Introducing SPC Analysis to the Reflow Process

By Dennis Ishler and Robert Sheahan

NEW DEVELOPMENTS IN MONITORING HARDWARE AND STATISTICAL PROCESS CONTROL SOFTWARE NOW PERMIT AUTOMATIC ASSEMBLY OF DATA ON THE THERMAL PORTIONS OF THE SMT PROCESS.

Statistical process control (SPC) is like the warning squeal of automobile brakes when the brake pad is almost worn out. The warning gives the driver time to get to a garage and avoid costly accidents. Similarly, with electronics production, even in a "perfect" process, machinery is subject to normal wear and failures that may not be immediately apparent. SPC can spot a process drift too subtle for direct observation, giving operators time to repair the process before defective parts are produced. For example, with continuous thermal profiling and SPC, a soldering failure can be detected within 30 sec. Without these tools, the failure might not be detected until the next scheduled profiling or when the electrical test reports a high percentage of failed boards, resulting in hours or days worth of bad parts.

SPC is a method that utilizes a set of data-based tools to optimize and maintain quality of manufacturing procedures. The key to its understanding is predictability: If a process produces good parts, deviations from predictions can be read as indicators that the process is changing and may soon produce bad parts. The first step in implementing an SPC program is to find the critical process measurement(s) and to gather data from a selected sample of production. Once a set number of samples are assembled, the data are plotted on one or several SPC charts (generally X-bar) and then on a frequency histogram.

'Assignable' Causes

Figure 1. Frequency histograms are compiled after critical process measurements of production samples are taken to determine if the process is deviating.

A process that is "in control" will fall into a normal frequency distribution resembling a bell curve when displayed on a histogram (Figure 1). If the data do not fall into such a distribution, there is an "assignable cause" in the process, i.e., a new feature prompting erratic behavior as reflected by the "non-normal" curve. Examples of assignable causes include consistently faulty material, a machine incapable of meeting process specifications or an operator using a machine incorrectly. The assignable cause in Figure 1 is a faulty thermostat.

After corrections are implemented, statistics can be used to develop an accurate overview of the process. Process-capability (Pareto) charts can show those features that require more urgent attention in the order of most-to-fewest defects. For example, correlation and regression analysis can answer such questions as whether minimum conductivity is more closely linked to peak or soak temperatures by showing how well different zones of the furnace maintain their desired thermal levels. Analyzing a process in this way concentrates process-improvement efforts on essential areas.

The Cpk charts in Figure 2 indicate that while thermocouple (t/c) eight is demonstrating a fluctuation well above the nominal, that at t/c 25 is well below nominal with very few parts outside the specification limits. Probable assignable causes: The oven zone t/c eight is monitoring may be draftier than that at t/c 25, or a mechanical problem may exist with the convection fans.

Figure 2. Process-capability charts can show how different zones of a reflow oven are maintaining their prescribed thermal levels. Note that t/c eight shows fluctuations well above the nominal while t/c 25 reveals a reading well below.

The Reflow Process

A process for connecting electrical components and assemblies to PCB's, the purpose of the reflow oven is to heat the product to a specified temperature for a precise period of time. When this combination is plotted, the result is a thermal profile whose maintenance is critical: If the PCB is not heated properly, the solder will not bond the components to the pads correctly; too hot, or heated too quickly, and the board and parts may be damaged.

History. Until recently, solder reflow has been the most difficult portion of SMT production to monitor for quality control. For many years, the conventional method of monitoring conveyorized thermal processes had been to attach thermocouples to a product. Using a wireless profiler, product and profiler would be run through the oven together to record the thermal pattern. Typically, this activity would be performed on a regular basis -– monthly, weekly or as often as once a shift -– to verify that solder joints of adequate quality would result. Oven capability must be verified regularly because it has been shown that even with modern forced-air convection units, the thermal profile can drift beyond acceptable specification despite a controller indication that nothing has changed. Also, product-profile verification was necessary when problems arose elsewhere on the production line. The guiding principle: With a decrease in yield, the oven must be profiled to determine the source of the problem.

Continuous Thermal Monitoring

Three problems occur with the conventional method of profiling:

A product profiler is time consuming and often results in production downtime.
Each profile run is the equivalent of a "snapshot" taken with a still camera. The effect: The oven user must assume that the oven is constant between snapshots, which will rarely reveal intermittent problems.
If regular profiling uncovers a problem. It is generally impossible to tell precisely how much product has been affected. All production between the first bad profile and the last good profile becomes suspect.

The solution to the failure of pass-through systems to provide a continuous product profile (plus their disruption of production) is a system that is roughly the equivalent of a video camera constantly running in the oven. In its standard form, the system consists of two 0.250" diameter stainless-steel probes, each containing 15 internal thermocouples mounted along the conveyor in proximity to the product (Figure 3). A thermocouple processing unit (TPU) that sends probe data to a computer and a self-contained Windows-based software package completes the system. The key difference between the package and conventional product profilers is that 30 thermocouples inside the probes continuously monitor process temperatures.

Temperatures at the conveyor are continuously displayed as "oven profiles" on a computer screen, and data are permanently recorded on the hard disk (Figure 4). During production, any temperature drift and its location are instantly visible to the user. The t/c probes are outside of the oven-control loop, which enables them to reveal critical temperature deviations often hidden from the thermocouples at the conveyor. The oven profile is updated every 30 sec and is permanently recorded in a history file on the computer hard drive.

SPC for Reflow

The oven-profile data from the thermal manager is available to any software application via live data output from the system, providing real-time data on the thermal process that can be used in a number of ways (e.g., SPC).

The thermal-monitoring output interfaces directly with an SPC software package that automated data collection, management and analysis. The package can also monitor stencil printers, solder-paste-inspection machines and placement equipment so that SPC is applicable to the entire production line.

The thermal manager can collect and send the following data to the software package: raw data from the 30 t/cs in the probes, the t/cs' maximum deviations and the belt speed. Up to 50 control charts can be tracked simultaneously by the software package on the computer screen. Chart backgrounds change color according to process conditions: green for normal, gray for warning and red for alarm. This combination of data options permits custom tailoring of the PSC application. Some manufacturers have chosen to monitor a single thermocouple at a critical point in the oven profile while others have identified multiple critical process parameters (Figure 5).

Traditionally, SPC has required the collection of 25 sets of data before points can be plotted on a chart and analysis can begin. Data collection has always been the most daunting task associated with implementing SPC because of the time and labor required, and also because manually collecting the data can cause an excessive amount of time to pass before charting can begin. Another issue in implementing an SPC program is the questions that sampling only a small percentage of production raise. All three of these issues have been resolved. The thermal manager outputs data every 30 sec, allowing adequate data for an initial charting can be gathered in under an hour. The data is gathered automatically, removing labor and cost issues, and because the data is continuously collected in real-time, sampling issues are minimized.

A fully implemented SPC program is used in two ways: for real-time process control by machine operators on the production floor and for historical analysis by statisticians. SPC permits machine operators to track process status, to spot problems to small to be noticed and to correct them before they cause a defect of affect yield. Data can also be sent to standard industry databases such as Excel or Access for analysis by statisticians at a later date for long-term process improvement.

Benefits of Applying SPC

Advances in both hardware and software have made efficient and cost-effective methods for applying continuous real-time SPC available to manufacturers. The task of implementing a SPC program for the reflow process has been simplified and now fills a gap in the application of SPC to the SMT production. Users will be quick to recognize the benefits of these powerful production tools including:

Real-time data collection, which acquires adequate samples for analysis in hours without disrupting production.
Rather than relying on time-consuming data accumulation or on a single statistic, real-time data collection includes several critical statistics as well as conveyor belt speed.
Real-time data are collected automatically, providing significant increases in accuracy for analysis at any time.
Real-time SPC simplifies the identification of long-term trends in the production process and for decreasing defect rates

Conclusion

Without real-time SPC of the solder-reflow process, a defect or failure is a manufacturer's first indication that something is wrong in the thermal process and that decisions for correcting the problem are, at best, guesses. Continuous data collection the other hand, ensures that a drift in the thermal process cannot go undetected. Users can prevent defects caused by failures in thermal processes before they occur. The advantages of real-time SPC for the reflow process are thus clear, and advances in both hardware and software technology have given manufacturers the tools to benefit from them.

# # #

DENNIS ISHLER is R&D engineer at KIC THERMAL PROFILING, 15950 Bernardo Center Drive, Suite E, San Diego, CA 92127; (858) 673-6050; Fax: (858) 673-0085.

ROBERT SHEAHAN is senior software engineer and chief statistician at Prolink Software, 148 Eastern Blvd., Glastonbury, Conn. 06033; (860) 659-5928; Fax: (860) 633-7309.

Company || PWI || Products || Ordering || Press || Support || Download || Library
Home || Contact || Links || Search