Recent horror stories remind us that designers really need to know it all.
A recent blog on the Circuits Assembly website got me going on one of my pet topics. This will come as no surprise to those of you who know me, but once I overcome inertia and a latent proclivity for procrastination, I usually have to take a nap, thanks to all the effort of overcoming said inertia and procrastination.
But not this time.
This time the subject close to my heart is design for manufacture. The blog, written by regular contributor Duane Benson of Screaming Circuits, is titled “Another Land Pattern Mixup." It is just so difficult to understand how, after all this time, we still cannot get right certain fundamental things between design and manufacturing.
Now I won’t pretend to know the particulars here. There may have been a mistake when the part was created; the part number may have changed somewhere along the line, or someone at the CAM station may have screwed up. I don’t know. But I do know that here is a board that has a land pattern that will not accommodate the component supposed to go in the circuit.
Duane also shows an example of what looks like a jumper strip that is probably a metric part, but at first glance, it appears the designer used imperial dimensions. It probably would work if there were only a few pins, but over the length of 11 pins, the conversion round off or tolerance means the part won’t fit into the board. At first blush, we naturally would say the designer screwed up. But what if somewhere along the line, purchasing or some other entity substituted a part that “looked the same” to save a penny or two? In many companies, a librarian is responsible for creating the footprints. Sometimes this person is not a designer or does not really understand all the implications of what they are doing. Whatever the cause, there is a preventable problem here.
I’m not going to put all the blame on the designer because the truth is, in many cases, neither side of this issue really understands the fundamentals of the other. Many people in manufacturing really do not understand who the designer is, much less what they really do. The only thing they know about the designer is what they read in the media or see in the Gerber files. And before we go further, I’ll add that, yes, some folks in the media write about design and manufacture and don’t know who or what the designer (or the fabricator, for that matter) is. They may fool some folks, but designers usually can tell the difference between those who have had to find a way to connect those last few traces and those who have not.
When it really gets down to where solder meets laminate, the buck stops at the designer. No matter what the problem is, more than likely whoever designed theboard is going to take the rap. In most cases, that is as it should be. Fabricators and assemblers may not need to know how to design a board, but designers need to know how every aspect of the PCB manufacturing process affects their decisions. One of my drafting gurus told me that, after I finished a drawing, I should stand back, look at the drawing, and ask myself if I could build the part from the information on the paper (yes, paper; it was the 1970s). The catch here is that if you don’t know what happens in the shop, you can’t answer those questions.
Over the past 20 years, PCD&F and Circuits Assembly have published countless articles about DfM. We’ve produced scores of classes at the PCB Design Conferences, and I’d like to think many designers know a lot more about manufacturing than they did before they read those articles or attended those classes. So it boggles my mind that we still see the kinds of problems Duane points out.
I know there are pressures to get the design out now! A host of other issues play into this as well. But what good is a design done quickly, if it has this type of problem? What’s the cost of doing another run of prototypes, both in dollars (or euros or yuan) and time? Every manager and bean-counter should be required to read Duane’s blogs and other horror stories, so they have a better understanding of repercussions of decisions made in the design process.
Having made those claims about our attempts to eliminate DfM problems, I’ll still take some of the blame. Our mission includes educating and bringing the sides together. Evidently we haven’t done a good enough job of that. In the coming months, we’re going to recommit to this mission. But it has to be something in addition to articles and sessions at our shows. Those are important parts of the puzzle, but we need more. Next month, I’ll let you in on what I have in mind.
Until then, stay in touch and, as humorist and author Tom Bodett says, we’ll leave a light on for you. PCD&F
Pete Waddell is design technical editor of PCD&F (pcdandf.com); This email address is being protected from spambots. You need JavaScript enabled to view it..
