reflow theta alignment

Hi everyone, this is my first time on this forum. I've only been an innocent bystander. Until now. I wonder if anyone out there knows of any technical publication (accessible over the net) that explains, in detail, the actual physics involved regarding the "pulling" of certain parts into alignment during the reflow process. The capillary action of certain solder pastes is of particular interest. I have a QA technician performing pre-reflow inspection of PCB's and is constantly requesting that manual alignment be performed (either editing the place program or "tweaking" the parts) on 1206 caps and resistors. I need hard evidence, published material to show them that this is not necessary.

reply »

Hi Jason! I can probably find something documented for you, but the easiest way to convince your QA people is just to show them. If you have a board around that has some 1206 locations all in a row, print the board with paste and then place the parts incrementally off the pads...you know, place the first one dead on, the next one 10% off, the next one 25% off, etc...and then reflow it. They'll be able to see with their own two eyeballs that you're not tryin' to blow smoke up their butts...it's simply a matter of the physics of surface tension, wetting forces, and component weight...and they'll also see how far off a 1206 component can be and still pull back into alignment. Another illustration that I've used before trying to explain and show why this works, is to take two flat pieces of glass (like microscope slides) drop some water between the plates, and then set one on top of the other. If you try to separate them in the vertical direction, it takes quite a bit of force. You've just shown them how the components remain on the bottom of a board during double-sided reflow. Then you try and separate the two plates in the horizontal direction, they'll snap back into place (provided that you don't go overboard in the distance that you slide them apart) and that will show them the forces that are at work "self-aligning" the part during reflow when the solder is liquid...it behaves the same way as the water working on the two glass plates. One thing about this phenomena, is for it to work as advertised, you need to have the correct pad geometry and footprint for the part, and the part needs to have good solderability. I wish you luck...(I know some QA people can be a little hard headed sometimes....GRIN) -Steve Gregory-

reply »

| Hi Jason! | I can probably find something documented for you, but the easiest way to convince your QA people is just to show them. | If you have a board around that has some 1206 locations all in a row, print the board with paste and then place the parts incrementally off the pads...you know, place the first one dead on, the next one 10% off, the next one 25% off, etc...and then reflow it. | They'll be able to see with their own two eyeballs that you're not tryin' to blow smoke up their butts...it's simply a matter of the physics of surface tension, wetting forces, and component weight...and they'll also see how far off a 1206 component can be and still pull back into alignment. | Another illustration that I've used before trying to explain and show why this works, is to take two flat pieces of glass (like microscope slides) drop some water between the plates, and then set one on top of the other. If you try to separate them in the vertical direction, it takes quite a bit of force. You've just shown them how the components remain on the bottom of a board during double-sided reflow. | Then you try and separate the two plates in the horizontal direction, they'll snap back into place (provided that you don't go overboard in the distance that you slide them apart) and that will show them the forces that are at work "self-aligning" the part during reflow when the solder is liquid...it behaves the same way as the water working on the two glass plates. | One thing about this phenomena, is for it to work as advertised, you need to have the correct pad geometry and footprint for the part, and the part needs to have good solderability. | I wish you luck...(I know some QA people can be a little hard headed sometimes....GRIN) | -Steve Gregory- |

reply »

Thanks for the advice, Steve. Hard-headed people in high-ranking positions spells trouble.

reply »

Whoo-doggies Jason, you said a mouthfull!! (GRIN) Anyways, 'member what I said about me being a pack-rat? Guess what? I found some documentation that I've had for a while that does document the self-alignment forces that goes on when reflowing. I have some pages from a study done by some engineers from Philips Product Application Group documenting a self-alignment comparison between two styles of terminations on a chip resistor network...you know, those little teeny, tiny RNET's that are the size of a 1206? There's a lot of people that don't know there are two styles of terminations, Concave and Convex... ..(the Concave ones are the best in my book) Anyways, I've got a hard copy of this page from the study and it has some good pictures of those RNET's before reflow, and then after reflow. Both a Concave and a Convex style RNET was placed misaligned on the pads and a picture taken, and then another picture was taken after reflow...and shows the Concave RNET had self aligned during reflow. So this SHOULD be good enough to show that ther are forces at work during reflow that WILL self align the part (provided that the misalignment isn't too terrible) What I did was take a picture of the page with a digital camera I have and converted to a bitmap, and I'll email it to ya'...the page won't fax very good because it's a photo (too dark). The camera I have is a Kodak DC50, so the resolution isn't all that spectacular, but it shows things well enough... -Steve Gregory-

reply »