BGA component removal and replacement are required during product development and repair operations. In general,a single rework on specific location is permitted without causing excessive damage to solder joints,board materials,and adjacent components. However,situations exist where a multiple rework is required on the same location. Multiple reworks could induce a thick intermetallic layer between the BGA solder balls and Cu pads on the boards,leading to reliability concern [1,2]. This paper examines the possible effects of rework cycles on bonding strength of solder balls,Cu pads and laminates. An effort was made to establish a sound scientific relationship among rework processes (cycle times),intermetallic thickness,and mechanical strength of BGA balls,as well as any possible intermetallic embrittlement. It is believed that these
results can help shed some insight on the maximum permitted limits of BGA rework cycles.

As companies transition over to lead-free assembly a certain amount of hand-soldering and rework will be performed. An article from Tech Search International last year did state that in Asia where lead-free soldering is much more common,handsoldering was more of a production problem than lead-free SMT or wave soldering or other soldering processes. In fact most user problem calls and requests for training these days are related to lead-free hand-soldering and rework. In many cases the assemblers are using materials from various solder suppliers with similar issues occurring in all cases. Often the problems are more than just material issues.
Switching to lead-free without proper preparation is not recommended. Although this seems easily understood some assemblers have attempted to transition without adequate training; resulting in line stoppages occurring only a few hours into lead-free hand-soldering. Operator complaints,loss of reliability and poor joint quality were experienced. This could be a production engineer’s nightmare but it need not be this way if the basic concepts of hand-soldering are revisited,some experience gained prior to the transition and adequate training of operators is performed before and after the switchover.
This paper is a compilation of questions often asked by assemblers in reference to hand-soldering with lead-free. These questions are in fact some of the issues addressed during lead-free hand-soldering on-site audits or training at assemblers using lead-free solders at their facilities. It offers practical answers as to enable the reliable implementation of lead-free.

The use of tacky fluxes is common throughout the industry in applications such as ball attach,BGA repair and hand soldering. These applications employ different heating profiles,meaning that the fluxes are required to endure a wide range of time and temperature conditions,while not compromising long-term reliability.
Tacky fluxes are generally made from the same types of materials that comprise standard solder paste products designed for typical SMT applications. Therefore,the processability and reliability of tacky fluxes that are subjected to a standard SMT reflow profile are well understood. However,when the same tacky flux is subjected to a shorter heating cycle,such as a hand soldering application,it is not necessarily known if the flux residues will have the same reliability as expected when subjected to the typical SMT reflow profile. This paper examines the long-term reliability of no-clean tacky fluxes when subjected to a variety of
processing conditions.

This paper describes a new lead-free SMT rework process which avoids component overheating,reduces board warpage and cratering,and prevents adjacent components from thermal damage. In the replacement of a BGA,on an organic substrate printed circuit board assembly,a solder alloy with the melting range lower than Sn-Ag-Cu solder is used. The lead-free solder pastes analyzed in this work include two Indium containing alloys,melting range 181-187°C and 138°C respectively,and a Bismuth containing alloy,melting range 195°-209°C. When the solder alloy melts,dissolution of the Sn-Ag-Cu balls occurs. The resultant composition and microstructure of the solder joints provide good mechanical and electrical properties with high fatigue life,while avoiding high temperature stresses on component bodies.
In this work CBGA937,PBGA196,and CSP46 components with SAC405 balls were assembled using SAC405 solder. Then rework was performed with three compositions of low melt solders using different profiles. The microstructures of mixed solder joints were analyzed using optical and scanning electron microscopy methods. Two solder pastes and rework profiles were chosen for thermal cycling. A limited ATC study was done in a temperature range of 0 -100°C. It was observed that fatigue life was component-dependant and when fully mixed,solder joints had better or equal reliability to that of the pure SAC405 assemblies.

In manufacturing flexible and rigid-flex boards,the metallization step using electroless copper poses a major challenge,namely the leaching of polyimide base material into the plating bath. Leaching is exacerbated by the combination of high alkalinity and extended dwell time in the bath during the plating cycle. The result of excessive leaching is the degradation of the quality of the plated deposit leading to compromised adhesion. This is presently overcome by increasing the renewal frequency of the electroless copper plating solution,increasing the demands on bath control as well as the overall cost.
Our process introduced in this paper forms a thin conductive Palladium film on the dielectric,eliminating the need for electroless copper deposition step. The conductive film readily accepts copper electro-plating and yields an adherent
deposit.
The process requires a short immersion time in an alkaline solution,alleviating any damage to the resin and minimizing the accumulation of leached byproducts. The palladium is specific to the dielectric resin and does not
adsorb to the copper surface eliminating the need for copper etching before electroplating. The treatment time is short leading to improved productivity. The process is stable and easy to control,resulting in enhanced adhesion and
increased inner layer connection reliability.

Jobs are becoming more difficult to complete to specifications as the complexity of parts increases. With electronic power components getting more advanced everyday reverse pulse technology is offering a great solution in today’s PCB shops. From five years ago or even one year ago there have been many improvements in reverse pulse power supplies. We are seeing very high quality pulses with repeatability. Some manufacturers are integrating intelligent pulse optimization systems into their power supplies. All of these functions increase the reverse pulse process from 30% to 40% better than before.
Prices of these units becoming more affordable and make reverse pulse an even more attractive option to complete those hard to do jobs. With the latest developments in user interface and IPOS in reverse pulse power supplies this integration into your facilities makes an easy transition from your old system to a new system. With today’s reverse pulse power supply technology and our global economy could reverse pulse be one of the solutions for your shop? The constant improvement of electronic components has resulted in many advantages for pulse and pulse reverse power supplies. Some advantages include output currents getting higher to meet the larger scale production facilities needs. The
integration of intelligent pulse optimization systems imbedded into the core software. This enables us to see pulse shape monitoring to achieve repeatability and self optimizing control loops to adapt to bath characteristics. All of these functions give a repeatable waveform and the ability to repeat the process and adjust for changing bath conditions. When working with a delta I 10000A with a slew rate of 100us or less,inductance must be addressed in order to maintain a repeatable pulse form. In order to understand the problem a little better inductance is defined as the following. “The property of a circuit or circuit
element that opposes a change in current flow,thus causing current changes to lag behind voltage changes. It is measured in henrys.” With this in mind there has to be consideration on how to compensate for this in the control of the power supply. IPOS and the latest designed power supplies are using separate voltage and current loops simultaneously this keeping the over and under shoot to a minimum. The next thing to be considered is the connection between the power supply and the plating cell. Depending on the current and the placement of the rectifier either coax cable or twisted pair should be considered. If the distance is not that significant twisted pair would be sufficient. If the distance is significant coax cable should be considered. Coax cable has a lower inductance compared too twisted pair. Most rectifier suppliers should have two types of cooling available when choosing reverse pulse. The first type is air cooled. The size of the rectifier will determine whether the unit will be convection or forced air cooled. The other option will be water cooled. Things to consider when choosing air cooling are the environment it will be placed in,ambient temperature,and cable length. If choosing water cooled,you are able to place the power supply right next to the plating cell to achieve lower inductance. However,water temperature and water quality needs to be considered.

The immersion silver process has become the first choice of final finishing for many original equipment manufacturers (OEM) because of its shortened process flow,good conductivity,even deposition,good solderability and bonding function,and suitability for vertical and horizontal-automated production lines. Some popular immersion silver solutions on the market contain acid that will attack copper to cause circuit break and trace undercutting. The acidity of these processes limit immersion times for silver deposition that can in turn lead to no deposition in the bottoms of blind vias. These processes also frequently use inhibitors or penetrating agents that can contribute to massive voiding and brittle solder deposits. This in turn leads to decreased solder strength.
The shortcomings of these acidic systems are overcome using a mildly alkaline immersion silver. This process has a pH greater than 7 making it possible for longer dwell times that ensures a complete deposition in blind vias. The new process does not contain inhibitors so that the deposited silver possesses higher purity,higher corrosion resistance,lower contact resistance,high soldering strength,and no electromigration. This technology is fully compliant for RoHS and WEEE regulations.

It has been the conventional wisdom of the industry that the predominate board parameter associated with through hole plating is the aspect ratio of the through hole. While this is intuitively understandable,very little has been done to analytically verify the assertion. This paper will explore this hypothesis by developing a first principles flow model using the Navier-Stokes analog for viscous fluids. The first concern is to define the proper flow regime,laminar or turbulent based on the Reynolds Number associated with the process. Once this is established,a proper mathematical model can then be identified. At that point,the governing equations are parameterized and the crucial parameters identified. Faraday’s Law will then be folded into the model to define an algebraic model for electrolytic plating. Finally,an order of magnitude analysis is presented to determine the relative importance of the critical parameters.

Many techniques have been developed to embed capacitors in printed wiring boards (PWBs) these days. The simplest way may be embedding chip capacitors in PWBs. However,since their thickness is over 200 m,the PWBs must be much
thicker. In addition,prepregs or semi-cured insulators for lamination should be prepared with cavities for embedding them.
Polymer composite materials with high-permittivity ceramic filler are a second candidate for embedding capacitors. However this technique needs large area to make decoupling capacitors because of their low capacitance. The best solution
should be embedding thin film capacitors having large capacitance. Now we have developed a novel dielectric insulator and thin film capacitors with the insulator. The capacitor has the structure of metal / insulator / metal (MIM). Both the metal electrodes are copper,and the thin insulator has a high permittivity for large capacitance. The total thickness,including the electrodes and the thin insulator,is less than 50 m; therefore,we need not use prepregs with cavities for capacitor embedding,and the PWB can be kept thin. We also evaluated the electric properties of the thin film capacitor. The capacitance density was 1 nF/mm2 at 1 MHz. The resonant frequency was higher than that of the chip capacitor with the same capacitance. The leakage current was about 1 nA/mm2 at 4 V. We also proposed the process for embedding the thin film capacitors in PWBs.

The global electronics industry is exhibiting a widespread and growing interest in the technology of embedded passives. This interest can be attributed to three primary benefits. First,embedded passives have far less parasitic inductance than discrete components,which enables electrical performance advantages in high speed digital applications. Second,embedding saves surface real estate,which allows for board size reductions. Third,the incremental cost of embedding additional passive components is typically negligible,which suggests the potential for system cost reduction in designs with high passive component counts.
The material we describe here is a high performance embedded passive material intended for embedded capacitor applications. It is a copper-clad laminate which utilizes an ultra-thin,high K-value dielectric material between the copper planes to deliver a capacitance density of over 6 nF/in2.
This presentation will focus on this laminate material and how it can be used to improve electrical performance and reduce board size (by replacing discrete capacitors),and will also address the impact on PCB reliability and system cost.