Electronic component placement

What is the general trend of chip component technology? This problem is actually quite simple: the size of the chip components and their packaging is getting smaller. In fact, the 0603 packaging form (side length: 1.5 x 0.75 mm) is the mainstream. And the 0402 patch (side length: 1.0 x 0.5 mm) ) Gradually more common, especially in the telecommunications industry.
However, most industries have not benefited from the technical advantages of smaller chip components. A simple reason is that suitable patch machines are not readily available.

Progress in matrix packaging:
SMT will naturally be placed in the ever-increasing use of complex packaging, such as BGA, m BGA, chip scale packaging (CSP) and flip chip. These matrix components greatly increase the number of outputs, and the traditional type The dense pin (0.4mm or less) QFP is ten times better than TSOP in some cases. In the era of reduced space, flip chip can ultimately provide considerable cost savings on the board. Matrix packaging is superior to dense QFP and TSOP because they have the ability to self-position, which is a feature that greatly increases placement accuracy.

BGA is the fastest growing new IC package, and its expansion speed is not expected to decrease in the 21st century. m BGA and CSP are also becoming mainstream, although a little slower. Price is a concern, the same reliability is also (although Japan The manufacturer has achieved some success in replacing the TAB component with CSP).
In the field of bare chips, flip chip and COB have the highest annual growth rate, and it is expected that in 2000, flip chip will grow to more than 2 billion units.

As for the relevance of the development of SMT patches, due to the smaller size and various matrix packaging, in addition to their benefits, there are problems with yield and output; production equipment for hybrid technology assembly and traditional type Both SMDs must be able to be completed with high-speed patches to achieve a reasonable cost. Reflecting this, market requirements are already shifting away from the traditional chip-based chip shooter technology, because such devices are not easy to catch up. New advanced packaging requirements.

Patch approaching special market
This is not to say that the chip-like component high-speed machine has lost its position. (It is important to remember that although advanced packaging is becoming mainstream, it is by no means dominated by all markets.) No matter how clear, the placement machine Industry is facing product challenges in four different markets: flexibility and tightness, high speed, high speed flip chip and ultra high speed. Each has its own processing requirements, each requiring a special placement machine to consider factor.

There is no doubt that the chip component high-speed machine is successful in high speed and ultra-high speed. When the component mixing is limited and the advanced matrix packaging is used to the minimum, it is assumed that the component spacing is 0.8mm or more, and the chip component is high speed. The machine achieves a high placement rate (40,000 cph or more). However, the feeding method is critical for the placement process and is limited on high speed machines because these machines cannot take components or adhesive tape from the matrix tray. Take the bare chip.

Even at high speeds, high speed machines can be a source of quality problems. Because their basic structure requires the board and feeder to move to the placement position, accuracy is compromised (the components may move as the board moves) In addition, for the application of dense-foot components or flip-chips, high-speed machines cannot achieve the necessary placement accuracy due to their original design, which is required for cost-effective circuit assembly.

For flexible/tight placement, two head configurations are necessary. Large components are handled by pick and place heads; smaller or special components such as CSP or m BGA are a fast spin. The range of patch heads allows for greater throughput. A flexible system will allow the patch head configuration to be quickly adjusted to meet changing requirements.

For flip-chip applications, high-speed machines can't be satisfied. Flip-chip applications require flux processing. These machines cannot add this feature to the turret-type patch head. Precise flip-chip placement also requires an advanced The visual system, it is difficult to add it on many high speed machines.

New generation SMT equipment
In response to the demand for advanced, high-speed flip-chip and ultra-high-speed circuit assembly, traditional high-speed machines are being replaced by newer generation machines. Some are derivatives from old equipment, but have greatly improved The ability to abandon the chip component high-speed machine, using a completely new approach. It has become apparent that the PCB itself is no longer an option. When the board is moving, smaller component packaging is likely Will shift, yield is compromised. One solution is to remove the horizontal patch head and replace it with a vertically rotating patch head or a patch head on an X/Y arch. In this configuration, the board is still stationary. The nozzle head randomly picks up the components. The patch head moves toward the board, instead of the opposite, the static feeder and the PCB are combined, resulting in a larger yield regardless of the component type.

In theory, the use of multiple nozzles allows the system to process many components "in flight". It can be packaged in odd-shaped packages (joints, transformers, etc.) and, if necessary, the nozzle can be designed to fit Special component packaging. However, some suppliers suspect this approach, preferring to change the placement head to accommodate component mixing. This is a viable option, and in some cases, the turret tip can be prioritized, but if In other words, it inevitably increases machine downtime. In addition, the head change may require recalibration and increase machine downtime. Component mix, frequency of necessary changes, and type of component to be mounted will determine which system type It makes sense. There is no single industrial solution.

Vision system is the key
In order to accurately place advanced surface mount components, it also relies on the use of complex vision systems. As the components evolve, the vision system must also adapt and improve. Mechanical positioning, once a necessary unit for electronic assembly, Then there is a viable treatment. Today, equipment manufacturers do not use an optical, camera-based system, or use a laser positioning system. Both have advantages, of course, there are differences.

Laser positioning allows for "in-flight" correction, the ability to handle all shapes and sizes of components, and the ability to accurately determine component position and orientation. However, even the most complex laser systems cannot measure pin and pin pitch.

The camera is able to. This is why suppliers still rely on camera positioning systems. And, as newer packages get on the production line, this preference has not changed. Lighting is the key to these systems, and today's placement systems use lighting technology. Combination. Backlighting, or illuminating components from above, and analyzing shadow images, is used for contour center positioning. For traditional SMD, this method works well. But for advanced components, backlighting lacks component packaging. Point or tin ball point image recognition ability. To solve this problem, laser illumination is used.

In operation, the component rotates in the beam of the laser. (Usually, this side illumination is combined with the pick-and-place head), the solder ball positioning of the soldered BGA, m BGA and flip chip, front light illumination, and Non-backlit lighting is necessary. To differentiate the solder balls, many systems use combined illumination; the components are illuminated from various angles so that the solder balls protrude from the background. In fact, multiple light sources allow for programmatic control so that each light source reaches each Ideal lighting for components.

Accelerate SMD assembly, dual channel better than single channel
Price pressure on all products is of course one of the necessary drivers of electronics. Think about personal computers, mobile phones or car stereos, and their steady growth. OEMs are usually only 6 to 12 months old. After that, they took out their new products, and their prices sometimes did not reach their previous things. It was followed that the faster the PCB was assembled accurately, the lower the final cost and the greater the profit.

Today, there are two techniques for passing the PCB through the production line at a faster pace. The first is to treat both sides of the board either in a pure solder reflow process or through a cure/reflux process of the SMD. The second method is Two different boards are assembled on one patch line to double the output.

Equipment suppliers work hard, develop better, faster methods, and place components precisely and with minimal production time per machine. In addition to faster placement heads, good feeders, and vision systems, improvements How the PCB moves through the machine is a method.
Improved conveyor technology is feasible; no matter how quickly the placement head posts components, high-speed placement heads and advanced vision systems are useless if there are no component placements or no boards to place them. Machine utilization can be achieved by making the processing board The speed is faster to increase; the time spent waiting for the PCB to move into the placement area is minimal.

Most SMD placement lines with a single conveyor belt are set up in the same way, using a separate production line board, the top surface of the mounting board to reflow, and the bottom surface being cured with an adhesive. This configuration creates several problems (except for Outside the maintenance and collaboration of the two lines): When the board moves from one line to another, sometimes called "intermediate storage", the number of people who need to operate the two lines and the resulting "downtime".

The replacement of the two-line pattern is a combination of the epoxy line and the return line, which is characterized in that the top and bottom surfaces of the PCB are over-sanded at one time. The disadvantages of this configuration include the length of the entire line, the difficulty of operation, and the requirement for additional personnel.

The board technology is another option. The combination board passes once and then the top and bottom surfaces are flipped over. As an option, the combined PCB can pass through the patch area twice, at the expense of slower production lines, instant PCB storage and downtime. Veneer technology, the production of composite panels is more expensive, and after the production run is completed, the panels must be separated.

The solution could be a two-channel conveyor (Figure 1) that can handle a dual PCB transport system simultaneously (synchronous mode) and/or patch the top and bottom of an assembly on the same machine (asynchronous mode) In synchronous mode, the same or different types of dual PCBs are simultaneously transported through the patch system. This gives the machine the greatest flexibility. In asynchronous mode, non-productive transport time is also minimized. A board moves into the machine, at At the same time as the patch, the second block of the same type is transferred to the machine. This is a feature that is not possible on a conventional high-speed machine.

Dual conveyor technology is certainly not for all applications. However, for high speed, high volume contract manufacturers and OEMs, the value of dual conveyor technology is clear.

Improved process technology
When most contract manufacturers and OEMs choose the best equipment for their production facilities, four classic standards must consider: speed (cph placement rate), accuracy, price/performance and quality. However, more and more manufacturers are The search for unimportant automation methods, such as process monitoring, process documentation, and data flow, raises the question of how existing devices can be connected to a process control system in a production facility, and how existing processes can be obtained. Parameters. Once, greater efforts were placed on optimizing automation related to real-time process monitoring and control, production documentation, and non-production time reduction. Production-specific properties and internal process policies and guidelines from three applications Important role:

Statistical Process Control (SPC). Deviations in the process are recognized early on. Possible confusion in production, such as the stopping of equipment, is prevented by monitoring appropriate process parameters.

Traceability. Which component is mounted on which PCB at which time? This information can only be obtained through documentation and material tracking during production, and is the only way to prevent large-scale traceability from failure to ensure traceability.
“Put down from the production line” makes it possible to increase the productivity of production equipment. PCB-specific data is distributed along the entire process parallel to a single product.

Protocol
It is well known that it is difficult to transfer information between devices produced by different vendors. Calls for communication standards are often mentioned, as expected, often in trouble in committee meetings, standards organizations and international debates. Why is semiconductor equipment? The members of the International Association of Materials and Materials (SEMI) have drafted an initiative for the definition of a communication protocol that can be beneficially manufactured because in the semiconductor industry, the connection equipment of different manufacturers is standard. Similarly, in PCB manufacturing,

the production process Monitoring and control are also necessary for thrift production.

Entering the Semiconductor Equipment Communications Standard (SECS) and the Common Equipment Model (GEM), a continuation of SECS II, which defines machine behavior on SECS II commands. Its purpose is to enable unified control protocols for control systems and equipment from different vendors.

Using a basic local area network and GEM controller, the SECS II/GEM communication protocol enables OEMs and contract manufacturers to complete full line level monitoring, ie, observe the entire production line and optimize production, rather than individually observing each piece of equipment. The agreement, SPC is easy to complete, because any changes in process parameters can be easily tracked, if necessary, corrected. In addition, "down from the production line" becomes a reality, for example, when PCB-specific information is exchanged across the entire production line, productivity It was added. Finally, traceability (locating an error and tracking it to the exact PCB and components) is possible. The result is a reduction in operating cost errors.The key to optimal use of the SECS II/GEM interface is the number of sensors installed in each machine. The more sensors mean the more process information, but the key is to decide how much information to exchange.

JOYSMT conclusion:
Advances in semiconductor silicon will continue to drive innovation in placement equipment. In fact, it has been seen early in the revolution in placement equipment that the leading industrial component packaging types are gradually changing. Undoubtedly, advanced technologies such as BGA, m BGA and CSP The packaging will continue to grow and eventually occupy most of the industry. The components will be smaller, and the new placement system will have to deal with components such as 0402, 0302 and 0201, with the least machine changes and downtime.