In the first article in the manufacturing Guide series, Sir Wei provided detailed information on supply chain procurement. Today, we will focus on the manufacturing process of PCBA.
Printed Circuit Board Assembly (PCBA) is formed by welding electronic components onto printed circuit boards (PCBS).
In other words, PCBA refers to the process of welding PCB empty board through SMT production process.
PCB is a customized component, you need to order from the PCB factory, the factory will produce according to the documents provided by the R&D.
The process of welding assembly of electronic components onto these boards is called PCB assembly, and PCB assembly has several objectives:
All components are in the correct position and direction;
The pins of each component are fully welded to the designated pad;
There is no excess solder that can cause problems;
There are no other residues, such as solvents or fluxes used in manufacturing, which can easily cause problems and may allow conduction in places that should not be conductive or cause corrosion.
PCB assembly is almost entirely automated, with one end of the assembly line feeding the PCB and electronics, and the assembled board coming out at the other end, requiring little human intervention.
The first step in the assembly process is to apply solder paste to the PCB.
1 – Apply solder paste
Solder is a metal alloy used to weld metal parts together electrically and mechanically.
Solder can be used to firmly weld the electronic pin to the PCB metal pad, while ensuring good electrical conductivity.
Solder paste is a viscous mixture of ultrafine solder powder and liquid flux, which is used to clean the metal surface of corrosion and dirt to form a good solder joint.
The first step in the PCB assembly process is to apply the right amount of solder paste to the correct location of the PCB. This process uses solder plates, also known as steel mesh, which need to be customized to the PCB designed.
The solder plate is a thin sheet of metal with many holes through which the solder paste is applied to the plate.
The template is precisely attached to the board, and the scraper evenly applies the solder paste to the template, then removes the template and cleans it, so that the right amount of solder paste is applied to the correct position of the board.
Now that the solder paste is applied, you need to install the electronics on the circuit board.
2 – Install components
The next step is to fit each electronic component into the PCB’s proper position, which is usually done by a SMT machine.
The components to be installed are usually supplied by a belt tray, each of which contains a component in a small raised container. Each container is sealed with a cover tape, which is removed until the component is attached.
Resistance on tape disc:
The SMT machine not only requires the correct components to be provided, but also needs to know the position of each component to be installed.
This is usually done with CAD data and bills of materials, which are also automatically generated by PCB design software.
Tubes and pallets are usually not as smooth as trays, depending on the particular mount.
If tubular or pallet packaging is used, it is best to study the suitability of the product with the production staff in advance.
Tubular packaging and pallet packaging:
When installing components, the laminator first determines the PCB position, then picks up the components with a small suction cup, places the components to the position specified by PCB and releases the suction cup.
The solder paste below the element acts as a binder to temporarily hold the element in place and is then reflow welded to solidify the paste.
In some cases, if the solder paste is insufficient to hold the element, the mounter may place a binder to hold the element.
The mounter is fast and precise, with a maximum of a few hundred components per minute, but that speed comes at a cost.
1) The first problem: the SMT machine is only suitable for surface mounting devices, not for large-sized perforated components.
Perforated components are usually placed and welded by hand, which is more costly than automatic surface mount device processes.
Replacing perforated components with their equivalent in surface mount devices is usually not a major problem, but this needs to be considered in advance during PCB design.
2) The second problem: all tape trays (or plastic tubes or trays) must be loaded into the laminator before the laminator works, which takes time.
The setup time is the same whether you have to assemble one circuit board or 1,000, so a few large batches are better than many small ones.
3) The third problem: the number of tape discs that can be installed by the patch machine is limited, ranging from 20 to more than 100, depending on the model of the patch machine.
If you need more types of components than the maximum number of reels the SMT machine supports, you must run the board through the SMT machine several times to install all the components.
To solve this problem, the best way is to reduce the number of components needed by using more of the same components in product design and development.
For example, you can try to standardize the resistance values and sizes of several resistors by connecting them in parallel or in series to obtain the desired resistance values.
4) Fourth problem: Sometimes you can only buy components on a reel-to-reel basis, but you may only need to use a small portion of them, and even then you need to buy a whole reel-to-reel.
Generally speaking, the number of components that can be loaded on a tape disk varies from a few hundred to several thousand. A plate of resistors might contain 5,000, but because resistors are cheap, costing only a few cents each, the overhead is not huge.
But a tape disk loaded with 500 GPS chips, if each GPS chip is 45 yuan, the purchase of the entire tape disk will cost about 22,500 yuan.
Even if you only use 100 of them, you’ll have to pay 22,500 yuan for the entire tape tray.
That works out to an average of 225 yuan per GPS chip.
Component dealers offer a workaround for this problem. They can either cut a portion of the tape tray and sell it, depending on the number of parts needed, or they can make a small tape tray just for the purpose.
For chips packed in tubes or pallets, the purchase quantity can be specified, but not all SMT machines meet this requirement.
3 – Reflow welding
The purpose of reflow soldering is to solidify the solder paste into a solder spot to weld components to the PCB.
Which includes heating the PCB and other two steps:
Make flux effective, do a good job of cleaning, and then evaporate;
Melt the underlying paste, then cool it to solidify into a block and weld the element pins to the pad.
The process is more complicated than simply heating the circuit board to a specific temperature and then cooling it.
In the process of heating and cooling:
Element heating and cooling should not be too fast, temperature impact will lead to failure;
At high temperatures, the flux has enough time to clean up and then evaporate;
The heat has sufficient time to permeate the entire circuit board surface, allowing the entire circuit board to reach the specified temperature. If heat does not permeate the board sufficiently, some components on the board may not gain enough heat to produce good solder joints.
Relationship between temperature and time during reflow welding:
Reflow soldering uses reflow furnaces, which can be programmed to set a temperature curve based on the type of solder paste used and other factors.
In addition to being programmable, the reflow furnace must ensure that each circuit board is heated evenly. High temperature gases (air or nitrogen) are generally used for heating, but other methods can also be used.
In actual production environments, reflow furnaces vary in size, depending on the volume. The reflow furnaces used in small batch production are small, like a microwave oven.
Large reflow furnaces for mass production have sustained production capacity.
Large flow-flow reflow welding furnace, the top cover is open:
Large commercial reflow welding furnace can continuously do reflow welding circuit board, circuit board is continuously transported to the reflow welding furnace by conveyor belt.
The reflow furnace has multiple zones that can be set independently of each other, as can be seen from the reflow temperature curve.
Under the top cover, the rear end of the fans can be seen protruding outwards, and these fans will blow air to keep the temperature even in the corresponding temperature zone.
When the circuit board comes out of the reflow furnace, hundreds of new solder spots appear on the circuit board.
Are these solder joints all good?
Are all components welded to the correct positions?
Sometimes it’s not so lucky. It’s better to check. That’s the next thing to do.
4 – Optical detection of AOI
When all the components are welded, the circuit board is placed in an automatedoptical inspection machine (AOI).
Check that all components on the circuit board are welded properly and in the correct position.
Through this process, all components in the wrong position or orientation will be detected and can be manually corrected or scrapped.
For example, in the figure below, you can see AOI’s test results displayed on the screen in front of the operator. AOI sees a component that is different from the previous circuit board and zooms it in.
Then ask the technician to check and answer this question: is there a problem with the way the component is installed, or has it been deliberately modified?
AOI results are displayed on the screen:
The “anomaly” detected this time is normal:
The marked component (resistance network) was produced by multiple suppliers and this time a new supplier was used for production.
The problem with the marking is that the font used to print “103” on this component is different from what AOI had seen before.
If AOI is told to remember this new style of “103”, the component will not be marked for the same reason in future PCBA.
Although this is not a major issue, sometimes these minor issues can stop production and AOI operators need to communicate with the head of design and development to confirm that everything is ok before they can continue production.
In some cases, less responsible AOI operators don’t talk to the people involved when they encounter problems that they think aren’t important, when in fact they can be so serious that they end up producing a lot of inferior boards.
In the production of products, every detail is crucial, production is not trivial.
Of course, it may be difficult to inspect components with invisible solder contacts.
The most difficult to examine are chips packaged in a “BGA” (Ball Grid Array) package, which has more than 1,000 invisible solder contact points.
Bottom view of small BGA:
PCB packages for medium BGA components, with hundreds of solder balls, can be difficult to inspect when these solder joints are under a square plate (BGA package).
Medium BGA package:
These invisible solder joints can be examined using custom-made 2D and 3D X-ray systems.
The tin bridge shown by the thick arrow in the image below accidentally welded two contacts together. Boards with such problems had to be manually corrected or scrapped.
Partial X-ray images of BGA:
Not all factories have X-ray systems, be sure to confirm this when selecting a product manufacturer.
At this point, most of the PCB assembly was done by automated means, and all the flaws in the process were noted down.
5 – Manual welding/assembly
In most cases, the assembly of certain circuit boards must be done manually, for example:
Repair the parts marked by AOI with reflow welding problems;
After the circuit board is manufactured, the design needs to be modified, such as the trace of the circuit board may be cut off and the addition of wires to change the circuit path;
Use manual welding of components not suitable for reflow welding, such as perforated components and temperature-sensitive components such as batteries;
Manual assembly of mechanical components on PCBA:
Hand assembly is more expensive than machine assembly, so designers and developers are always looking for ways to reduce the number of components that need to be assembled by hand.
Even so, some manual work is unavoidable in PCB assembly.
For example, connectors intended for end users, such as USB connectors, are often perforated rather than surface-mounted devices, which are more durable and have less risk of falling off the board when worn.
At this point, the circuit board has been fully assembled, if problems are found in the subsequent tests, it is necessary to return to the factory for correction.
In the next two steps (cleaning and cutting), you need to prepare the PCBA for assembly into the product.
6 – to clean up
A well-functioning PCBA must be clean, and it is best to clean it after the PCBA to remove any flux that has not burned off during reflow soldering.
Residual flux on the circuit board may cause corrosion, and unnecessary circuit paths may compromise the reliability and performance of the entire circuit.
Some fluxes are “clean free” and do not corrode or conduct electricity, even if they remain on the circuit board.
However, it is best to clean them up because they can still adversely affect highly sensitive and high-speed circuits.
7 – cutting
In product manufacturing, it is often the most economical way to manufacture and assemble multiple circuit boards on a single board, also known as a jigsaw, and then cut them into individual boards.
The number of boards on each board depends on the size supported by the PCB manufacturer and assembler, the size of each PCB, and other factors.
For example, an 18×24 board can hold up to 20 2.5×5 PCBS.
The picture below shows a board with five PCBS. There is a cutting line between each circuit board, and there is an opening on the corner of each circuit board, which is convenient for cutting later.
The V-cut board includes a shallow V-shaped cutting line at a specific position on the circuit board, which is convenient for subsequent operations to cut the circuit board according to the cutting line.
Curve boards use tools to cut any shape on the circuit board.
A single PCB after cutting:
PCB assembly is usually performed on a single board (containing multiple PCBS).
The individual PCB is typically cut out before testing using a cutting tool or manual separation (which puts more pressure on the board and components).
Cutting tools can be used to separate circuit boards in a variety of ways, including sawing, curve cutting, laser (laser) cutting, drilling and cutting with a special device.
The way the boards are split and cut makes the edges of the separated PCB boards different, which designers and developers need to be aware of.
During the cutting process, when cutting or separating, the edges of the PCB will be subjected to some stress and strain, and mechanical deformation may occur nearby.
To solve this problem, you need to leave at least 50 mils of space between the board edge and the trace/pad, and at least 100 mils between the board edge and the component.
Whether to use V-cut splitter or curve splitter should be carefully weighed:
Using a curved board creates smooth edges, whereas v-cut boards create rough edges, but this is usually a problem only when the board is visible to the end user, such as when the user connects the board to a connector.
Curve boards require greater spacing between circuit boards to accommodate cutting tools;
When curved boards are used, the circuit board may be significantly bent, which will have an adverse effect on PCB assembly.
To prevent these problems, it is important to design a PCB in close cooperation with the assembler, who knows their equipment very well.
Now that the PCBA is complete, it’s time to test to see if they work as expected.
Author: Sir Wei, public account: GANI Business
This article is originally published in Nuggets by @GANI Business. Reprint is prohibited without the author’s permission