Yes, that is a pretty broad statement coming from someone who over the last several years has worked as an external consultant and designer-for-hire for a living. This article will address failure to carefully vet a potential manufacturing or “turnkey” partner and/or failure to transfer sufficient information and requirements to such a partner—a very common problem I have seen again and again with my clients over the years, and sometimes becoming the shoulder cried upon by several relatives and clients during that time.
I will start with a client for whom I designed a battery-powered circuit some years ago. The circuit included what was at that time a very special low-current-draw op amp used in an oscillator circuit to drive a sensor. We designed and built the circuit with a stateside assembly house to whom we shipped a parts kit and circuit boards. All units returned to us functioned as designed. The customer was happy with the performance, took our design documentation and files, transferred these to a low-bidder assembly shop (in China, but could really have been in several countries with low labor costs), and received back a pilot run of hundred-ish quantity.
Upon return, pilot boards failed to operate and we were tasked to figure out why. Troubleshooting located the op-amp circuit not oscillating. The relaxation oscillator design should never have a stable or damaging condition, regardless of power-up characteristics, as long as operating voltage is reached. Observed through a microscope, the SOT-323-5-packaged op amp looked kind of strange, and markings did not match the datasheet. We called Linear Technology, manufacturer of the op amp specified, and were told the markings did not match any of their parts. We searched the Web trying to find those markings and came up empty. Analysis was that an unauthorized substitution had been made to a similarly packaged, but not functionally equivalent, part.
What happened? A small assembly shop was given a deadline and likely was unable to obtain this special part in their distributor chain. In order to feed the workers, a same-package device was substituted on the boards to allow delivery of completed boards to ship to, and bill, the customer (our client).
What could have prevented this result? As always, interactions with partners or suppliers must be properly managed with business needs and capabilities in mind. To avert such a disaster, our client should have sent a knowledgeable representative (engineer or technician) to the assembly shop to personally hand over and explain the data package, make sure the shop understood what was required of them, and then remain on site overseeing the pilot run build. This representative would likely have observed the supply problems and discussed changes or substitutions with engineering, averting disaster prior to soldering wrong parts on the board.
Same Song, Different Verses
In another scenario, a design included a ceramic filter capacitor on the input to a linear regulator for high-frequency bypass. The circuit nominally was for a rectified 24 V ac at rated load transformer output, for which computations showed absolute worst case—all line and device tolerances would not exceed 43 Vdc. The circuit design called for a 50-Vdc-rated X7R capacitor at this location. Boards assembled in the USA were built and satisfactorily tested, and the first production was conducted as a small ten-ish board run overseas. Boards came back and were being programmed.
During program debugging, the capacitor on two boards exploded from the surface of the board, with two fragments ricocheting between circuit board and housing whilst still burning. It left gray/white residue on the green solder mask and gray/black residue on the cream-colored ABS housing. The fragments eventually burned to crystalline ash after numerous bounces, partially embedding the crystalline ash and small oxidized metallic pellet in the plastic housing at two locations. The overseas assembler claimed to have installed a direct substitute for the specified capacitor and sent photos of the reel of parts. Upon visiting the website of the manufacturer printed on the reel, when looking up the part, a pop-up window warned that parts should only be purchased direct, and that counterfeits had been experienced in supply-line purchases.
Attempts to replicate the problem over broad voltage and temperature ranges with known parts totally failed until a 6.3-V, 1-µF Y5V device was subjected to 128ºC and 50 Vdc. Here, the failure resulted in flames briefly and a large piece of the capacitor chip blew out of the body of the part. The conclusion was a counterfeit part had been installed, and budget did not allow more extensive investigation. Again, no representative followed the project for the pilot run.
In one other similar example, a company decided to move manufacture of a low-cost automotive electronic part overseas. Production of the new part eventually came on line after a few hiccups. A year or so later, a counterfeit part was found that duplicated the original design even to the extremely small and detailed company logo. Investigation found that the manufacturing plant had “leaked” the 3D model information to another facility that was manufacturing counterfeit parts. Most interesting was the quality of the counterfeits was almost the same as, and the appearance identical to, the company’s parts. In this case, a representative had been sent to the facility for pilot production, but not maintained on site long term.
I am positive that there are innumerable similar stories out there. Such problems have become so prevalent as to significantly impact our industry (electronics as a whole). Several classes of customers/clients have found it necessary to take on pretty severe internal and supplier initiatives and controls. It seems obvious that on-supplier-site monitoring and problem resolution is critical to avoid such issues, although this heavily impacts production-cost equations. Some companies have also started on-shoring previously off-shored products.
With these types of problems potentially affecting purchases, I now personally review country-of-origin on everything I buy and am very picky where possible. Where that is not possible, I often find myself deciding to choose the non-brand-name, non-brand-price alternative from the same country. This is born from disgust at the quality I have experienced when manufacturers move manufacturing overseas without sufficient supervision, a situation seemingly endemic to many product families. Cost-of-quality issues are at least as important as purchase cost for me. Am I that much different from your target customer?
What has your experience been?