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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Micro Leadless Frame - solder wetting control

ianchan

#19832

Micro Leadless Frame - solder wetting control | 8 May, 2002

Hi mates,

have this SMT production issue :

A) We have a leadless-chip-carrier(LCC) IC package that is classified under the Land-Grid-Array(LGA) package family.

B) The LCC is 4mm x 4mm, and has 28 tin plated pads located beneath the outside perimeter of package body (think of BGA without the solder-bumps beneath the body).

C) The side of the package body perimeter has exposed copper leadless terminals corresponding in width and aignment to each of the 28 (tin-plated) pads beneath the body. The customer insists that this bare copper is its natural condition for the leadless terminal walls.

D) the solder we get has inconsistent solder-fusion onto the exposed portion copper of the leadless terminal walls. customer has classified the gap as the "unsolder" defect.

E) we did analysis as follows :

E.1) there is copper oxidization of the leadless terminal walls. it varies heavier with random frequency on each batch of LCC delivery to us (consignment).

E.2) we did evaluation by manual solder-dip some of the LCC with copper oxidization and found that solder-dip "Tin plated" LCC has excellent solder fusion onto the (now tin coating) leadless terminal walls. the Tin coating negates the copper oxidization layers as Tin assists solder fusion, we therefore get good solder fusion, ie. no "unsolder" gap :P

E.3) If we do get some tin coating "smearing" onto the copper leadless terminal walls at the raw materials stage (still in Tape and Reel format), it is a "bonus" side-effect of the die-slice-cutting during the LCC mfg process. customer is not going nor willing to change any raw materials yet.

E.4) we took a transparency of the : a) Stencil aperture openings b) PCB (Au/Ni) pads sizes and found the two coverage areas are mis-register, when superimposed on each transparency.

E.5) the PCB pads have "up/down/in-out" Pads irregular pad layout; even when the gerber data shows all the Pads are supposedly equal in Pads dimensional lengths and equal dimensional PCB layouts.

E.6) the stencil was generated using customer supplied gerber data of the silk-screen layer. this gerber supplied to us assumes all pads are of equal layout and individual pad sizes/lengths. this assumption is flawed as using a scanner machine we find there are 5 different varient pads dimensions measured, and PCB pad layout is irregular.

E.7) we have calculated the expected deposited paste volume for each individual PCB pad, and had the stencil fab house scan each pad to get accurate dimensions. a new stencil is fab according to the scan data of the actual PCB pads.

E.8) irregular PCB pad layout is found NOT matching the actual LCC component pads layout. some PCB pads is not covering enough land width to accomodate the intended covergae surface area, assuming SMT mounting of the LCC component is accurate with ref to the overall fudical marks.

F) however despite the controlled volume paste printed using , we still see gap between the solder fusion at the contact line to the BCC copper leadless terminal wall. the volume control gives good solder fillet at the contact region nearest to the PCB pads. only problem is now the gap "no fusion of solder" onto the copper walls.

G) we have adjusted the reflow profile to a longer reflow time(60-76sec) and higher peak temperature(212-215) deg-C.

H) recommened specs of paste supplier was violated only for the peak temperature adjustment to (216-218) deg-C to heat up the copper leadless terminal walls and observed that this gives good solder fusion onto the walls. gap unsolder is greatly minimized but we are going for 100% elimination. we tried another fluxing system from same paste maker but cannot get a smooth shinny solder fillet finish.

I) we discussed results with the customer. but they are asking us to try and get the wetting onto the leadless terminals (ie. no gap solder condition) as they have huge surplus stock of the LCC components. they also claim its a off-shelf part number, though we cannot find the LCC part no. in the supplier website database.

J) bottom line of this thread, based on the info given here, can anyone give some suggestions what we might be missing here? any experiences on LGA to share, that can help us out here?

reply »

ianchan

#19834

Micro Leadless Frame - solder wetting control | 8 May, 2002

Hi mates,

someone told us to use mild HCL acid 5% to 10% solution and submerge the LCC parts into this solution. hopefully this will clean off the varient copper oxide layer on the LCC leadless terminal wall.

after the HCL soak, submerge the parts into water(dilute HCL).

after water soak, submerge into IPA (IPA helps absorb water residue).

after IPA soak, oven dry bake for 10mins at 60deg-C.

anyone have opinions in this suggested method? before we will try it out later. hope can get some +ve results, and would appreciate any -ve technical reasoning on this method?

reply »


RDR

#19840

Micro Leadless Frame - solder wetting control | 9 May, 2002

One thing I have done in the past although it may not be feasable in your case, is to process through normal reflow and then saturate parts with a high solids water soluble flux and re-reflow with hot air, iron, or even another pass through the oven. Pretty hokie but it worked for me once in a low volume situation. I do find it somewhat hard to believe that the customer and the component supplier expect solder fillets to be formed to oxidized copper.

Russ

reply »


RDR

#19841

Micro Leadless Frame - solder wetting control | 9 May, 2002

One thing I have done in the past although it may not be feasable in your case, is to process through normal reflow and then saturate parts with a high solids water soluble flux and re-reflow with hot air, iron, or even another pass through the oven. Pretty hokie but it worked for me once in a low volume situation. I do find it somewhat hard to believe that the customer and the component supplier expect solder fillets to be formed to oxidized copper.

Russ

reply »

TomN

#19843

Micro Leadless Frame - solder wetting control | 9 May, 2002

I'd be a more than a little cautious about doing this. Had a somewhat different application once and found out that as the solution was cleaning/etching the surface, it also leached (thru the lead frame) into the body of the part. What we thought was a quick solution ended up becoming a component reliability problem in the field. What we had to do was tightly control 1)immersion time & 2)concentrate level, and add additonal rinse cycles. If your stuck doing this, I would recommend you run a DOE followed by component level validation tests. CYA !

reply »

#19856

Micro Leadless Frame - solder wetting control | 9 May, 2002

janchan, don't know how many LGA's/LCC/BCC's you are placing but a short term solution that I can recommend is the application of solder bumps to the component and than place it after board has gone through reflow or even at the same time.

Our company currently applies solder bumps to BCC's and LGA's that are used to rework boards that fall out during the process.

Good luck..

reply »

ianchan

#19858

Micro Leadless Frame - solder wetting control | 9 May, 2002

Hi Guys,

thanks to everyone for their help and advise!

we are running these LCC parts in the 1000's, so its going to pose a big problem to either: 1) tin-plate the oxide layer, 2) HCL away the oxide layer,

thanks for the option shown, dun think solder-bumping can help, coz the gap unsolder is reject by customer at the side walls of the leadless terminals, and not beneath the tin-plated LCC pads. customer dun allow recycle use of machine dropout LCC parts either, so our drop-out rate is controlled very tightly.

Russ idea looks interesting and its gonna take ages for the customer approval for any DOE results. Will look into situation further.

Cheers everyone!

reply »

Reflow Oven

One stop service for all SMT and PCB needs