What happens when new silicon comes out of fabrication? Do the designers go to the lab and start evaluating the new silicon, or do they wait until the product engineer or lab tech takes the data and presents the results? My hope is that the first situation is most common. However, I do realize that many engineers are busy and sometimes rely on the second option to get the data off the silicon. I think this is a mistake.
Valuable insight is gained by going into the lab when the first silicon comes out to evaluate the results. If problems occur, the designer of the blocks is the best person to assess and draw conclusions on what might be happening with the circuits that were designed. In addition, by looking at the first silicon results in the lab, the designer can establish the correct procedure for ensuring the part is tested correctly. The problem with turning the first evaluation over to a tech or other engineers not familiar with the designs is that the results may not reflect the intended results, because of setup issues.
Furthermore, having the designer of the various blocks in the lab capturing that first silicon data helps solidify in the designer's mind that the circuits work as planned. If the circuits do not work, the designer can start the debugging process early on and gain valuable insight into what may be the problem. The development of skills offered by hands-on evaluation of the various circuits and interactions with the package and board is gained only by taking the time in the lab evaluating what you developed.
The idea of having designers in the lab was commonplace years ago where the development of the circuits was at the board level with individual passive and active devices laced out on a circuit board. (It still is commonplace today for high-frequency RF circuits.) For these type of designs, a designer could easily see what was being designed and could directly test the various nodes of the circuits being designed, provided the circuits could tolerate probing. However, improvements in technology have moved many of the functions once designed on the final board into the integration environment. This means that testing or at least having access to test every node is becoming more difficult.
As a result, I believe, designers are losing the feel for the interactions that occur on the boards used in the final product. I believe the designer is losing the ability to understand how the board interactions affect the circuits being designed. In addition, the designer who does not go to the lab to learn these valuable lessons loses the ability to design circuits that are robust enough to interact with the board and may create inferior products.
I am not saying that the designer who goes into the lab becomes a superior designer to one who never goes into the lab. However, I am saying that the experience gained from lab evaluations and understanding how circuits interact with the board is invaluable. The knowledge gained from board and outside environment interactions will help the designer understand how to make circuits work. In addition, the experience gained by looking at circuits to understand why they do or do not work on the evaluation and/or production board is important in becoming a much better designer. As I have stated, I believe the lab experience cannot be replaced and is extremely valuable to the engineer's growth in a field of expertise. Futhermore, let me say from experience that the lab can teach you things about circuit design you cannot gain anywhere else.
Do you agree with these statements? Do the engineers you work with go into the lab to evaluate first-pass silicon results?