Leave the Rules of Thumb to the Experts

2016-03-02 16:03:12 by

engineering design rules
engineering design rules

You know how it is.  You’ve done something over and over again.  You’ve gotten the same results.  You’ve put in your 10,000 hours and I guess on some level that makes you an expert.  Maybe you’ve developed a shortened process, maybe you’ve come up with a new way to do it,  maybe you have a few rules of thumb you can share with others.  Rules of thumb are a sort of shorthand that experts use to guide others in the process and help them get a quicker solution.  The chances are rules of thumb work almost all of the time, but what about the one time you get caught in a bind.

Rules of thumb are often just what you need in a pinch.  I can tell you from experience that you can change out a cup of buttermilk for a 15 Tablespoons milk mixed with 1 Tablespoon lemon juice (allowed to sit for 5 minutes).  Are there instances where this method won’t work as well? There sure are, but the vast majority of the time it will work just fine. I like to think of rules of thumb as quick tips, substitutions or guiding principles developed from experience not scientific data.   In order for you to properly use them you need to understand the underlying theory and not rely on them blindly.

In engineering rules of thumb can easily lead to over or under designing a system.  For example, when setting up your microstrip trace spacing we all know that it should be greater than the dielectric height.  The underlying thought is trace spacing vs crosstalk flattens out when you get to enough spacing.  So if you always use 3x spacing you should be good right?  Well, yes and no.  You can easily be over designing in this instance. The only way to shake out the real answer is to run simulations to understand the process.  Over designing results in extra cost and under designing can lead to product failure.
You’re best bet is to understand where rules of thumb originate and how to use them in a way that won’t cost you time, money, or project failure.  When in doubt always simulate, simulate, simulate.


Read Comments

Leave a Reply

Your email address will not be published. Required fields are marked *