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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Fine pitch paste release problems.

Hi, I have checked the archives, but need some advise. ... - Dec 16, 2002 by Grant  

#22724

Fine pitch paste release problems. | 16 December, 2002

Hi,

I have checked the archives, but need some advise.

We have a 256 pin 1 mm pitch BGA part of our boards, and we have having problems with paste release our of the holes for the BGA. We just cannot seem to get it right, and we have to keep the stencil incredibly clean, however even then, we cannot get reliable paste release.

We have a 5 thou thick laser cut stencil, with stainless steel. The holes for the BGA are .4 mm, and the pads are the same at .4 mm (15.75 thou)

We cannot really enlarge the holes any more, as the stencil is already the same size as the PCB pads. We could go larger but the holes would be larger than the pad. We could re make the PCBs however this would take time.

Our stencil printer is manual, as we have not chosen an automated printer yet. We wanted to get some production done on this first, then get an automated unit. We are using a metal blade.

Does anyone have some advise on how we could solve these problems. What factors are cause poor paste release with fine pitch components.

Thanks for any advise.

Regards,

Grant Petty Blackmagic Design

reply »

k_h

#22729

Fine pitch paste release problems. | 16 December, 2002

Are you using "on-contact" printing? (is the stencil touching the PCB during squeegee travel)

reply »

Stephen

#22731

Fine pitch paste release problems. | 16 December, 2002

What paste are you using?

reply »

#22733

Fine pitch paste release problems. | 16 December, 2002

square apertures provide better paste release than round in most applications. See IPC 7525 Stencil Design Guidelines

reply »

#22746

Fine pitch paste release problems. | 16 December, 2002

Hi,

Yes, the stencil does touch the PCB. Is there another method?

Sorry to sound a bit dumb on that, however I don't have a lot of experience with automated stencil printing, and manual has been working well up to now with this BGA.

Regards,

Grant Petty Blackmagic Design

reply »

#22747

Fine pitch paste release problems. | 16 December, 2002

Hi,

We tried AIM no clean, and one other paste which I cannot remember the name offhand. We have been using Koki lately, and that's been the best. It seems to leave less residue behind in the stencil.

Regards,

Grant Petty Blackmagic Design

reply »

#22748

Fine pitch paste release problems. | 16 December, 2002

Hi,

Thanks for the info, and I will check it out. Do you know where the IPC 7525 spec is? Does it explain why square is better?

I know there is some theory I am missing here, and that's why we have problems. I am just freaking out trying to establish what makes a good print, and what is wrong in our situation.

Thanks for the info.

Regards,

Grant Blackmagic Design

reply »

Randy Villeneuve

#22753

Fine pitch paste release problems. | 17 December, 2002

Personaly I would try a different solder paste. Give Indium a call and get a sample of their SMQ92J No clean. You will be impressed. Alpha and Kester also have good products but we have a long history with this paste and we have printed millions of boards with .016" pitch or micro BGA's parts with no problems.

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

Stephen

#22756

Fine pitch paste release problems. | 17 December, 2002

Are your stencils electropolished? I still haven't used an electroformed stencil, but I understand they can be worth the extra money, because of the better release characteristics.

reply »

#22766

Fine pitch paste release problems. | 17 December, 2002

Square apertures can have a larger area within the same space than a round aperture.

You can buy IPC 7525 from IPC.

The theory you are missing has to do with: * Aspect ratio * Area ratio

Search the fine SMTnet Archives for background on these ratios.

Further, as an earlier poster noted, paste selection is an important factor in obtaining proper release in fine pitch applications. I want to say there there is discussion on this in the fine SMTnet Archives, also.

reply »

iman

#22769

Fine pitch paste release problems. | 17 December, 2002

Check :

1) design aspect ratio / are ratio required to fabricate the stencil you r using.

2) check if the stencil maker is using your checked aspect ratio / area ratio? if not the same, what are they using? any QA report confirming actual dimensions matches their supplied theorectical calculations shown to you?

3) check if your storage conditions and paste control is proper and correct? ours is 4-7degC for our paste.

4) check if your humidity and temperature control for the printer machine environment is somewhat 22-26degC / 40-60RH%?

5) check if your printer setup coplanarity is even and squeegee contact to stencil level? try running a dial gauge indicator to check if there are "humps" along the surface area the squeegee is travelling during printing? also check for PCB levelness (warp/bow/twist?).

6) follow the advise to get square stencil aperture openings.

7) see if electrofoam stencils are a viable option for u? such stencils do help.

note: many a time, lousy design can still be overcome with some creative process engineering techniques. Yeh, i know, its a bandaid.

Good luck!

reply »

Steve

#22770

Fine pitch paste release problems. | 18 December, 2002

I dont know what kind of screener you use but my 1st impression is that the board and stencil are separating too fast,the paste doesn't have enough surface tension to stay on the board, you're tearing the paste off the board. If you can, slow down the separation speed between stencil and board. That's my best guess. Good Luck.

reply »

#22772

Fine pitch paste release problems. | 18 December, 2002

One thing that will also cause this is very slight stencil mis-alignment. Sometimes the paste will have a tendency to hang up in the apertures if you are over hanging just a little on the solder mask. Some solder mask are very slick and will not pull the paste out of the apertures like the component pad will.

reply »

#22781

Fine pitch paste release problems. | 18 December, 2002

That issue (gasketing is what I usually see it called) was the driving force behind us getting new printers. Our old semi-auto clamshells just weren't repeatable enough, particularly over changes in squeegee travel direction.

If the paste in the aperture isn't in contact with something on the pcb (and it can't all be touching something if the edge of the stencil aperture is resting on the top of a mound of HASL tin/lead) it won't pull out of the stencil.

reply »

#22786

Fine pitch paste release problems. | 18 December, 2002

I have had a similar problem in the past, since I have never worked with a semiautomatic printer there could be a few things I recommend that will not work for you. 1. Make absolutly sure that you have proper support under all of the BGA's. 2. Print at a very low speed but do not go below 12mm/sec. 3. Use the slowest separation speed possible. 4. Spend the money for an electoformed stencil, I have conducted tests that proves they are worth the money and last 6-8 times longer than stainless. 5. Use stainless steel squeegees if you are not using a step stencil. 6. Make absolutely sure the board is properly fixtured and cannot move during the print cycle. 7. Increasing your print pressure could also help you gasket to the pad better and force the paste down to the pad so you can get a good release. 8. Verify that you are within the rated temp/humidity levels for the paste you are using.

reply »

Brian Doyle

#22789

Fine pitch paste release problems. | 19 December, 2002

You can also get that IPC standard in the SMTnet Bookstore (along with a couple hundred others)

http://www.smtnet.com/bookstore/index.cfm?fuseaction=view_item&item_id=848

reply »

Guy

#22794

Fine pitch paste release problems. | 19 December, 2002

Really? Where does it say that? I have had the opposite experience and 3.2.2 from the IPC-7525 states for square or rectangular apertures - "Having a radiused corner for all apertures can promote stencil cleaning". It actually recommends circular apertures for BGA. for Fine Pitch square with rounded corners are recommended but this is not so much to improve release but to reduce the likelyhood of shorts. The distance from one paste brick to another is increased and effects of misregistration reduced.

reply »

Guy

#22795

Fine pitch paste release problems. | 19 December, 2002

Huh?! The area of a square that fits in a circle is clearly less than the circle. Paste release is better in a circle than a square because the circle has maximum area per perimiter lenght.

reply »


RDR

#22796

Fine pitch paste release problems. | 19 December, 2002

Your area ratio for this aperture is fine (.78) usually any value greater than .66 will provide a reliable paste release. I would make sure however that the stencil is polished. I agree with other posts regarding your setup and/or paste type is probably the culprit.

Russ

reply »

Guy

#22797

Fine pitch paste release problems. | 19 December, 2002

Now, this is good solid advice. If following this advice does not help there is something wrong with 1. the paste (reology failure) 2. the stencil (rough walls) 3. the board (bad design or fab) (in order of likelyhood)

reply »


RDR

#22798

Fine pitch paste release problems. | 19 December, 2002

We actually look at it like this, The area of a circle that will fit into a square is less. (slight overprint with square aperture with radiused corners)

Russ

reply »

mark

#22800

Fine pitch paste release problems. | 19 December, 2002

It sounds like the stencil wasn't polished. Most stencil sources that only offer laser cut stencils do not have the electro polishing chemistry because they don't want to deal with the waste treatment issues. I suggest you verify that the stencil is polished. If the vendor says you don't need polished stencils - find another vendor! There is a good chance "you don't need polishing" because they can't do it.

reply »

mark

#22801

Fine pitch paste release problems. | 19 December, 2002

It sounds like the stencil wasn't polished. Most stencil sources that only offer laser cut stencils do not have the electro polishing chemistry because they don't want to deal with the waste treatment issues. I suggest you verify that the stencil is polished. If the vendor says you don't need polished stencils - find another vendor! There is a good chance "you don't need polishing" because they can't do it.

reply »

#22805

Fine pitch paste release problems. | 19 December, 2002

Hi Grant Petty,

I read about your problem, You might have problem with the design. First on the pad. What is the diameter of the BGA pad in your PCB. We have the same size of PCB and we use 0.584mm Pad dia. and a 10% reduction on our aperture. We use a semi-auto printer as well but has no problem with regards paste release on any solder paste brand.

reply »

Roger

#22809

Fine pitch paste release problems. | 20 December, 2002

Although everything that has been said earlier is sound advice for achieving good printing results, I strongly believe your problem is your manual printer and its inability to accurately control the speed at which the PCB separates from the stencil after the print stroke. I call this speed the 'snap-off' speed. After the end of the print stroke, a slow snap-off speed (0.5 - 2.00 mm/S) moves the PCB vertically down from the underside of the stencil until completely free of the stencil. This vertical, slow controlled separation I'm sure will overcome your problem. It works for me. There are several machines on the market that perform this separation control. I'm sorry to say but I think your answer lies in a new printer.

There is an independent organisation in the USA that might be able to help with your process - try http://www.aciusa.org

reply »

D. Adams

#22812

Fine pitch paste release problems. | 20 December, 2002

Check your stencil apertures making sure they are trapazoidal and smooth. I have hadjobs where the laser stencil was improperly cut (poor laser settings leaving a rough hole even after electropolish. Be sure to specify a .0005 trapazoid in your stencil Docs. I was using squares for everything but stopped specifying them when I did not see a big difference. I would still use square apertures on applications where higher paste volumes are desireable such as cerammic BGAs

reply »

P.SANKARA NARAYANAN

#22818

Fine pitch paste release problems. | 21 December, 2002

The easiest simplest solution i can think of is the Print plate from OKI. Why dont you try the same. Visit http://www.metcal.com. Instead of printing on PCB you can print on the BGA.

reply »

Mark

#22826

Fine pitch paste release problems. | 22 December, 2002

Hi: Your stencil specs are good; Did you check the stencil thickness? It would be worth a ck since there may be an error from the stencil mfr

If the paste is not releasing with good stencil specs, these are possible causes:

1 The paste isn't touching the pads (never gets there) - try flooding hard (like ten deg angle w squeegee, fast and hard). Clear with a firm 90 deg stroke (straight up w the sq)

2 If everything is very clean; sometimes the stencil will gasket against the pc soldermask, and air will trap in the aperature preventing the paste from touching the pc lands. To check this, skew stencil off alignment a little and try printing.

3 Microscopic dust on the pcb can cause similar prob. Be sure boards are clean.

Transition Automation Inc.

reply »

#22827

Fine pitch paste release problems. | 22 December, 2002

I am breaking a cardinal rule with this reply, but here goes. I believe almost every one is missing this one. Roger has hit the nail on the head. The gentleman is useing a clam shell printer and when the stencil lifts (too fast) and at the angle of separation--we have solder left in the aperatures. Another engineer who is trying to use a printer that was not designed or built for bga's. Yes I know some people can use them and I know most don't work correctly. Too many of these print and process problems return to the same problem that we see all year--wrong printer for the end results. Some may agree and some may not, anyway good luck.

reply »

Fluid Dispense Pump Integration

Boundary Scan