From potential energy to value added by test (Guest commentary)
Dr. Fang Xu, Teradyne -- Test & Measurement World, 9/19/2007 11:23:00 AM
Recently, there have been many discussions in the press about how people have changed their perceptions of being an engineer. More and more young people are reluctant to consider engineering as their lifetime profession, and it seems unlikely that many among those who are currently studying engineering are willing to become future test engineers. Moreover, how many of our talented test-engineering colleagues have quit the test profession?
It is unfortunate that interest in engineering in general and test engineering in particular is diminishing, because test itself has become increasingly challenging and important. In the semiconductor industry, for instance, we predict that the cost of testing an equivalent transistor will eventually be higher than its manufacturing cost.
It is my contention that test is at least as important and valuable as any other stage in production. I employ the concept of generalized potential to demonstrate that using test to screen scrap from a production line to make the final product is not fundamentally different from extracting silicon from its dioxide in order to make wafers. They all elevate the potential and add value to the product in the same way.
We have learned in our physics class that if we elevate an object of a mass m from level 0 to a height h using a force of mg, we need to engage a certain amount of work:
where g is the standard gravity (approximately 9.8 m/s2 at the earth's surface).
If that object falls to level 0, it can produce work of equivalent energy value. So we say that when this object of a mass m is at height h from level 0, its potential energy is:
Then we learned that the weight of that object is due to universal attraction between the object and the earth. Surprisingly, the reference to define the zero of mutual potential energy of that object and the earth is when their distance is infinite. Therefore, the potential energy of that object at a distance r from the center of gravity of earth is:
where M is the mass of the earth; G is the gravitational constant, 6.6742 x 10-11m3s-2kg-1.
Indeed, this means any object on earth, including us humans, has a negative potential energy according to the definition in physics!
The more lessons learned, the more this statement rings true: almost every formula to calculate potential energy in physics starts with a negative sign; electrons around the nuclei; atoms in a crystal; planets around the sun; solar systems in the universe! As electronic engineers, we are all familiar with the electrostatic potential energy:
where q1 and q2 are electrical charge of two particles, and d is their distance. In the case of electrons around the nuclei, two charges have opposite polarity, so the potential energy will be negative.
But then is there something wrong with these negative signs? What will happen if we change all the negative signs in formulas of potential energies to positive?
Here, I imagine some people have goose bumps simply by reading the previous lines: If we change all the negative signs to positive, we will observe an explosion bigger than the big bang! It turns out that if anything could exist, then it must be stable, like our universe; for example, if there are positive charges, then there must be negative ones to balance them. A stable system must have these negative signs for their potential energy.
These negative signs in physics do not bother us in our daily life. Indeed, the history of technological advances of man is a history of our ability to use or manage energy potential differences. For instance, on a macroscopic level, we can use the energy of a waterfall to produce electricity; then we may consider using the electricity produced to elevate bricks a distance of a half kilometer in order to build a skyscraper. That same electricity could also be used to extract oxygen and transform silicon dioxide to mono crystalline, which has higher potential energy and high potential as a product. In reality, to achieve the chemical reaction:
The energy required is:
This represents the first major semiconductor manufacturing process among many others. In general, all major semiconductor manufacturing could be regrouped into two categories: They are either additive or subtractive. These processes include production of mono crystal, oxidization, photolithography, etching, diffusion, metallization, passivation, graining, and packaging.
On a microscopic level, semiconductor devices functions are based on manipulation of quantum potential energy levels to detect, process, store, and transfer data.
In our daily life, we have the good, bad, and ugly. We usually consider positive as a synonym for good and negative as a synonym for bad. Since a positive value multiplied by a negative value always produces a negative value, we can think that the potential of a system composed of good things and bad things is also negative. Regardless of the sign on the potential of that system, we can add value to any system if our work and effort is changing that potential to a higher level.
I can give many examples to show how human daily activities improve the potential of systems around us: chasing criminals and putting them in prison; extracting silicon from sand; screening out scrap from product. In reality, testing could be considered as the last subtractive step in the whole semiconductor manufacturing process.
Now, the question is: How much work do we need to do and is it worth doing? The answer to the second part of the question is, of course, yes. Some negative elements can create such negative potential that they can be detrimental to our society or to our final product. For instance, organized criminals or terrorists in the global community render the global community less safe; one flawed transistor in a chip of one billion transistors that renders the chip useless.
So, we spend energy and money—sometimes even human life—to chase criminals or terrorists in order to avoid another disaster like 9/11. We also screen out any chip containing a bad transistor in order to avoid an even higher cost in the final product—even if that cost will be higher than the original cost of making that chip.
The previous analysis demonstrated that test, as a part of production process, moves product from an initial potential to a higher level of potential. The value added by the test needs to be measured by potential elevation in that product. I do not have a formula to calculate potential of product that corresponds directly to the one in physics. Even if I could formulate one, others would disagree with my version. But one way to measure the potential of a product is by how much people are willing to pay for it. After all, how much would people want to pay for a production output if we were to tell them that the products have not been tested?
According to today’s standard, semiconductor companies are talking about zero defects. Although I do not have a formula for that potential, I think I can define that potential as zero for our future reference. That will be our ultimate achievement in terms of product quality.
If we think this way, then test activities have a great value. People doing this job should be proud of it. There is no question that we should reduce the cost of test. But we need to attract and retain the talented people who can invent the test techniques by giving them the respect they deserve.
Fang Xu is senior scientist at Teradyne’s semiconductor test division. He has been an invited speaker at multiple international semiconductor test conference occasions. He holds multiple patents in instrumentation techniques and architectures. His publication of “Algorithm to Remove Spectral Leakage, Close-in Noise, and Its Application to Converter Test” at the 2006 IEEE Instrumentation Measurement Technology Conference received the best paper award. He received the License en Science in EEA and Diplôme d’étude approfondie in Instrumentation from Université Paris Sud, France, in 1983 and 1985, respectively, and Docteur en Science from the same university in 1990. He worked at DRUSCH SA from 1985 to 1994 and designed NMR spectrometer and imaging systems.
Related Content
Related Content
Sponsored Links
TMW Resource Center
Browse Resources by Type:
- Back-to-Basics: Measurement-Based Channel Modeling for Signal Integrity using Agilent ADS
Agilent Technologies | Web Event
VIEW NOW
Protek, Inc. | Data Sheet
VIEW NOW
Protek, Inc. | Data Sheet
VIEW NOW
View More
