Area Array Component packages (BGA, CCGA, Flip Chip, Chip Scale Packages etc.) have become more and more prevalent in electronic circuits. Although they offer invaluable savings in real estate usage on the boards they present limited inspection opportunities. Optical inspection is limited to peripheral joints and hence all the body of standards for the acceptability of solder joints (IPC-610) have limited application here. We are therefore forced to use other media to penetrate and make “visible” the joints hidden from direct view. X-Ray inspection is promising to be a very useful tool in this area. However, the interpretation of the X-Ray images and correlation of the images to the quality of the solder joints is an area which is still under development.

Measurements and Investigations

In the current report we focus on the size of the circular X-ray images of the ball grid arrays. The investigation was carried out using the following:

  1. Intel 304 Pin PBGA
  2. HASL finish PCB with 0.625mm BGA land Pattern
  3. Glenbrook X-Ray Machine
  4. Eutectic Solder Paste

The BGA balls are generally spherical in shape before reflow. This study focuses on ball grid arrays with all eutectic solder bumps. During the reflow the solder balls and the solder paste deposited on the pads merge and form ellipsoids with volume equal to the sum of the solder ball volume plus the volume of the solder in the solder paste deposit.


During the reflow stage the weight of the component pushes down on the solder balls on which it is floating thus “pushing the sides out” and changing their shapes to ellipsoid. The diameter of the ellipsoid (2D) is dependent on the weight of the part (W) and the sum of the volume of the original ball (Vb) and the paste deposit (Vp) .

D = f (W,Vb,Vp)

The dependence of D on W was investigated by placing different weights on top of the BGA and reflowing  it on the test PCB. By using the same component each time and printing paste with the same stencil the other two parameters (Vb) and (Vp) were kept constant during the  trials.

The PBGA weighed about 3.2 grams. The ellipsoid diameter “2D” was measured for different weights of up to 350 grams placed on top of the part. The results are shown in table 1 below.

Component Weight in gr “W” Ellipsoid Diameter in mm “2D”
3.2 0.98
6.4 0.99
9.6 1.00
12.8 1.00
19.8 1.00
23.0 1.01
62.5 1.04
198.2 1.10
353.2 1.15

Table 1 The effect of component weight on the ellipsoid diameter

As the table indicates the effect of part weight on the reflowed ellipsoid diameter is very small. The ellipsoid diameter increases by about 3% when the weight of the part is increased ten fold. This result is in line with the large surface tension of the molten eutectic solder and the large number of pins on which the pressure of the component weight is distributed. We can therefore ignore the effect of the parameter W and restate our ellipsoid diameter relationship as follows:

D = f (Vb,Vp)

Now since


Where “q” is the fraction of solder in a given volume of solder paste after the evaporation of the flux. “q” is typically in the 0.3 to 0.5 range. Ve is the volume and H is the height axis radius of the ellipsoid.



As the terms (1H) and  (34qh)  have the same orders of magnitude then the dependence of D on R and r (that is solder deposit volume and original ball size of the part) is significant.

This dependence was verified by the following experimentation. Two trials were conducted and measurements of the X-ray image were made. One with no solder deposit on the PCB (that is r = 0) and the second one with a deposit of 0.15mm thickness of paste through a circular stencil aperture of diameter 0.6 mm.

Solder Deposit on Pad ( “r”) Ellipsoid image diameter ( “2D”)
0 0.91 mm
0.3 mm 0.98 mm

Table 2 The effect of solder paste stencil aperture radius on reflowed ellipsoid diameter

X-Ray Image Diameter Characterization

Above observations will now be used to describe a method for evaluation the BGA solder joint using the radius of the reflowed ellipsoid. The defects that would be identified by this method are:

  1. No reflow
  2. Blocked stencil aperture
  3. Insufficient paste deposit
  4. “Dishing” of the component. That is development of curvature in the component body leading to loss of co-planarity in the ball grid.

Each of the above mentioned conditions leads to a well defined range of values for the observed X-ray image of the ellipsoid diameter. For example, in case of no reflow the observed ellipsoid diameter would equal to original ball diameter. That is in our example:

2D = 2R = 0.80 mm

For blocked aperture where there is no paste deposit, r = 0 and the measured ellipsoid diameter would be:

2D = 0.91 mm

For good reflow (target condition) the observed ellipsoid diameter would be:

2D = 0.98 mm

Using above data, we now devise a procedure for characterization of the X-Ray image. Depending on the diameter of the X-Ray image five different zones can be defined. Each zone points to a particular aspect of the solder joint (Table 3).

Zone “-3” zone “-2” zone “-1” zone “0” Target zone “+1” zone
Ellipsoid diameter in mm 0.80 0.91 0.95 0.98 1.10
Solder joint condition No reflow Blocked paste stencil aperture Insufficient paste deposit Good reflow Dishing

Table 3 Zone definition

Depending on where the X-Ray image of the joint falls within the zone determines its characteristic. The different zones are shown in Fig 1.

If the X-ray image of the ellipsoid has a diameter of 0.8mm or less, then the image lies in Zone “-3” and we have a non-reflowed junction and the oven profile needs to be re-adjusted. If the ellipsoid diameter value is between 0.8 and 0.91 then we are in zone “-2” indicating a blocked paste screen aperture – the screen needs to be cleaned. If the value is between 0.91 and 0.95 we are in zone “-1” which indicates that the deposited solder paste volume is not sufficient. The paste printing process needs to be re-examined as many factors including defective screen apertures, inadequate screen cleaning, printing parameters, paste kneading procedures etc. could lead to this condition. A value of around 0.98 would place the joint in the target zone “0 “indicating good reflow. Values of 1.1 mm and higher would be in the zone “+1” indicating the occurrence of “dishing” in the vicinity of the solder joint which could be caused by inadequate storage or insufficient pre-baking of the part.

Ellipsoid Diameter

  • Ellipsoid Diameter (mm)

The actual values for different zones for each part depends on paste screen (that is the volume of the paste deposit) and the ball size on the BGA. These values can either be calculated from the formula (1) above or they can be measured by reflowing the part first with no paste (Zone “-3”) and then with paste deposit (Zone “0”) as performed in our example. The values for the other zones can then be extrapolated from the two measurements.

If a sophisticated X-Ray machine is not available, then the diameter of the X-ray image can be measured simply by placing a reference metallic gauge on top or adjacent to the BGA. For example, if a pin gauge of diameter 1mm is placed along the side of the BGA then the actual diameter of any ball in the image is determined by ratio of the measured image diameter of the ball and the image diameter of the pin. This comparative method reduces the measurement errors that may be introduced by “half-shadow” regions of the image.


The actual diameter of the ellipsoid formed after reflow under an area array component provides valuable information for solder joint quality assessment.  In conjunction with visual inspection methods and other X-ray signature identifiers, the method proposed in this article can provide indicators for process development and serve as a tool for solder joint quality evaluation during production.  Although an advanced X-ray imaging system would definitely be an asset, the proposed process can be implemented using an entry level X-ray machine.


Circuits Central R&D Team

Electronics Manufacturing
toll-free: 1 (888) 568-6550