Stop. Doubt. Plan, then Draw: SEAM8 Test Board

How fast could you draw a 400 pin BGA board? 2-5 hours? How about 30 minutes? or 5…

400 pin 0.8mm BGA end-cap board design in Altium 18

“Sure, I can do it fast, 2 hours” that was my response when I’ve understood there was a need for a little PCB. This PCB will have the scope of short-circuiting the two ends of every differential pair routed in one of our main components in LHCb UT.

The Pigtail, this is the name of that component as we have a superb fantasy record, is the flexible circuit connecting the front end electronics infrastructure to the back end one. It is vital thus to be able to test connectivity and DC resistance on every single trace the PCB is connecting.

Our approach was then to design a fanout board to which we would connect our testing instrumentation, and an end-cap one; here we’re talking about the latter.

A board this simple, just a 2 Layer PCB with one single component having some pins shorted together, is anything but challenging. The problem comes with time. We don’t have such a resource, thus being fast and right the first time is essential.

Why am I writing this post? As usual, something “bad” happened.

Driven by the urgency of the project, I began working on the PCB right ahead. After asking info on the pins to be shorted to a colleague of mine, I’ve started connecting all the positive ends in each differential pair to their negative mates.

The mistake was I didn’t stop thinking about the faster way to do it in the first place.

Having already the Pigtail design, I could have just modified it. Renaming every net by removing the suffix _P or _N, so that the two ends of a differential pair would have been considered the same net (this is a flat 5-second work with “Find & Replace”) and shorted together as desired.

Bad, slow and not easily testable approach to the problem, even though it didn’t need any active thinking

The difference between the two approaches is tremendous. 2hours of development time for the first one (which I’ve followed), and 1 minute for the second (which I later did as the first route was impossible to debug).

All of this just for the schematic, which is one of the first things one should do when designing a PCB. Luckily for the PCB routing part, I stopped myself, looked at the schematic, and clicked “Auto-route”.

Optimal approach: re-use what you already have and use the functions provided by your CAD environment (Copy-Paste, Replace-Rename, Move, Select-All-in-Rule, etc…)

This saved me roughly 20 minutes and took me 20 seconds to figure out and execute as the PC was doing all the work.

Why? Not being a critical high-frequency board, the only thing that mattered to me was the connection between those pins, which is something a computer can easily nail in a few seconds.

Bottom Line:

Always, Always, ALWAYS stop and doubt that the way you’re following is the best one. Openly assess its pro and cons and move forward in the smartest way possible.


(Including Altium 18 files, gerbers, NC-drills, Schematic and PCB)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.