Subject: | |
From: | |
Reply To: | |
Date: | Mon, 24 Feb 2020 18:41:57 +0000 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Dear Victor,
Trying to understand your actual need behind your somehow odd question...
Would you be losing your balls one by one and trying to estimate how long you can last ?
;-)
Or maybe just surprised by the high number of unused BGA bumps or balls for a given component?
In my experience, some memory devices can be housed in large BGA packages with only à handful of balls actually used... In cases where the BGA ball matrix concentrates in the center of the BGA and is far away from the edges, there may be some unused balls for mechanical support.
Best regards,
Arnaud Grivon
Le 24 févr. 2020 18:33, "Stadem, Richard D" <[log in to unmask]> a écrit :
I have never heard of a design guideline in the ratio of unused balls to circuit-connected balls. Not sure why there should be a ratio from an electrical performance standpoint?
Solder balls that are not connected to the BGA circuity are used to balance out the package wetting forces during reflow, as well as to distribute CTE forces between the PWB and the BGA package. This is a BGA design characteristic, most likely any ratio would be in the component design guidelines in JEP95 or more specifically JEDEC Standard DR-4.14J.01. It may also be in some ANSI or other standard. Here is a link https://www.jedec.org/standards-documents/docs/dg-414h This is a free download after registration.
I don't have a copy, but somewhere in the JEDEC design standards for BGAs it should be listed, that is, if such a ratio actually exists.
I wish I could be of more help, but that much I know.
Odin
-----Original Message-----
From: TechNet <[log in to unmask]> On Behalf Of [log in to unmask]
Sent: Monday, February 24, 2020 11:02 AM
To: [log in to unmask]
Subject: [TN] Sacrificial solder bump ratio in Nand gate BGA device
Fellow TechNetters,
Is there an industry standard/guideline for the number of sacrificial solder bumps versus total solder bump count. For instance, 64 sacrificial and 88 active bumps for a total of 152 bumps (42%)?
In advance I thank you for your timely response in this matter.
Victor,
|
|
|