News

DPMO is the superior means for determining equipment performance.

As the electronics manufacturing market continues its migration to finer pitched devices with 0201s and CSPs now considered mainstream technology, the accuracy and repeatability of modern production equipment is essential to a robust process. This is true of most equipment used for SMT and packaging applications, but is absolutely critical for screen printing equipment. As I have said several times in this column, nearly 60% of defects can be traced back to errors in the printing process. So, one can conclude that if the printing process is spot-on, the defect rate will be reduced dramatically.

Historically, a screen printer's accuracy and repeatability have been measured using statistical process control (SPC) tools, confirming a Sigma or CpK rating. Six-Sigma is the repeatability rating that everyone is after and even that can be confusing since equipment manufacturers do not all abide by the same upper and lower limits of variability (i.e., the tighter the limits, the more difficult it is to achieve Six-Sigma status). That said, it is generally accepted that Six-Sigma @ ± 25 µm ensures that a machine has a high probability (Six-Sigma) that every time it repeats, it will be within 25 µm of the ideal position.¹ And, while this is, to date, the best measure we have had for confirming accuracy and repeatability, there are a couple of cautions to using this rating as the gold standard for equipment qualification: 1) as stated above, not all equipment manufacturers adhere to tight limits of variability, and 2) most screen printer manufacturers confirm this rating for dry alignment only and not a full, wet print cycle.

But, what if we took those standards and raised them a level? What if a new benchmark of equipment qualification could actually guarantee a printer's true process capabilities before installation?

What I am proposing is a measure of performance that has been used in other industries (namely, automotive) and, to some degree, in electronics assembly: DPMO. Simply put, defects per million opportunities is the number of defects divided by the total number of opportunities for a defect, multiplied by 1,000,000. This measure can be extremely valuable, as it permits the prediction of yield at the product design stage - before the equipment is commissioned. And, because each product, each process and each manufacturing environment is unique, DPMO can accurately predict how the screen printing equipment is going to perform for a given process, based on these manufacturing variables. By all accounts, it is an invaluable tool to a manufacturer.

In our industry, placement companies use DPMO as a measure of placement accuracy. It is widely accepted, however, that placement does not contain as many variables as printing and, therefore, DPMO for placement is a fairly straightforward calculation. When using DPMO to evaluate screen printers, the influence of such variables as aperture size, aperture geometry, paste volume, different paste rheologies, squeegee pressure, squeegee type (blade or enclosed head), print speed, separation speed and stencil material, to name a few, can be reduced by using tests based on a on a standard internal procedure to ensure the equipment is measured against a known good process.

DEK firmly believes that DPMO would be a worthwhile standard by which to evaluate screen printing equipment and would give customers a better indicator of the long-term capability of the equipment and their process.

In fact, because we think this is such a viable measure for printer performance, DEK has begun work with DPMO analysis. Here is our approach: Each component type has a pass/fail volumetric criteria set against it. To ensure that the process is capable, all aperture geometries comply with the 0.66 aspect-ratio rule. Each aperture is defined as an opportunity and then measured individually to determine DPMO. Ensuring that the printer has a predetermined process DPMO figure enables the assembly facility to predict and control yield rates.

Suffice it to say, however, that just as in the case of Six-Sigma qualification, DPMO levels could be easily manipulated using very wide tolerances. So, make sure that you understand to what tolerances the testing was performed. Let's face it, a paste volume of 25% is much easier to achieve than 75%, and accurate prints for very fine-pitch part sizes are more difficult to achieve than those for larger pitches. But, if an equipment manufacturer is adhering to extremely tight tolerances and testing against today's mainstream processes, a DPMO rating could be incredibly useful.

Adapt or die. That modus operandi has served our company and the industry well. Manufacturing improvements and technical advances do not happen by sitting still. Perhaps it is time to take a new view on screen printing equipment capabilities assessments. What do you think?

References

1. Bruce Brigham, "Demystifying Six Sigma," Circuits Assembly, February 2005.

Clive Ashmore is global applied process engineering manager at DEK (dek.com). His column appears semimonthly.