An Ounce of Review, A Ton of Cure

The job definition varies from company to company, but every component engineer (CE) is expected to be proficient in some standard disciplines. The broader the scope of experience and more thorough the knowledge of electronics and component behavior and characteristics in particular, the more useful the CE becomes to an engineering operation.

Every engineering operation will follow some formal or informal procedures when producing and introducing a product. One common element of these procedures is a design review process or series. A design review serves several purposes, the chief one being verifying that the design is moving toward its stated goal and keeping to its schedule. Review meetings may limited to engineers only, or they may involve invitees from other departments. At a minimum, the design, systems, project, CAD, and component engineers should be part of every design review. This article discusses the CE's involvement in the design review meeting.

The one essential visual aid that everyone will expect to see at the meeting is a schematic diagram of the circuit being designed. The host engineer will usually take the team through a block diagram version of the circuit describing the operational theory and features. The schematic will then be discussed one page at a time with an eye toward identifying potential problems or oversights. The CE will examine the schematic for new parts, specialty parts, and even parts that should not be used because of factory-related problems or pending obsolescence.

The CE will also look for parts that may require heat sinks. Interconnects are also of particular interest, because mating contacts between boards or from cable to board need to use the same metals as the contact plating. The overall power budget is also an issue. Environmental compliance at the part and system level should also be a consideration for design approval. If the CE can identify legacy parts that can be replaced by higher levels of integration, then the circuit's reliability may be increased by having fewer points of potential failure.

If the CE is not in attendance, a host of unanticipated problems may arise that may necessitate a respin of the design, the printed circuit card, the interfaces between cards, and the development schedule.

Let me cite one example. A redesign requirement to reduce power and board size was commissioned to an outside consultant. After three months, the consultant brought in a prototype for a first look and design review. The company was thrilled with the results. Entire chip sets were replaced with single, multi-pin, multi-module circuit packages. The power requirement was cut in half, and the board size was reduced by more than half. I was called in to look at the prototype and do a preliminary cost analysis.

This was the first time I was involved in the project. I looked at the board and immediately saw something that required a response. The large multi-module chips on the board had part numbers with an “-ES” suffix. I knew that suffix meant “engineering samples.” So I asked, “Are these parts in full production yet?” Blank stares all around the table. I left the meeting and called the manufacturer's rep. While on hold, I pulled up the part's specification PDF. Right across the page, in big red letters, it said “PRELIMINARY.” When the rep came to the phone, I asked if the part was available. The answer was one I did not want to hear: “We will be in production in the second quarter of next year.”

The rep offered to get me 10 more engineering samples if I was in trouble, but that was all the manufacturer could spare. I returned to the meeting with this bad news, and then came the scramble and the finger pointing. This stealth development by a highly qualified and skilled consulting engineer had backed the entire project into a corner.

One of the best-practices during a design cycle is to identify all the essential people who need to be in the design review. More importantly, if the design is being done in a corner or in stealth mode, get at least one component engineer reviewer involved as early as possible. In this case, an ounce of prevention could have been worth a ton of cure.

1 comment on “An Ounce of Review, A Ton of Cure

    December 19, 2012

    Any design engineer will tell you that time spent in review is paid back ten fold (or more).  There is an inverted pyramid which clearly shows that more time is required to fix a problem later in the development cycle. 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.