Posted by Balapmurugan: Fri Nov 10, 2006 1:27 pm    Post subject: Board level drop test failure

Hi,

We encountered board level drop test failure in 16 ld QFN device (Body size - 4 * 1.6mm). The test conducted as per procedure Jesd22-B111. Failure rate - around 50%. Failure mode - Open in any one of the corner pins.

However when the units are gently pressed against the test contact, (Both in handler & bench verification) the same is passing. X-ray does not reveal any abnormalities. However C-Sam pictures reveal lead delamination.

What are the likely causes for this intermittent behaviour of the device. It is quite difficult to decap this device on the second bond side.

Any idea to overcome the failures in the drop test. Process control measures that need to be put in place to prevent recurrence of this defect.

Drop test is done as the units have application in portable electronic gadgets and to simulate the drops of the final end product.

Posted by FARel Engr: Sun Nov 12, 2006 12:52 pm    Post subject:

Hi Balapmurugan,

Based on your description of the failure mode, it looks like that your units have suffered a second bond heel break during the drop test. They can still pass when you apply pressure on the leads because this pressure makes the broken ends come into contact with each other again.

The delam at the second bonds certainly needs to be addressed, but another possible factor in this is the quality of your second bond heels. If they're too thin due to overbonding, then they can easily break. You really need to decap the second bond to analyze this properly - this will allow you to determine why your second bond heels are so vulnerable.

FARel Engr

Posted by Balapmurugan: Mon Nov 13, 2006 8:38 pm    Post subject:

Hi FA Engr.,

Yes, you are right !

We did the decap and observed heel crack in one of the corner pins. However the real cause for this is not known - Whether the heel crack was induced by lead delam.

As the units mounted on the PCB have undergone the mechanical stress due to drop test, (1500g force, sinepulse, 0.5msec duration, 30 times) the delamination has been induced on the leads.

Any suggestions to overcome the lead delamination and improve the wire strength to withstand the mechnical stresses induced in the test.

Regards,

Bala.

Posted by Logan218142: Sat Dec 16, 2006 5:30 pm    Post subject:

you may want to look into a better interlocking design of leads.

Is your package a punch VQFN?

Posted by Balapmurugan: Mon Dec 18, 2006 11:39 am    Post subject:

Hi,

We are exactly working towards the same.

Also we increased the GW dia from 0.9 mil to 1 Mil and changed the cap.face angle from 8 deg. to 11 deg and the units passed the drop test. (though there was not much difference in the Avg wire pull strength).

I am not sure if the excessive gold underneath the heel in the cross sectional area helped to withstand the mechanical stress.

It is a Saw type QFN.

Bala.

Posted by Mike_balbuena: Tue Dec 19, 2006 1:56 am    Post subject: delamination

How's the intensity of your delamination before and after test? It would be nice to see how good your molding compound adhesion vs the leadframe.
The delam could be inducing stress on your heel as well. If you use copper leadframe they have a very bad reputation with respect to molding compond adhesion due to oxidation. To improve your molding capability you might want to change compounds with better coupling agent and high filler content.
Just a thought...

Regards,
Mike B.

Posted by Balapmurugan: Tue Dec 19, 2006 11:25 am    Post subject:

Mike,

Even after the assembly without any stress, we notice minor lead delamintion and that's why we are working on improving the leadframe design by providing lead locking hole in the lead finger.

Mold compound type is fixed by customers. We noticed improvement after plasma cleaning the parts.

Bala.