Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

SMTnet
»
Electronics Forum
»
BGA Voids

BGA Voids

Subscribe to Thread

Hello, I'm having problems with voiding on an assembly th... - May 19, 2010 by patf

Since they don't effect the functionality or reliability of ... - May 20, 2010 by ScottE

Unfilled vias consistently produce higher quantities and lar... - May 20, 2010 by davef

Hi Pat According IPC-610D voids biger 25% are consider... - May 20, 2010 by Erli

Erli, Thanks for your feedback. The BGA hybrids were bak... - May 21, 2010 by patf

iv40, Thanks. The PCB that we are populating the BGA hyb... - May 21, 2010 by patf

Hi, Just have a look at the tech sheet on VIP, looks like... - Jun 04, 2010 by karthikthebest

patf

#61870

BGA Voids | 19 May, 2010

Hello,

I'm having problems with voiding on an assembly that has a GPS hybrid with balls much like a BGA. This is a RoHS assembly using no-clean solder (Kester EM907). The voids are at or above 25%.

I use an 8 zone oven where I peak in zone 7 and use zone 8 as an extended peak. I've tried RSS and RTS profiles with no change in the voids. I've tried following the profile recommended by the paste manufacture and profiles recommended by the GPS hybrid datasheet using an ECD Megamole 20.

I contacted the GPS hybrid manufacturer and got the following response:

"Since the soldering pads of the product are realized in Hole-In-Pad technology the origination of voids on balls is normal."

Has anyone heard of voids on balls being normal for Hole-in-Pad technology???

In the attached image notice the large voids. Also, notice the many separate voids in the balls at the bottom of the image. Have you ever seen anything like this?

We built assemblies for years with BGAs and have never had voids like this.

Thank you, Pat

Attachments:

reply »

ScottE

#61884

BGA Voids | 20 May, 2010

Since they don't effect the functionality or reliability of the product, I'd just ignore them.

reply »

davef

#61886

BGA Voids | 20 May, 2010

Unfilled vias consistently produce higher quantities and larger voids than filled vias, in some cases 5 times as much. [Effect of Filling Via-in-Pad on Voiding Rates in PWB Assembly for BGA Components; C Shea, R Raut, L Piccione; ALPHA -A Cookson Electronics Company]

reply »

Erli

#61893

BGA Voids | 20 May, 2010

Hi Pat According IPC-610D voids biger 25% are considered defects. You need verify you Pre Heat and try to reduce the ramp. Its very importante verify if the BGA`s are avoid hummity.

Erli

reply »

iv40

#61897

BGA Voids | 21 May, 2010

hi,

I have similar problem 2 year ago.BGA256 , hole in pad technology.I order PCB manifacturer to fill hole with solder mask.And now everything is ok.

regards

reply »

patf

#61904

BGA Voids | 21 May, 2010

Erli,

Thanks for your feedback. The BGA hybrids were baked before production. I had a particular date code that we dis-continued using. This date code was consistently voiding near or above 25%.

The other date codes still produced voiding on average around 12-16% void sizes which is a process indicator.

The manufacturer of the BGAS seems unconcerned if the voiding is under 25% citing the the "Hole-in-pad" issue.

All of the other assemblies that I build have no measurable voids. I do get a occasional void of around 5%.

Thanks...

reply »

patf

#61905

BGA Voids | 21 May, 2010

iv40,

Thanks. The PCB that we are populating the BGA hybrid onto does not have holes in the vias. I think the manufacturer indicated that on their hybrid it has holes in pads.

This hybrid is a PCB assembly with BGAs and other components populated under an RF metal can. The solder side of the hybrid has the BGA bumps that I reflow to attach to our PCB.

Here is a link to the hybrid look at the bottom of the page for Telit_GE865_quad.

http://www.telit.com/en/products/gsm-gprs.php?p_ac=show&p=47

Thanks...

reply »

karthikthebest

#61996

BGA Voids | 4 June, 2010

Hi,

Just have a look at the tech sheet on VIP, looks likes a continuing story until IPC defines the criteria in a better way

Attachments: