Solder Paste and fine pitch components

Where I worked previously we had some boards with a couple 15 mil pitch parts. The first batch we got all had undersized pads. We had lots of bridging. We had less bridging with bigger pads. The ball size in the paste makes a bigger diffence regarding whether the stencil would clog up. Talk to your solder paste rep, the good solder paste companies have good reps to help you out. Make sure the stencil aperatures are not bigger than the pads. You may have to make them the same size. It is counter-intuitive to think that bigger pads (which means they are closer together) would have less bridging but that is the way it worked for us.

reply »

1) For solder short, pls check the paste height. Too much vol will cause short. For 15 mils use 5 mils thickness stencil if possible. Check profile preheat time to prevent hot slumpof the paste.

2) type 4 powder will be better for 15 mils pitch comp

reply »

I place 15-20 mil a lot. You need to look at your pad to apature reduction. Your stencil thickness. The type of paste your using. We use 2-5 mil reduction on each side as a general rule on fine pitch. This varies depending on pad width. We also use 5mil thickness on the foil. I have found that the omnix 5000 or the new omnix 5002 is very good for fine pitch parts. If you are moving your fine pitch prior to reflow due to bad placement in the machine you are going to see shorts. To much placement "z" will squish the paste out around the pads and cause shorts also. I hope this helps you a little.

Bob Merriam SMT Manager

This message was posted the Electronics Forum @

reply »

My suggestion is to conduct a Design For Manufacturing (DFM)study on the assembly in question. Pay particular attention to the pad sizes, component lead dimensions and satisfy yourself that those two factors are correct. Given that pads match component needs with respect to solderability, the next steps are straightforward. If design factors are incorrect, either fix the design or you are forced to find work arounds that may compromise your ability to produce a defect free product.

Secondly, pay close attention to your stencil apertures. Oversized or 1:1 ratio apertures in 15 mil pitch can and will cause bridging. Learn about and incorporate aspect ratio rules for your stencil design. Search the archives in this website for info on aspect ratio and stencil design rules.

Third - validate correct alignment of stencil to PCB. An offset of 3-5 mils in any direction can cause paste to be deposited that encourages defects during reflow.

Fourth - back to design of PCB. Are there solder dams of mask between the pads? If yes, that helps. If not, pay close attention to stencil aperture width and be sure that the aperture deposits less paste by 1-2 mils per pad, centered on the pad.

Next - paste volume is critical. A height measure of paste is necessary to validate that your printer is doing what it should. Height control, combined with a controlled Length and Width is key. Overall paste volume is inferred by measuring height only. Accuracy is determined using a paste measurement tool that calculates volume from actually paste deposit measures. Too much height or blow out on the sides of the deposit indicates squeegee pressure not optimized, aperture not aligned to pad, etc.

5 mils stencils work well with fine and ultra fine pitch parts. Normally I use 6 mil because of consideration for the rest of the SMT components. If you go too lean in height, all the other components run the risk of insufficient solder. Therefore you will have to open the apertures for all other parts to compensate. It is far easier to modify apertures of fine pitch parts only on a 6 mil stencil than to modify apertures of all components on a 5 mil stencil.

Talk with your stencil supplier. The CAD operators who design the final artwork to create the stencil are an invaluable source of information. Pick their brain and get recommendations of what seems to be the industry standard practice with regards to the component types in question.

Good luck and check back with us to update your success or need of further information.

reply »

People have made good comments. Additional points are: * Your bridging is probably being caused by one or more of the following: [1] printing too much paste, [2] smearing the paste during placement or subsequent handling, or [3] paste slump during the steps between printing and reflow soldering. * Be wary of guidelines for pinching apertures without commentary about the size of the component lead and size of the pad on the board. * Consider IPC-7525 - Stencil Design Guidelines as a worthwhile input. * Stencil fabrication is a key variable affecting paste release. * Print independant variables are not necessarily the same for both 16 and 20 pitch. * Selection of solder paste type varies with pitch, as follows: Lead Pitch||Mesh Type||Particle Size 0.025||3||-325/+400 to 500 0.020||3||-325/+500 0.016||4,3||-400/+500 0.012||4||-400/+625

In trying to troubleshoot this, consider telling us about: * Pad dimensions and finish. * Component lead size and solderability protection. * Aperture dimensions, stencil thickness, fabrication. * Paste type and mesh. * Printer set-up. * Etc

Finally, can you generalize about the bridging? Mostly: * Here or there. * On the side / sides of component leads. * On front / back of component leads. * Where ever ...

reply »

I'm with you dave...please provide specifics as to defect condition... Are you getting good wetting to both land and lead? Is there any repeatability to lead/location/device? Do you have paste inspection capabilites to verify deposition? Is it a Paladium leaded component?

Scott

reply »

We had lots of problems with bridging on 15.8mm pitch, QFP17-160 at the start.The board was .032 thick, double sided-six boards to a panel.

Here is the problems (not neccessarily in the biggest to little but as i remember them) and solutions that helped us as our DPMO after all the problems were fixed was next to 0.

1)Humidity and temperature in our SMT room. It was notice that during high hummidity/high temperature(very little hummidity control in the room) the solder paste would turn to soup after 1/2 hour of printing.Cycling solder paste on the stencil seemed to fix this problem.

2)Solder paste supplied was very soupy.After one minute of stiring the solder paste usually sticks to the stir stick. Two solder paste lots that i know of, the solder paste would just fall off the stir stick.When the past was applied to the pads slumping happened right away consequently our bridging increased.

Solution-Used another lot of paste and let these two lots age a bit

3)Perhaps the biggest problem( due to the size and flimseness of the panel)lack of support pins under key support areas casuing two sides to be full of solder paste (of course casusing many bridges) and two sides not enough solder paste for good heel fillets. Installed big hokin support pin right under the QFP and increased support pins in other areas.

4)Reduce stencil thickness from .006 to .005. Used elctropolished stencil. Pad size on the board for QFP was 0.008 wide by .070 in length. Aperture opening was 0.0075 x .070 on stencil. This seemed to help quite a lot.

5)QFP placed off pads slightly (combined with the four previous problems) had to skate back onto the pads. When skating back onto pads the QFP would drag the solder from next pad ensuring a nice bridge. Treated each board in the pannel as a board for programming purposes (Universal GSM used with QFPs supplied in waffle packs). This seemed to help place the QFP right on the pads.

6)Our regular solder paste used in house is the type 3. With all the above problems occuring a type four solder paste was used.After all the above problems were rectified, we went back to the regular type three solder paste. No future problems were noticed using type three solder paste.

Alas after all this work and many, many boards produced defect free the product went our China plant for future builds (Lots of manual soldering and manual mechanical work thus the overhead was a lot less expensive). No problems competing with China on the SMT side but as soon as the paws were used, we could'nt compete.

Jim M.

reply »

I would like to thank everyone for their time and replies

reply »