|
|
|
Why Are There So Many Chip-Scale Package Ball Diameters?
|
 |
Mark Bird Contributing Editor |
It seems that with each new CSP package family (or variation of a present family), a different ball diameter is defined.
This subject, in fact, was an agenda item assigned to the JEDEC Committee at a joint EIAJ/JEDEC meeting held in Portland, Ore., last year, where JEDEC explained why there are different ball diameters for various pitches in its CSP specifications. (The JEDEC presentation is shown in the table.)
The table is a summary of all present JEDEC chip-scale package outlines, including all JEDEC Design Guides that have different eutectic ball sizes defined.
The nominal pitches covered are 0.50, 0.65, 0.75, 0.80 and 1.00 mm, while the eutectic ball nominal diameters covered are 0.60, 0.50, 0.45, 0.40 and 0.30 mm (each of these has a tolerance of ± 0.05 mm).
All specifications listed in the table are JEDEC Design Guide Standards (95-1 Sec. X) or Package Outlines (Registrations MO-XXX and Standards MS-XXX). The specifications are available, without charge, from the JEDEC web site [jedec.org].
The table illustrates that as the lead pitch becomes finer, the number of different options (in terms of ball size) are reduced. This is because the ball's physical size is limited by the lead pitch to prevent shorting during CSP mounting. Thus, at the finest pitch of 0.50 mm, only one ball diameter option is recommended (0.30 mm). At the 0.80 mm pitch, however, four different sizes can be employed (0.30, 0.40, 0.45 and 0.50 mm).
|
CSP Ball Diameter Vs. Pitch
|
| |
Ball Diameter |
|
|
| Pitch |
Min. |
Nominal |
Max. |
Co-Planarity |
Specifications |
| 1.00 |
0.50 |
0.60 |
0.70 |
0.20 |
MO-151, MO-192, MS-028, 95-1 Sec 14 |
| |
0.40 |
0.45 |
0.50 |
0.12 |
MO-216 |
| 0.80 |
0.45 |
0.50 |
0.55 |
0.12 |
MO-205, 95-1 Sec 5, 95-1 Sec 6 |
| |
0.40 |
0.45 |
0.50 |
0.12 |
MO-216, MO-219, 95-1 Sec 5, 95-1 Sec 6, 95-1 Sec 7 |
| |
0.35 |
0.40 |
0.45 |
0.10 |
MO-210, 95-1 Sec 5, 95-1 Sec 6, 95-1 Sec 7 |
| |
0.25 |
0.30 |
0.35 |
0.08 |
95-1 Sec 4, 95-1 Sec 5, 95-1 Sec 7 |
| 0.75 |
0.40 |
0.45 |
0.50 |
0.12 |
MO-207, 95-1 Sec 7 |
| |
0.25 |
0.30 |
0.35 |
0.08 |
MO-207, 95-1 Sec 7 |
| 0.65 |
0.35 |
0.40 |
0.45 |
0.10 |
MO-207, 95-1 Sec 5, 95-1 Sec 6, 95-1 Sec 7 |
| |
0.25 |
0.30 |
0.35 |
0.08 |
MO-207, 95-1 Sec 5, 95-1 Sec 6, 95-1 Sec 7 |
| 0.50 |
0.25 |
0.30 |
0.35 |
0.08 |
MO-195, MO-207, 95-1 Sec 5, 95-1 Sec 6, 95-1 Sec 7 |
Ball Diameter Drives Co-Planarity
The summary table indicates that the eutectic ball diameter, not the pitch, drives the co-planarity requirement-something that is not normally seen when looking at a specification. The larger the ball diameter, the larger the volume of solder needed, and the greater the lack of co-planarity tolerable in board assembly.
The major consideration, not obvious in the table, is the end user's package overall profile height requirement. Package overall profile height is defined as the distance from the bottom of the ball (seating plane of the package) to the highest point of the package.
The overall profile height is specified as a maximum value, and is normally driven by the end customer's application require-ments. The lower the profile height needed, the smaller the ball diameter must be.
Due to the low profile height requirements of portable products, many JEDEC package outlines have been created with the same body sizes, pitches and ball counts, although they possess different ball diameters.
In general, the largest ball diameter possible should always be selected to provide the best solder joint reliability for the overall profile height and pitch requirements.
Obviously, even with the JEDEC CSP Design Guides in place, there will be (and are) ball diameters that differ from those found in the table. The many possible variations will continue to defy total standardization.
While it may not be readily apparent, a number of standard ball diameters have been adopted for use by the industry. Moreover, as end users' overall profile height requirements continue to become smaller, CSP ball diameter variations will necessarily follow the same trend.
|
Mr. Bird is an Amkor Technology Fellow and the director of technical marketing at Amkor in Chandler, Ariz. [mbird@amkor.com]
|
|
|