Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Board Stretch...Is there a relevant IPC spec?

#17310

Board Stretch...Is there a relevant IPC spec? | 22 August, 2001

Can anyone point me to a relevant IPC spec that specifies an allowable board stretch, measured between metal features on outer layers (e.g., between fiducials and component pads on the topside of a board)?

I've got a non-linear stretch problem that's making it impossible to get a good print. When I put one of the rogue boards on our CMM, I find that the global fid locations look pretty good (usually within 0.001" of the CAD locations), but I find regions near the center of the board where the pads are shifted as much as 0.005" relative to the CAD location.

I'm confident in my CAD data and my stencil. I can print and place the 0.5mm pitch devices (13x 240-pin QFPs on a 15" x 11" board) just fine on the great majority of boards. On a few boards however, the print is way off locally, the placements are off in the same direction (not using local fids...yet), and the CMM measurements agree with the print and placement errors. It's definitely a localized, non-linear stretch problem.

It appears that I cannot expect my board supplier to buy the rogue boards back. They point out that there's no real spec covering this. Our drawing doesn't mention it, and the only IPC spec that my supplier found is not really applicable (IPC-D-300G...last revised in '84, and obviously not applying to SMT boards.). Is there another IPC spec that I can reference? If not, does anyone have a well-worded drawing note to cover this problem?

Michael

reply »

#17315

Board Stretch...Is there a relevant IPC spec? | 22 August, 2001

First place to look is IPC-6012A w A#1 "Qualification & Performance For Rigid Printed Boards".

I'll look at it later to see what's there.

reply »

Hussman

#17322

Board Stretch...Is there a relevant IPC spec? | 23 August, 2001

I've always asked the board house to send me the Gerber data they used to make the board. I've found that things got moved for reasons unknown to either of us (e-mail gremlins?). 9 times out of 10 I take this Gerber and have my stencil manufacturer re-etch a stencil to match (if the board is OK electricly and doesn't cause quality problems). It only costs about $300 and saves me time to take care of other issues.

reply »

#17326

Board Stretch...Is there a relevant IPC spec? | 23 August, 2001

I don't see anything in IPC-6012A specifically about stretch. Maybe paragraph 3.4 (Board Dimensional Requirements) applies, which tells me I need my own drawing note. Crafting this note could be tricky, and I was hoping that someone has already tackled this (?).

Michael

reply »

#17327

Board Stretch...Is there a relevant IPC spec? | 23 August, 2001

In this case, I know my supplier's data is good. They've been supplying this particular PN for a long time now, and the stretch problem appears to be random within a given date code (i.e., most of the boards in each lot are good, but a few exhibit this problem). My supplier hasn't determined the root cause yet. They noted that the outer layers are imaged with photo tools (artwork on mylar), and suggested that the problem might be eliminated if they use their LDI (laser direct imaging) equipment instead. I'm not familiar with the LDI process; is it supposed to be more accurate? Are there any drawbacks to LDI?

reply »

Michael Parker

#17330

Board Stretch...Is there a relevant IPC spec? | 23 August, 2001

Bingo - the mylar is the problem. Ask the fab vendor what storage and humidity controls they use on the photo tools. The films can changed geometry from day to day depending on the ambient environment. What controls are used during production? The random changes within a lot may very well be tracked back to morning startup, just after break restart, etc. Go for the digital controls. If the problem still persists, then a bigger issue is out of control at the fab house.

reply »

#17338

Board Stretch...Is there a relevant IPC spec? | 23 August, 2001

Yes, 3.4 does not seem to apply. Consider that 3.5 might be more appropriate.

reply »

Plasma prior Conformal Coating

Reflow Ovens thermal process improvement