F, 216A-F, may be prime candidates implementing as PLDs or RLDs, since these components are most likely to require customization or modification when adapting the integrated testing system from one type of testing and/or tested device to another. This programming may be done via any programming scheme, e.g., via the opcode/micro-code scheme discussed above. This programming scheme may be implemented via ATE communications link 228, e.g., utilizing serial ATE data vector channels or via user-interface communications link 232. In the case of an opcode programming scheme, the corresponding program would include the appropriate opcodes for communicating with and/or controlling microcontroller 224, identifying the appropriate PLD(s) and/or RLD(s) 220, 820 and providing the microcontroller with the appropriate data for (re)configuring each PLD or RLD. Given a particular type of PLD or RLD, those skilled in the art will readily understand how to (re)configure PLD or RLD 220, 820 via processor 804 and microcontroller 224, such that further explanation is not necessary herein.
As mentioned above, in one embodiment integrated testing system 108 may be set up and controlled using serial channels of ATE 116 typically used for downloading data vectors to DUT 104. Referring to
Serial communications link 900 may utilize any number of channels compatible with the corresponding ATE. For example, communications link may include six channels denoted REQ (request), ACK (acknowledge), MODE (nature of the request), S_CLK (communications clock), S_TX (transmit channel) and S_RX (receive channel). The MODE channel is used by the master to signal to the slave what it is that is being requested (e.g., a file upload, download, test execution, test abort, etc.). The S_CLK is driven by the master in order to stream data out on the S_TX channel or to stream data in on the S_RX line.
An important goal of utilizing serial communications link 900, when provided, is to provide a means for implementing high-level, minimally interactive control of integrated testing system 108 using ATE 116. Thus, although communications link 900 may be relatively slow, the fact that interaction between ATE 116 and integrated testing system 108 is minimal, the slow speed is essentially of no consequence, since most, or all, of the analysis of measurement data may be performed by compute engine 208 aboard the integrated testing system. For example, if desired, ATE 116 may be used to set up a sequence of tests and to pass test parameters to integrated testing system 108 only once in order for the integrated testing system to test multiple DUTs 104 of the same type. In addition, it may not be necessary to upload test results from integrated testing system 108 to ATE 116, unless a user would desire this. Instead, result comparison may be performed locally within integrated testing system 108.
In one embodiment, a testing process 1000 may proceed as shown in
Depending upon the nature of the computations performed, one or more out-of-limit responses of DUT 104 may almost immediately allow integrated testing system 108 to determine that the DUT fails the test. In this case, at step 1050 integrated testing system 108 may discontinue testing and notify ATE 116 that DUT 104 should be rejected. At step 1050, ATE 116 may also move the rejected device to a “reject” bin (not shown). However, if integrated testing system 108 determines DUT 104 passes the test, at step 1060 it may notify ATE 116 of this so that the ATE may move the passing part to a “pass” bin (not shown). In an alternative embodiment (not shown), ATE 116 may make the pass/fail determination for a particular DUT 104, rather than integrated testing system 108. In this case, measurement data may be read back to ATE 116 via ATE communications interface 228, e.g., serial communications link 900 of
At optional step 1070, integrated testing system may determine if all DUTs in the group to be tested have been tested, e.g., by comparing a count of DUTs tested to a previously input total of DUTs to be tested during this cycle. If not, process 1000 may loop back to step 1020, at which point another DUT 104 may be loaded into DUT socket 120 and tested. If step 1070 determined that all DUTs to be tested have been tested, process 1000 may end at step 1080. Alternatively, testing process 1000 may be globally interrupted by ATE 116. If another one or more of the same type of DUT are to be tested after process 1000 has stopped or been stopped, testing may be resumed with or without repeating step 1010. However, if another type of test is to be performed and/or another type of DUT is to be tested, step 1010 will have to be repeated using one or more new setup/testing programs suited to the particular test and/or device type. Of course, as those skilled in the art will readily understand, any of a number of testing processes may be implemented depending upon the design or application of integrated testing system 108. In addition, although testing has been illustrated relative to implementation via ATE 116, testing may alternatively be implemented via user interface 140 in a similar or different manner.
Referring to
Although the invention has been described and illustrated with respect to exemplary embodiments thereof, it should be understood by those skilled in the art that the foregoing and various other changes, omissions and additions may be made therein and thereto, without parting from the spirit and scope of the present invention.