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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


QFN aligment issues

Views: 8441

Hi guys, I need your advise on this, I have a very uncommon ... - Jun 29, 2012 by Jorge_Quijano  

Look at this thread [ ... - Jun 29, 2012 by davef  

#66705

QFN aligment issues | 29 June, 2012

Hi guys, I need your advise on this, I have a very uncommon PCB design (at least for me, I've never seen anything like this...) it is 32" x 18" x 0.062" and has 9 QFNs that are driving me crazy, I'm not able to get a good alignment, I can get a medium alignment in one QFN and a terrible one on the next one at the same PCB. I'm guessing that maybe the stencil has some offset or maybe the board, we are getting many bridging errors, I have tried 2 stencils but same results, is it common to have offsets from the stencil or PCB manufacturer for this uncommon size? I'm using a carrier to hold PCB while printing to avoid warpage but, could the small thickness affect the alignment? PS this board has no fids so I'm using pads as reference, haw could this affec my alignment? I'm sharing some pics to give you a better idea.... please share our comments, regards!

Attachments:

reply »

#66708

QFN aligment issues | 29 June, 2012

Did you order this stencil? Do the gerbers you have for the board match the actual board?

reply »

#66710

QFN aligment issues | 29 June, 2012

Yes, I ordered this stencil, I checked the "cut" file vs the gerber before the stencil was made, and it matches. Attached is an overlap of the stencil over the original gerber.

Regards

Attachments:

reply »

#66711

QFN aligment issues | 29 June, 2012

Look at this thread [http://www.smtnet.com/Forums/Index.cfm?CFApp=1&Message_ID=66586 ] that on it's surface has little to do with what you're talking about, but actually may be talking to the root of the issue that you're concerned about.

reply »

#66712

QFN aligment issues | 29 June, 2012

The stencil doesn't seem to match the pcb. On the pcb there seems to be pads that are wider than others whereas the openings in the stencil seems to be uniform. As Phil mentioned, make sure the gerbers match the stencil. As for no fiducials, that shouldn't be a problem as long as you're using pads that are unique in size and isolated from other pads.

reply »

#66713

QFN aligment issues | 29 June, 2012

I guess I would atribute it to either board stretch/warpage or someone messed up, whether it be the gerbers aren't quite right or the stencil wasn't cut exaclty as the checkplot was laid out.

reply »

#66714

QFN aligment issues | 29 June, 2012

This issue can be common on such large size boards. The newer laser we use today to cut stencils are generally accurate to +/- 5microns X and Y. Probaly way more accurate than your board which will have a +/- into thousands of an inch. Try send a board to your stencil vendor so when they cut it they can overlay to make sure it matches. This is very common today due to board movement and the small components not allowing you the luxury of misregistration. Also make sure you Pick and place is calibrated and you have good board support.

reply »

#66726

QFN aligment issues | 2 July, 2012

I disagree with the statement that using pads as fiducials shouldn't be a problem. If he's using pads on the PCB, than he is most likely using corresponding apertures on the stencil. During production, these apertures will start to have paste accumulate around the edges. This may very well throw off the alignment. A board this size, with QFN components on it, should have fiducials. I suggest sharing the IPC-7351, (Generic Requirements for Surface Mount Design and Land Pattern Standard), with your PCB Designer, and/or Customer, and stress the inclusion of Fiducial Marks on all future designs/revisions. The problems that you are experiencing with alignment are the exact reason fiducials are recommended and encouraged.

reply »

#66727

QFN aligment issues | 2 July, 2012

using fids instead of pads I agree would be more accurate. However if the board and stencil do not register then nothing will align. We have started to fo many stencil to board matching due to board distortion either before or during production. We also recommend a slight apperture shrinkage just to help with some alignment issues.

reply »

#66728

QFN aligment issues | 2 July, 2012

Thanks for you inputs guys. We left our customer know about the design issues, they will add fids on the next version, I hope it helps. meanwhile we'll deal with those issues. I tested a new stencil today and It has a better alignment than the older one but it still is misaligned. How can you make a stencil starting from the PCB to make it match, is the PCB scanned or something?

reply »

#66729

QFN aligment issues | 2 July, 2012

There are a few different ways depending on the vendor. If they have the board and overlay it once it is cut. Make adjustments if necessary. If it is a double sided board do you see any movement after the first reflow? This should also be taken into account. You can call me at 603 401 1011 to discuss if you wish. Hope you all have a safe and happy July 4th celebration

reply »

#66758

QFN aligment issues | 5 July, 2012

That's a long board! Based on your description, it sounds like board stretch may be your issue. Look at where the paste is printed on the board. If you have board stretch, you'll see the print off to the left on the left side of the board and off to the right on the right side of the board (assuming you're lined up in the middle of the board). Are your defects more on the QFNs near the left or right sides of the board? That may be another indicator of board stretch.

reply »

#66759

QFN aligment issues | 5 July, 2012

What size frame are you using and what thickness and size is your stencil. I would ask your stencil vendor for their positional accuracy over 20-36 inches x-y. We have the capability also to make a photoplot in a controlled environment and over lay it on the board to see if it matches the gerber cut file. you can call me to discuss 603 401 1011 anytime

reply »

#66760

QFN aligment issues | 5 July, 2012

The PCB has 9 QFNs, distributed arround the whole board, but the ones in the edges are the ones with the problem. My stencil is 0.005" (I use to have a 0.006"), yesterday I realize after talking to my vendor that the stencil is manufactured in 2 steps because the frame does not fit into the machine, so they cut half the stencil + rotate the frame + cut the other half, I'm guessing that may cause such inaccuracy... dont you think so? anyway I will send the stencl and a board the them to check if the PCB or stencil is according to gerber file.

reply »

#66762

QFN aligment issues | 5 July, 2012

It is very possible We do have lasers designed to cut upto 7feet long. However they are mainly backplanes.I would think this could be your issue although I still think them having the board would help. I am not sure where you are situated. If you wish us to cut a stencil for you free for your evaluation feel free to contact me. I have over 40 years in printing espcially electronics. I have built lasers and developed the Eform process for Stentech I also have years in Etching. Most of my time is taken working with customers on issues like the one you have...you are not alone.

reply »

#66763

QFN aligment issues | 5 July, 2012

Thanks let's see what my vendor finds after verification and I will let you know. by the way I'm located in Juarez, Mexico. regards!

reply »

#66764

QFN aligment issues | 5 July, 2012

We did have a facility in Juarez until a few months ago. I know Juarez well. We do have a facility in Guadalajara. I used to do Mexico until it became a little unsafe. Please do keep in touch 603 401 1011 rob@stentech.com. if I can help you or anyone lese at anytime I am a phone call/email or a forum away (*;*)

reply »


lt

#66785

QFN aligment issues | 9 July, 2012

Your selection of pads to use as Fids will be critical. In your pictures of the QFN component, the pcb pads are primarily non-solder mask defined, but on 3 sides of the QFN the corner pads become solder mask defined due to the copper ground flood.

Review the pads that you selected to use as your fids. The pads of the pcb must "look" as similar to the apertures in the stencil as possible. If the pcb pads are not similar to the apertures, then their centroids will be shifted which WILL shift the position of the stencil on the pcb.

2 other things:

Place the pcb on a workbench. Place the stencil on top of the pcb. Manually align the stencil to the pcb. Is it possible to get all 9 sets of qfn apertures positioned on top of the 9 sets of qfn pads? If yes, then your stencil is cut correctly relative to the pcb. If no, then you'll never get a good print. Side note: I realize that manually aligning a 31" pcb/stencil will be a challenge.

If your screen printer allows, increase the number of fids used for aligning the stencil to the pcb....for example, instead of using 2 diaganol fids, use 4 fids in each of the corners. By increasing the number of fids used for alignment, the error induced because the fids/pads on the pcb do not exactly match the fids/apertures of the stencil should be minimized.

reply »

#66789

QFN aligment issues | 9 July, 2012

We used to have a manual printer machine, we could never align the stencil over the PCB well, now with the new stencil we have less defects but still appearing on the Top list. I tried to use more fids as you recommend and one issue I faced was that there is a silkscreen mark near almost every pad, so the Screen Printer camera did not regognize the pad correctly, I tried to set up the light to omit such marks but alignment was inconsistent...

reply »

#66795

QFN aligment issues | 10 July, 2012

until you know if you stencil does align with the pads on the board all the fids in the world will not help you. registration is key. We do it a lot at our facility in Toronto as we have lots of measuring equipment. In cases like this. We lay the stencil over the board to check for alignment. We also measure distances from point to point using the original and edoted gerber files. We also measure from the fid to a point or points to make sure the fids are in the correct location. laying the stencil over the board on a flat workbench is pretty easy. I usually use a 40x shop scope for alignment.

reply »

Matt Kehoe

#66938

QFN aligment issues | 2 August, 2012

Solid solder deposit technology allows you to adjust alignment after placement and before reflow, without causing shorts or smearing paste. Just an idea....

www.sipad.com

reply »

 Reflow System

MSD Dry Cabinets