Testing Environment
The system can be setup to accommodate a variety of philosophies about how to present data to the operator.

In the most simple case, the operator places the UUT into the fixture, then uses the Start and Safety switches on the front panel to start the test. The fixture actuates, and the test begins. Once the test is completed, the screen shows a large red FAIL or green PASS indication, and the fixture is de-actuated. Testing status is shown on the CRT, along with red (fail), green (pass) and amber (busy) lights on the stack-light and front panel.

The operator can then choose to ReTest, or move to the next UUT. Most users configure the system to automatically print out a test report of component failures on the system printer if the UUT fails. The operator then attaches the failure report to the bad UUT, and sends it off for repair. The next UUT is then put into the fixture and the process starts over again. This simple operation cycle is easy to use by unskilled operators. Paperless repair is also possible including built-in serial number tracking.

The system can be set up to halt on each failure if you would like your operators to be able to repeat steps, or repair the UUT as faults occur.
You can also view a real-time Pareto report of failures during each batch of UUTs. By observing this sorted table of specific failures, you can quickly detect repetitive process faults.

Test reports can be automatically generated in a variety of configurations, or can be manually selected by the operator.

Panelized UUTs are accommodated during testing operations. As the test is performed; you can observe a graphical status representation of the panel as each UUT in the panel is tested. At the end of the test, each UUT in the panel is shown as a pass/fail/skip, and result reports are separated by UUT.

While there is a wide variety of capabilities for the operator, you can use the system’s login capability to tailor the resources available to each user. Not only does this provide ease of use based on operator skill level; it can provide integrity to test programs and the system configuration so they cannot be modified by unauthorized personnel.

The system can log serial numbers of assemblies, either through manual entry, or via an optional bar-code reader.

Statistical Process Control
The system can be set up so that it logs statistical data. When this is enabled; you can obtain several types of reports. Reports can be limited by beginning date and time and ending date and time. You can also report on all UUTs, or choose particular ones to analyze.

The Production report lists which UUTs have been tested, the failure rate and how many defects have occurred. This is valuable to determine overall production, production by shift, production by UUT, and production failure rates.

The Pareto report lists the faults sorted by occurrence. For example, you might find that your greatest source of error on one assembly is R101. You might find that frequently the roll of parts used to feed R101 is incorrect, or perhaps the part is difficult to install, or a particular shift is having more trouble than others.

The X-Bar/Sigma reports are used to show, by individual analog measurement, the mean (average), standard deviation, 3-sigma limits, Cp and Cpk. This data is graphically displayed with a predicted distribution curve and high/low test limits.

While this information can be used to monitor process measurements, it is more often used to help fine-tune test program tolerances. By observing the data, even with a relatively small programming sample size, it is practical to set control limits that are applicable to a wider range of UUTs.

Raw statistics data is logged on the disk in ASCII format, comma-delimited, so that you can write your own custom analysis software if desired.

Test Program Generation
The system includes all the software necessary to write and modify your test programs. Many users have CheckSum build test fixtures and write test programs; however, many users do these functions in-house or use local contractors to help in this effort.

Test programs are generated in an interactive spreadsheet-like environment, with each line specifying one test step. Typical test steps include RES (resistance test), CAP (capacitance test), CONT(inuity test), JMPx (jump based on some conditional), MEMx (memory math), RELAY (specify relay closure), DIGO (digital output). The line contains other information relevant to the step. For example, a RESistance test step would include two test point names and numbers, a measurement range, nominal (expected) value, and high and low test limits.

The test programming language is rich in features. In addition to normal measurement and stimulus test types, features include mathematics functions, file I/O, jump based on measurements, math, or operator input, display of messages, operator input, interactive adjustment routines, calling of external programs that you have written, and a host of other capabilities to make programming easy and flexible.

Each test program can have up to 30,000 test steps, and test programs can be transparently linked together to provide unlimited length programs, or to allow you to make libraries of program segments that you can reuse.

If you have CAD data for your assemblies, it can be used to generate the preliminary test program and a wiring report. The system accepts ASCII net list and component information from many popular CAD formats including OrCAD, P-Cad, Mentor, HP-BCF, Cadence, Racal-Redac, Viewlogic, Tango, ComputerVision, Pads2000, Scicards and Fabmaster. Even if you are using another CAD package, it may be able to generate output for one of the supported formats. The automatically generated program contains test steps for the components in the net list, and a wiring report. Once the fixture is built and wired, you can load the generated program, then fine-tune test steps as necessary. Typically, about 70% of the generated steps initially pass. Once the program is interactively optimized with appropriate ranges, polarities and guard points, you can self-learn CONTinuity and ICs data, and the power-down test is ready to use. Functional programming is hand-entered to meet the specific needs of the UUT.

Entry or generation of programs can be done off-line in your office. Optimizing the program is done on the test station and involves choosing the best methods of making measurements. While in the test program editor, you can execute tests. If you are not satisfied with the result, you can enter a menu that displays the measurement taken with a variety of techniques and ranges. You can quickly choose the best technique/range, add guard points, change polarity or add delays to obtain the best test results. Other tools include X-Bar/Sigma, measurement time, and time/voltage displays for each basic measurement type.

Panelization facilities include a step-and-repeat mode. This allows you to specify the panelized format and the initial wiring points for each PCB in the panel. Once you have written and debugged the first UUT in the panel, the system will then automatically generate the steps for the other PCBs. At run-time, the operator can elect to skip PCBs in the panel that are not populated or known-defective.