PCB Assembly Services - Screaming Circuits: A Few Hints of The Centroid File


A Few Hints of The Centroid File

NOTE: I've recently (8/18/2010) updated this post and the downloadable PDF to match the IPC 7351. We will properly assemble both ways, but this now matched industry standard.

Every now and then, we get questions on the centroid file (AKA Pick and place file or SMT locations file). Most CAD applications will create one for you. If you use Eagle, download our ULP and run it to create a centroid from your board.

If you want to poke around and need some hints on what's what you can download our Understanding the Centroid file r2. Here are a couple of illustrations from the guide. Fist, the point of origin needs to be centered in the part.

Chip originCopy of inline SMT connectorIt should be centered in a box that contains the outline of the pins as well as the body of the part. The chips on the left are easy. The connector, to the right, is a little more ambiguous, but as you can see, it's centered around the imaginary box containing the area.

Top-side rotation goes counter clockwise as shown on the left image and rotation on the bottom side is simply aChip rotation Chip rotation Bottommirror image left to right, with clockwise rotation.

Diodes and other passives should have their zero rotation horizontal, with pin one (if there is one) on the left. Passives orientation r2

That would place the cathode left for diodes and the positive side left for electrolytics and other polarized two lead parts.

Duane Benson
If you get dizzy spinning counter clockwise, go to Australia


TrackBack URL for this entry:

Listed below are links to weblogs that reference A Few Hints of The Centroid File:


I haven't found that pin configuration in any of my IPC docs. Given that, I would, as you suggested put pin 1 in the 12 o'clock position. It occurred to me that you might be talking about a thru-hole part. If so, I'd still go with 1 at 12 o'clock, but it won't matter for the centroid because we hand insert thru-hole parts.

Yes, they are in a triangle. The only other choice would be a straight line which would be a SIP and is mentioned in the PDF file.

Pin 1 and 2 are along the centerline of the part with pin 3 on one edge of the package at the mid-point between 1 and 2. The pins are numbered counter-clockwise. Picture Pin 1 at 12 o'clock, Pin 2 at 6 o'clock and Pin 3 at 3 o'clock.

So Pin 1 can only be oriented at 12 or 9 o'clock positions. I would say the 0 degree rotation for this part should be oriented with Pin 1 at 12 o'clock position. But this is not clear from the PDF file.

Hi Rick; Are the pins in the outline of a triangle? If not, what shape do they fall in?

I was working on a new board design and have a component with three pins (not a SIP) with pin 1 on the centerline of the part. I'm a little unclear on how to orient this part for zero degrees. Your doc says, "Three-pin and odd shaped parts have pin one in the upper left like dual inline packages." With pin one on the centerline it either has to be at 12 o'clock or 9 o'clock. Which is considered to be "upper left"? I am guessing this part should be treated like a PLCC with pin one in the 12 o'clock position.

BTW, your PDF is great, much better than the IPC document. Thanks for the good work!

Hi Rick;

Both the blog post and the PDF now match IPC 7351A.

I am having trouble finding a consistent reference for how to spec component rotations. I did finally find a spec, IPC-7351, which defines pin 1 orientations and rotation, but doesn't address how the bottom side of the board is referenced. This spec disagrees with your "Understanding the Centroid File" note with respect to passive orientations. They say pin 1 (+ on caps, cathode on diodes) should be to the left and all passives should be horizontal, not vertical as your article says. I can't generate design files that fit both your requirements as well as the IPC spec which I believe others use. How do you suggest I resolve that?

Post a comment

If you have a TypeKey or TypePad account, please Sign In.

« Via in Pad with Passive Components | Main | Ye Hardware Engineer on Quest for Firmware »