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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


LGA voiding

Views: 3747

Hello experts, I have a 50mmx50mm LGA on a PCBA. I Would li... - Dec 31, 2019 by Evtimov  

Here is the profile ... - Jan 14, 2020 by Evtimov  

#84074

LGA voiding | 31 December, 2019

Hello experts,

I have a 50mmx50mm LGA on a PCBA. I Would like to decrease voiding in these joints and I wanted to hear about your techniques to decrease voiding. Stencil is currently 5mil. SAC305. ENIG finish.

Attachments:

reply »

#84075

LGA voiding | 31 December, 2019

Assuming you've already looked at optimizing your profile and stencil design, have you looked at metal load in the paste?

I hadn't, but just read that increased metal load (associated with reduced powder size, i.e., moving from T3 to T4) has been shown to correlate with increased voiding. If you're using T4, can you switch to T3?

reply »

#84088

LGA voiding | 2 January, 2020

thank you Steve,

when you mention Stencil design, what do you refer to? How about the Reflow profile? Any specifics that I should look for. Everything is in my process window, but some detail towards this package might be very helpful.

reply »

#84090

LGA voiding | 2 January, 2020

As far as stencil design for LGA's I was thinking in terms of verifying that you've used the same aperture design criteria that you would normally use for LGA's so you know you didn't introduce some new variable.

I've made changes from my norm based on datasheet suggestions (thinking "who knows better than the manufacturer?") and had disastrous results.

With respect to profile, mostly just making sure you give the flux time to outgas before the metal freezes (without introducing wetting issues by boiling off the flux too soon.

You don't have any unplugged vias in the pads, do you?

reply »

#84091

LGA voiding | 2 January, 2020

Hello,

Can you give more detail on LGA parts in question? is 50 mm x 50 mm is overall size of the parts? What about pad pitch and how big is the pads individually? I assume that 50x50 mm is overall size. How about the stencil opening for the LGA pads, is it circular opening, or square radius? What is the opening size?

If my assumption is correct, then you may encountered insufficient solder release on the LGA pads. Usually for fine pitch BGA/LGA, 4 mils stencil is usually used, depends on pitch and size of the pad itself. That's probably the cause of voiding.

reply »

#84093

LGA voiding | 3 January, 2020

What's your target for voiding? I cannot see it violate 25% of solder joint area. Process setting and PCB design can impact the voiding, many factors, such as PCB pad, stencil aperture and thickness, re-flow profile, solder paste, etc.

reply »

#84095

LGA voiding | 3 January, 2020

Some of the voids are questioned and they might easily be over 25%.

reply »

#84096

LGA voiding | 3 January, 2020

Zack,

circles 36mil copper 35mil circle stencil aperture. Pitch is 80mil. Stencil is 5mil and apertures are spread 46mm under the 50mm component size. In my mind this might be too much paste. I am afraid that if I print 4mil I might get open joints at the ends due to the 50x50mm part size might bend in reflow.

reply »

#84098

LGA voiding | 3 January, 2020

Besides correct stencil design, solder paste chemistry can have a profound effect on voiding.

People will tell you reflow profiles. I have never in my career seen a reflow profile have any effect on voiding unless it is grossly off.

reply »

#84099

LGA voiding | 3 January, 2020

We use low voiding formula from Indium.

reply »

#84106

LGA voiding | 5 January, 2020

Yes, you may need tell us how the PCB designed, how it re-flow(even it's Indium voiding formula), how stencil is.

The void ratio means area ratio, not diameter ratio, so you need calculate it. If you have high ratio void, show its picture.

reply »


SMTA-Norah

#84135

LGA voiding | 7 January, 2020

Were the boards baked out before soldering?

Is there plated over vias in the pads?

This message was posted via the Electronics Forum @ SMTASMTA

reply »

#84140

LGA voiding | 8 January, 2020

We don't have plated vias on these pads.

reply »

#84162

LGA voiding | 10 January, 2020

Hello,

When I calculate based on the data you gave, your stencil aperture area compared to pad area is 94.52%. Your aperture area ratio is 1.76 using 5 mils thickness foil. OK, your stencil and your paste release is not the culprit here. Your problem is somewhere else, it is possible that the reflow oven setting need to further optimized.

reply »

#84167

LGA voiding | 10 January, 2020

Zack,

would you recommend longer TAL or higher peak temp or both together?

reply »

#84170

LGA voiding | 11 January, 2020

Hi Evtimov,

Could you share your current reflow profile setting, and also your part technical datasheet on its heat tolerance? It might as well help other engineers to review and give their opinions.

reply »

#84185

LGA voiding | 14 January, 2020

If you are using standard alloy and flux types and you are within the ballpark of 30-90 seconds above liquidous you should be fine. As I said above unless your profile is grossly off, it is not the cause of your voids.

reply »

#84186

LGA voiding | 14 January, 2020

Here is the profile

Attachments:

reply »

#84187

LGA voiding | 14 January, 2020

Your profile looks textbook.

reply »

#84188

LGA voiding | 14 January, 2020

Still considering going from 5mil to 4mil stencil. Concern is the 50mm part might have issues with contact on the outside rows.

reply »

#84203

LGA voiding | 17 January, 2020

Yeah, that's a nearly uniformed ramp up profile, I can't see anything's wrong with that.

reply »

Dual Lane Reflow Oven

Reflow Oven