SPC controls the solder

A vendor of PCBs has used SPC software to identify where to make improvements in its manufacturing process.

Solder height on a Printed Circuit Board (PCB) is crucial to how well that board performs as it ultimately affects how well components sit on the finished board.

So when one manufacturer of PCBs realised that a proportion of its finished boards were failing final quality inspection due to inconsistent solder height, it turned to NWA Quality Analyst’s Statistical Process Control (SPC) software to help out.

SPC uses statistical methods such as control charts, capability analysis and exception reporting to monitor and control a process by enabling engineers to graphically analyse systems behaviour. This allows the impact of process improvement decisions, regulatory compliance and cost reduction to be gauged.

The variations within the process identifies where quality improvement could be made. SPC also shows how the process functions over time so it’s possible to monitor the effect of improvements and predict how the process will run in the future, based on how it ran in the past.

SPC helps to visualise whether the process is currently capable of producing output that meets or exceeds specifications (and customer expectations) 100 per cent of the time. SPC control charts identify data that falls outside defined limits, indicating variations in the process that indicate where quality standards are not being met, or where improvements can be made, so that appropriate action can be taken.

To investigate the irregularity in its process, the PCB manufacturer first sampled five units made at the same time and periodically measured them for solder height. The data was then analysed with Northwest Analytical’s (NWA) Quality Analyst using X-Bar/Range charting to determine whether the height variation indicates an out-of-control condition.

The quality control staff know that a certain amount of variation is natural, but an out-of-control condition could indicate a problem that, if left unresolved, could become very costly.

Data from a full day’s run was charted in an X-Bar/Range chart (Figure 1). This chart showed that the process was out of control in two places on the X-Bar chart, but in-control on the Range chart. This indicated that the process average was influenced by something on these occasions. 

The X-Bar/Range chart demonstrated how accurate and precise the process was in terms of: process accuracy, or how closely the data fell to the mean of all samples (X-Bar) as well as the process precision, or how much spread or variation there was between the five samples (Range chart).

The X-Bar chart in Figure 1 showed that the process was providing an average solder height in each sample that varied greatly from sample to sample. The chart also showed that two samples were out of control.

Figure 1: The process is out of control on the X-Bar chart, but in-control on the Range chart. This indicates that the process is being influenced by something at these points

This influence resulted in inadequate solder height. The problem might be with the raw material, with the machine applying the solder or even with the measuring device itself. Upon investigating the records, the process engineers found that the inadequate solder heights all came from batches using raw materials from a single vendor.

Further analysis of batches from that supplier revealed its inability to produce consistent material. As a result, the manufacturer dropped that supplier.

With the out of control condition removed, the process engineers analysed the process further. The process was run and charted for another day using only raw materials from reliable vendors. The process was now in-control (Figure 2) but suffered from excessive variation and was not capable of producing output well within specifications. (Note the Cpk of 1.05.)

Cpk is part of the capability index which measures how well the process is performing against specification – the higher the index the more capable the process. Most organisations look for an absolute minimum of Cpk equal to 1 and would be aiming for a Cpk over 1.33. So a Cpk of 1.05 indicates poor performance within the manufacturing process and means that there is scope for improvement.

Figure 2: After removing materials from an unreliable vendor, the process was in-control but suffered from excessive variation and was not capable of producing output well within specifications

Several SPC charts were used to view the data in different ways to establish possible causes for the poor performance of the manufacturing process. This resulted in the process engineering team establishing that the machine needed maintenance and recalibration to significantly improve the manufacturing process.
 
Once that had been achieved, the results from the next day’s production (Figure 3) indicated that the process was capable of producing output well within specifications. (Note the Cpk of 1.438).

Figure 3: After machine maintenance and recalibration, the process is in-control and capable of producing output well within specifications

The SPC software provided the PCB manufacturer with a tool that could be used to identify where quality improvements could be made to its process, as well as allowing its engineers to document how the improvements led to better quality.

But the flexibility of the SPC software, and its ability to maintain data links with corporate databases, LIMS and SAPS software, also allows engineers  to use it to measure and address numerous other issues in manufacturing. It could also be employed, for example,  to determine which component in a system causes the largest number of failures. Alternatively, the factors that affect the cost and profitability of a system could be analysed and prioritised.

For more information on NWA Quality Analyst visit http://www.adeptscience.co.uk/products/qands/analyst/