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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Skewed transistors

tsongshe

#24194

Skewed transistors | 18 April, 2003

Hello,all.these days I met many skewed transistors after reflow soldering.we 've tried many ways attemting solving it thoroughly...changing the aperture pattern,solder volume and even changing the size of pcb pad,but the results are not very good.even though sometimes it seems we had made it,but it'll come out another time.these skewed transistors happens on all kinds of products...we 're dealing with.finally we found that these skewed transistors happened when using transistors from difffent suppliers...and the pad patterns of transistors are also diffrent.still we can't get the root cause.it'll be kindly appreciately if someone can give some suggesttions. is there anyone can tell me how I can attach a picture of diffrent kinds of patterns of these transistors?

http://paymail.china.com/extend/gb/std/bryan.c.x.she/Mi0021LJKSrZiXzqsyhlAaTwISUdPt/_T_m_P_/b/bryan.c.x.she/Sent/20/tansistor.JPG

Thanks &Best Regards Bryan c.x. Sheh

reply »

k_h

#24195

Skewed transistors | 18 April, 2003

Transistors? Are you referring to SOT-23 packages? Are the solder paste and part properly registered before reflow?

reply »

#24196

Skewed transistors | 18 April, 2003

tsongshe

#24197

Skewed transistors | 18 April, 2003

I'm sure the parts and solder paste are well registered. these transistors are commonly to252 package,and sometimes to263 package.I 've a picture of these two diffrent kinds of pad patterns of to252,but i don't know how to post it here... Thanks ®ards!

Bryan Sheh

reply »

tsongshe

#24198

Skewed transistors | 18 April, 2003

Thanks Dave.F. I 've carefully read what you recommended,but still can't get the answer.it's two different suppliers with diffrent pad patterns of each components.one of them can work well,while the other cant do.I think it must has something to do with the pad shape,but i don't know why!???coz we can't use the only one suppliers parts,we cant get out these problems! Thanks®ards! Bryan

reply »

#24199

Skewed transistors | 18 April, 2003

Brian

What did we recommend? We just gave you a link so that you can see how to get your pictures on-line, so that we can see what�s going on.

Anyhow, if we understand what you saying, [We're slow today. So, please be patient.] your situation is: * Sometimes you get TO252. * Other times you get TO263. * Each part has different lead configurations / pads. * One part works with the pad configuration you have, but the other part does not work. * Part that doesn�t work skews. [We bet it's the TO252 that's floating on the TO263 pads.]

Assuming your TO252 and TO263 pads look like these, http://www.necel.com/discrete/english/products/transistor/pkgquick.html

If this is correct, and your boss will not listen when you told him that this is insane: * Talk the designer into putting a set of pads for the TO252 and a set of pads for the TO263. * Use one stencil and two placement programs.

If that doesn't work, but you can still relay-out the board; try to jury rig a unique pad design that might work for both the TO252 and the TO263. [If you're forced into this choice, start with the 7x7mm pad and play with the smaller two pads. Test your alternatives by cutting and gluing [use Tak Pak, cyanoacrylate ester] pieces of copper foil on bare laminate, slopping some paste on the copper, setting components on the various choices, and reflowing the laminate.]

If that doesn't work and you can't relay-out the board, glue the floater in place with chip bonder before reflow and move on to something else, if they'll let you.

reply »

#24201

Skewed transistors | 18 April, 2003

"glue the floater in place with chip bonder before reflow"

gets my vote. We have decals for DPAKs and TO-252's that don't match either specification, and chip bonder is the only answer if they won't spin the board for you.

reply »

tsongshe,Bryan

#24206

Skewed transistors | 20 April, 2003

Thanks. *We use TO252 package,but sometimes the skewed components also happen on TO263 package. *we use TO252 package parts from different suppliers,and when we change the parts from one supplier to the others,few skewed transistors happen again. *the shapes of the pad on parts are different,as attached picture indicate. http://picturecenter.kodak.com/view/kpcViewLarge.jsp;jsessionid=baaaKOK9SAQnyS?machineid=7&page=1

I want to know why such skewed parts would happen? How can we deal with it ? We can't use the chip bonder etc. for cost considerations Maybe to change the apertures can solve this problem to some extent? you'll be kindly appreciated for your suggestions. Thanks&Best Regards Bryan sheh

reply »

tsongshe,Bryan

#24207

Skewed transistors | 20 April, 2003

Hi,all, I'm very glad here this site to meet so many friends,and learned much from you.It's very lucky that oneday I found this website(forum),and so I come frequently to share many precious experience with you.Thank you very much. Bryan.c.x.she@china.com REGARDS! Bryan Sheh

reply »

#24208

Skewed transistors | 20 April, 2003

Your link doesn�t work. It gives an application timed out message.

Pads on the board that are too large or spaced improperly for the component usually cause skewing. Tombstoning and skewing are first cousins. Look here: http://www.smtinfocus.com/smt_failures/smt_failure_tomb_stoning.html

Chip bonder is not expensive to purchase. We can appreciate that it is expensive to apply in some situations. In those situations, we expect the design of the board would be refined.

We doubt that changing your apertures will affect your skewing problem.

reply »

Bryan

#24212

Skewed transistors | 21 April, 2003

Thank you Dave F

reply »

k_h

#24218

Skewed transistors | 21 April, 2003

You might try reducing your solder paste volume. To much solder paste gives the part a pool to float in, nullifying the positive effect of surface tension. I use an aperture on stencil for the heat sink which is 5mils reduced on each side and a 25mil wide "X" on the diagonal. This gives you 4 triangles pointed towards eachother on a 90 degree polar array. My pcb terminal footprints are 1=1 on all sides except the heel where we add 50mil, this is where the solder fillet is for the gull wing type terminal. I use a 1=1 stencil aperture on these terminals. I think by reducing the solder volume (you can experiment by manually removing solder paste in a concentric pattern prior to placement) your allowing for the power of the solder's surface tension to take effect agian and help with alighnment assuming your pcb foot prints are not to large. I have not had any skewing problems with this in my full convection reflow process.

reply »

Jon Fox

#24219

Skewed transistors | 21 April, 2003

I remember from my days of machine setups that these packages' CAD data is actually off center from where the placement needs to be. Usually CAD center comes up with the parts more to the large collector/ground pad so the the smaller gull wing leads are just touching. How is the placement of these parts after placement and before reflow. Are they placed with enough gull lead touching their pads or is the heel of the lead just making it to the edge of the pads. Just remember that the pick up position of these parts is not the same as the center of mass. Other than placement, look at solder volume. We use a 10% reduction with a 25mil "cross" half way up and over to appature void to help keep the part from pulling too far over when reflowing.

reply »

k_h

#24223

Skewed transistors | 22 April, 2003

I'll throw this out there also: http://picturecenter.kodak.com/share?invite=HE2r4mzhP5YLJkvkhU1o It shows void slots (purple) in the copper on the perimeter of the smaller transistors heat sink. Should keep the smaller transistor from moving while the larger one simply covers them up.

reply »

jedduan

#24228

Skewed transistors | 23 April, 2003

I think it will make some progess if you modify the program and let the component shift towards the lead orientation as possible while P&P.

reply »

Jade Series Selective Soldering Machines

SMT fluid dispensing