Quite often when a laboratory is approached with a seemingly straight-forward problem, rather than propose a simple test to provide some relevant information, they outline a research program instead.
I donÆt know if itÆs overzealous salesmanship or simply that they donÆt know how to answer the customerÆs question (remember, the real customer may be internal rather than external to the laboratoryÆs organization). Too many times programs are based on the type of equipment available in the lab, or the specific expertise of the individual to whom the question has been posed. They immediately jump into familiar tests, sometimes with seemingly no regard as to the potential value the data will provide to the customer.
Whether you suspect that you might know the answer or not, testing programs should be approached simplistically. Like any other technical problem, the program should be approached in a scientific manner --- 1) form an hypothesis, 2) test the hypothesis, 3) form a decision about the hypothesis , and 4) either modify the hypothesis and repeat steps 1 through 3, or form a conclusion.
Problems should be addressed with an initial ôquick-and-dirtyö approach. Make a quick, yet thoughtful, assumption as to the problem. Perform one or two simple, inexpensive tests that could confirm or eliminate the hypothesis. Usually simple inexpensive tests donÆt provide all the information you need to answer the customerÆs question, but they can provide direction which can eliminate unnecessary, more expensive testing.
If the test doesnÆt confirm your hunch, now is a good time to rethink the problem, possibly get some fresh ideas from colleagues, and try another approach. If you hit a wall, honestly confront your customer --- donÆt continue testing in the hope that some answer will suddenly pop up.
If, however, the preliminary data supports the initial hypothesis, additional, more elegant testing may be substantiated --- but check with the customer. All too often the customer doesnÆt want the sophisticated answer; he too may only be looking for direction, and that ôdown-and-dirtyö approach may be all that he needs.
Another advantage for keeping the customer in the loop is that new information is gained each time you talk. Sometimes itÆs because you may have misinterpreted some information given, or more often than not, significant information is not revealed by the customer because he assumes it may not be relevant. Nevertheless, constant communication will help you get the information you need to solve the problem, as well as help you to better understand what level of sophistication your customer needs to satisfy his problem.
Remember, the purpose of a test program it to satisfy the needs of your customer, not to resolve all the scientific issues that may arise.
This brief provides a high level summary of advancements in Mentorâ€™s AMS Verification solutions since we saw you all last year at DAC 2019 in Las Vegas. Products featured: Analog FastSPICE Platform, Symphony Mixed-Signal Platform, Solido Variation Designer, and Solido Characterization Suite. Download Now
This resource will help you find key sessions-panels at DAC from researchers, customers, and our expert technologists; and various resources detailing the recent innovations in HLS, Verification, and RTL Low-Power since DAC 2019. Download Now
Among the challenges in the design flow has been aligning the metrics for design-for-test and for functional safety. This paper describes using logic built-in-self-test as both a functional safety mechanism and as a part of in-system testing, which allows for alignment of metrics required for safety certifications. Download Now
IoT systems are multi-domain designs that often require AMS, Digital, RF, photonics and MEMS elements within the system. Tanner EDA provides an integrated, top-down design flow for IoT design that supports all these design domains. Learn more about key solutions that the Tanner design flow offers for successful IoT system design and verification. Download Now