Reply to thread

Message: <blockquote data-quote="John Roberts" data-source="post: 74072" data-attributes="member: 126">Re: Faulty channel strip #5Sorry the calculus on reliability engineering is not roughly that simple. How those exact same parts are applied can matter a bunch. This is an old theme for me to explain. "Quality" is often confused with presence or absence of higher cost features, so while there is a correlation with cost and feature content, quality technically has a subtly different definition, pretty much how well does a product meet expectations for it, or do what it is supposed to do. While the popular definitions of quality includes excellence or exceeding expectations, I find that several product characteristics like ruggedness or overhead (extra capacity) are actually costly features expected by professional customers for professional application (and paid for), so not quality per se, unless explicitly promised and lacking.   Quality control or quality assurance functions in professional organizations, pretty much monitor internal processes to confirm that products get built according to the design, and perform to design targets. This quality assurance can extend all the way to monitoring web forums to identify missed or incidental problems (problems associated with shipping half-way around the globe are not exactly new either.). Cost and quality has an interesting relationship. It costs more to build a product wrong and repair/rework the mistakes, than build it correctly the first time. So coincidentally the cheapest manufacturing process involves building it right once (pretty obvious huh?). This is not a new concept either and techniques like SPC (statistical process control) not only test products for pass/fail, but measure process variable to confirm that test units not only pass, but are exactly in the middle (or better than) process target range. If the process shows any deviation from optimal, adjustments can be made to restore the process to dead center of the target, before ever making a single product that would fail to function or not work.  SPC, when a company can use this approach effectively can be a lot cheaper than 100% barrier testing and rework. The quality of cheap mass produced consumer products like razor blades can be very high... when was the last time you got a faulty razor blade? Of course we are talking about far more complex SKUs here, with far more opportunity for faulty components and/or humans to make mistakes, but the concepts are the same and many component vendors deliver low PPM reject level quality (I recall better than 1PPM from some good vendors on very high volume components, while even great vendors drop the ball sometimes). JRPS: Actually razor blades are not very cheap, for just a few cents worth of plastic and steel, but hopefully you get my point.</blockquote>

Verification

Top Bottom