The title is very well put. These days you have to use reference designs to meet time to market demands. They are generally well done. However you must assume that there are possible mistakes and that you must still do your due dilligence and perform a design review. Also many reference designs are really just schematics to an eval board that probably has many more bells and whistles than you need. You must know how to cut out those without compromising the circuit.
Reference designs are great, because they provide a quick example of how to wire up a circuit. Sometimes they will even work for your application.
However, people who just use a reference design as-is without learning the fine points will face the consequences. As Stevie Wonder said, "When you believe in things that you don't understand, then you suffer."
Good reference designs are informative. The older data books had many useful reference example schematics. It seemed though as more and more devices were added over the years, there were less and less pages devoted to these. Then along came the internet and online data sheets were no longer restricted to a page limit in a data book.
I did get mislead once. The reference design for the 78L05 (+5 volt regulator) did not include the fact that when sharing a common load with a 79L05 (-5 volt regulator) they were sensitive to which supply was applied first. I had a merry chase into why some of my cards were failing to bring up the 78L05 during live insertion. Turned out if the negative supply card edge connector pins contacted prior to the positive supply pins the 78L05 would hang due to reverse voltage at its output due to the common load, an ECL to TTL translator.
The kicker was that I searched the 78L05 appnotes and reference designs to no avail. The solution was finally found in the 79L05 data sheet, a germanium diode in parallel with the 79L05 intended for this specific problem. Could never figure why the National Semi editors would place the solution to a 78L05 problem into the 79L05 data sheet.
The usefulness of 'reference' information to the engineering community can easily be seen by the immense popularity of such classic publications as
The famous "Quadzilla" publication
RF related books on RF transistors, Power, etc.
Power Semiconductors handbook
SCR and Thyristors
Op Amp Series
and on an on.
Of course these were mostly data books but with some decent applications data to start designs with, which ultimately became good marketing publications as well. Engineers looked forward to receiving the next year's editions.
These were very helpful when designs were made from simpler components rather than the highly integrated and programmable parts today. You had to understand the purpose and performance of each part in the system to achieve your goal as opposed to buying a pre-packaged integrated chip that now does almost the same thing. But...such is progress.
Certainly if you peeked into the bookshelves of some of the older engineers you will find these books still occupying a place of prominence. While parts may have changed, there is still much to be gleaned from these types of publications. If nothing else....a history of our profession as a progression of parts.
I have always seen reference designs as "your mileage may vary". I do not expect to be able to just insert them into my design as is and that they will work flawlessly. There is always some tweaking that will need to be done to solve the problem you are working on. In short - a good starting point that will get you anywhere from 50% to 85% of where you need to be.
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...