Troubleshooting Tips

Jerome L. Hartke, President, Media Sciences, Inc.

Published in one to one, September 2006

Contents of this Document
Why Is the Disc Bad: How to approach interchange problems.
Types of Defects: Flaws responsible for read failures.
Testing for Defects: Electrical, logical, visual, and mechanical tests.
Additional Tests: Special tests for recordable and rewritable discs.
Recording Drive Flaws: Writers can produce bad discs.
Data Analysis: Only careful analysis can detect multiple flaws.
DOs and DON'Ts: Procedures to follow and to avoid.
Corrective Action: The right way and the wrong way.
Return to Home Page




The Good, the Bad, and the Ugly




Why Is the Disc Bad?

User complaints are often just the tip of the iceberg. Frantic firefighting methods often make problems worse. Only a disciplined approach, followed by a solid fire prevention program, can generate low cost, high quality products and confidence in predictable media interchange. Reliable data from an experienced manufacturer or independent testing laboratory may be essential to good decisions and effective corrective action. Bad data or speculation invariably leads to erroneous conclusions and misdirected actions that can degrade quality.

Interchange problems can be caused by a single flaw, but usually result from multiple quality deficiencies. A clear picture based on comprehensive test data is required if problems are to be pinpointed, prioritized, and corrected. Follow-up tests must then be conducted in order to confirm both the diagnosis and also effectiveness of corrective actions.

Problems often surface when a disc is readable sometimes but not every time. This common symptom often occurs when discs are marginal, therefore tests should be conducted on both “bad” and “good” discs to determine if they are similar or very different.


Return to Top of Page

Types of Defects

Read drives can have many types of defects, but the absence of public read drive standards may cause good discs to be blamed when interchange fails. Nominal pit geometries and low error rates predict successful interchange, even with marginal read drives. Regularly scheduled, comprehensive tests are required in order to achieve this goal.

Although CD BLER or DVD Inner Parity Error values are frequently used to characterize quality, they rarely expose the true problem. This is because they only identify total errors, most of which are easily correctable. Important electrical tests identify both uncorrectable and significant but correctable errors, and also evaluate disc parameters that affect read drive servos.

Marginal or unacceptable electrical parameters may generate high error rates and read failures in some drives but not in others. Logical tests often identify recording software problems. Visual examination may reveal manufacturing or handling flaws. Mechanical tests can uncover a host of issues such as unbalance, tilt, and eccentricity. Few discs pass every test, therefore a complete picture of media quality is essential to effective troubleshooting.


Return to Top of Page

Testing for Defects

No single test, or even a “top ten”, can be relied on to identify the root causes of interchange failure. Troubleshooting is similar to a detective patiently following a long thread of clues, not a snapshot pass/fail determination of manufactured part quality

Electrical parameter measurements employing signals internal to a test drive indirectly evaluate pit width, length, depth, and sidewall slope. No one dimension is measured by a single parameter, thus nominal pit geometries require that all parameters be well-centered in order to achieve predictable, successful interchange.

Electrical error tests also use internal test drive signals to evaluate both the number and severity of physical defects. Successful error correction depends on small tangential defect lengths, therefore DVD Inner Parity Fails (PIF) and Outer Parity Fails (POF), or CD E22 and E32 errors, are much more significant than DVD Inner Parity Error (PIE) or CD BLER error rates. Soft, non-reproducible errors related to servo disturbances may not occur in high quality test drives, therefore low error rates alone do not predict successful interchange.

Logical volume and file structures are defined in ISO 9660 and UDF standards, and are controlled both by recording software and by operator selections. These standards are frequently violated, but forgiving operating systems nonetheless mount the volume, masking logical defects that may prevent interchange in less tolerant receiving systems.

Visual inspection is one of the most important tests. Careful examination can disclose handling problems that limit interchange and longevity, as well as marginal defects that give permissible error rates in a high quality test drive but unacceptable results in lesser quality drives. Defects outside of the information area are not detectable by electrical tests, but indicate latent problems that will cause problems when the information area expands.

Mechanical tests often disclose flaws that do not degrade performance in high quality test drives but can cause read failure in other drives. Center hole and rim dimensions, thickness, track pitch, eccentricity, axial deviation, and angular deviation (tilt) mechanical measurements are required to assure conformance to standards and specifications.


Return to Top of Page

Additional Tests

Recordable and rewritable discs present special challenges. Unrecorded tests are necessary to assure conformance to pregroove and unrecorded addressing requirements. Dye or alloy layers must be tested by first writing the discs in high quality writers and in a clean environment. Surface damage or contaminants doubly degrade quality; first when the disc is recorded, and again when it is read.


Return to Top of Page

Recording Drive Flaws

Writers may be responsible for bad discs. A drive may have degraded, might have been bad initially, or may have been used at the wrong recording speed. Write a few prequalified, high quality discs to full capacity, and then thoroughly test them to detect write drive problems. Regular incoming qualification and regular monitoring of writers can prevent many problems.


Return to Top of Page

Data Analysis

Because most failures are caused by multiple problems, do not focus on the first detected flaw. A Media Sciences study of 100 failed CD-R discs disclosed that 78% had multiple flaws. Develop a comprehensive data set including both numerical and graphical test results. Look for both marginal and out-of-spec values, as well as unusual radial variations. Only then implement corrective action, and conduct follow-up tests to confirm that they have been successful.

Ongoing process control is necessary to establish and maintain a high level of confidence in your product. Regularly sample and test to build a baseline and to detect departures from the norm before they cause interchange failures. Confirm that the instruments that produce your test data are properly calibrated and correlate with the appropriate laboratory. Bad data leads to bad decisions. Good decisions require good data.


Return to Top of Page
DOs and DON'ts
  1. DO recognize that good management is an essential part of effective troubleshooting.
    DON’T allow poor management to become part of the problem.

  2. DO realize that problems are opportunities to improve quality, and communicate that philosophy to all.
    DON’T deny that problems exist, arbitrarily assign blame (finger point), or demand a fast fix.

  3. DO expect that problem identification and correction will be time consuming and will require significant support.
    DON’T establish a panic culture that deals only with symptoms instead of the real problem.

  4. DO evaluate all sources of defects, beginning with the original image and including all following steps.
    DON’T begin with the assumption that a specific element of the manufacturing process is responsible.

  5. DO assure that test equipment is accurate, and that results correlate with independent labs.
    DON’T assume that test data is valid just because the equipment “works”.

  6. DO conduct comprehensive tests, objectively evaluate the data, and then make logical decisions.
    DON’T speculate or decide “by committee” with an outcome determined by the most forceful participant.

  7. DO understand how to interpret each test, and rank every result.
    DON’T rely on just pass/fail criteria, or assume that one test failure is the cause of the problem.

  8. DO expect multiple nonconforming results, and cull the significant from the trivial.
    DON’T focus on the first defect observed, or “the one that we saw last time”.

  9. DO prioritize quality issues and address them layer by layer, verifying each step before proceeding to the next.
    DON’T attempt to fix everything at once because adjustments frequently interact.

  10. DO use all available resources, including supplier and user feedback.
    DON’T degrade quality by relying only on the resident guru or the opinion of some industry “expert”.

  11. DO establish a corrective action plan that includes follow-up tests.
    DON’T expect that good intentions always give good results.

  12. DO cooperate with vendors who may be responsible for problems, giving them adequate time to fix them.
    DON’T impatiently swap vendors, resulting in a possible degradation of quality.

  13. DO communicate positive progress, both internally and to your customers.
    DON’T attempt to conceal quality issues; everyone has them.

  14. DO capitalize on troubleshooting outcomes to reduce costs and to enhance your image as a quality-conscious supplier.
    DON’T miss this opportunity to realize a significant return on your investment.

  15. DO reward all members of your team who contributed to successful resolution of the problem.
    DON’T miss this opportunity to strengthen your team in preparation for the inevitable, future problems.

  16. DO carefully evaluate lessons learned once pressure is relieved, and incorporate them into your business practices.
    DON’T condemn your organization to endless repetition of the same quality issues.

  17. DO expect both GOOD and BAD discs in the future, and be prepared to address issues as they occur.
    DON’T pass UGLY discs along to your helpless end users just because you believe that they “work”.

  18. DO expand this list based upon your experiences.
    DON’T expect the opinion of one author to fill all your needs.
Return to Top of Page



Corrective Action

A team approach involving all available internal and external resources is faster and more effective than speculation or reliance on an “expert”. Use accurate and objective data, as outlined above, to avoid denial and finger pointing between suppliers and departments. Follow the RIGHT WAY –– avoid the WRONG WAY. Quality is never static. It is either improving or degrading.


RIGHT WAY   WRONG WAY
     
Problems Are an Opportunity   Problems Create a Crisis
That Support   Resulting in
CONTINUOUS IMPROVEMENT   DENIAL
     
Team Approach   Assign Blame
     
Problems Are Corrected   Problems Accumulate
     
Team Is Strengthened   Team Is Weakened
     
Quality Improves   Quality Degrades


Return to Top of Page


Media Sciences, Inc. –– Dedicated to Quality