Flex circuits are used in a wide variety of products, including automotive assemblies, digital cameras, cell phones, calculators, flat-screen displays, printers and disk drives. According to the market research firm BCC Research (Wellesley, MA), the worldwide market for flex circuits will increase from $5.9 billion in 2005 to $11.2 billion in 2010, for an average annual growth rate of 13.5 percent.
The primary method for attaching flex circuits to printed circuit boards (PCBs), ceramics and glass is hot-bar soldering. Also known as pulsed-heat bonding, the process can be used to solder surface-mount edge connectors, battery packs, ribbon cables and small coaxial cables to PCBs. Because it's a selective soldering process, it's ideal for soldering heat-sensitive parts, such as CCD camera chips, or optically sensitive parts, such as Luxeon power LEDs.
Theoretically, the process can also be used to solder leaded components, though it's rarely done. "There are issues with coplanarity of the leads," explains Steve Hoover, pulse bonder product manager at Fancort Industries Inc. (West Caldwell, NJ). "You need a way to hold the component in place during the process. Usually, you have to glue it to the board first, so there may be issues with the adhesive and the heat."
Hot-bar soldering is a semiautomatic process. The equipment is similar to a benchtop resistance welder. It consists of a titanium thermode, which is mounted to a pneumatic cylinder and connected to a power supply. The parts are shuttled in and out of the machine on a linear slide with one fixture or a rotary indexing table with two fixtures.
A microprocessor-based control system allows the operator to program all elements of the soldering profile. These include:
- pressure.
- temperature.
- heat-up time.
- reflow time.
- cool-down time.
To begin, the operator loads the PCB into the fixture and applies flux to the joint area. (The solder is already present, having been applied to the pads and reflowed during the surface-mount assembly process.) Next, the operator loads the flex into the fixture, slides the fixture into the work area, and actuates the machine. The cylinder then pushes the thermode against the parts with a force ranging from 2 to 160 pounds. The force is applied during the entire cycle: heating, reflow and cooling. On most equipment, the force is applied with an accuracy of 0.25 to 0.5 pound
"The pressure is usually very light-less than 5 pounds of force," says Chris Heesch, product sales manager for hot-bar soldering at Miyachi Unitek Corp. (Monrovia, CA). "There are some occasions when more force is needed. For example, if the PCB is not flat."
The thermode is custom-made to fit the dimensions of the parts, and thermodes are easily changed on the equipment. The thermode is typically 0.05 to 0.08 inch wide and no more than 4 inches long. Ideally, the thermode should be longer than the bonding area by the space of one pad width on either side of the joint. The thermode does not have to be as wide as the pads to be soldered. It only has to cover approximately one-third of each pad.
"If you're having trouble getting a good bond, changing the width of the thermode may do the trick," says Hoover.
Once the thermode contacts the parts, it begins heating up. It typically takes 1.5 to 2 seconds for the thermode to reach peak temperature for reflow. This temperature varies depending on the solder alloy and the parts. If the leads of the flex are exposed, the peak temperature might be 300 C. If the leads of the flex are not exposed and the thermode must transfer heat through the Kapton, the peak temperature might be 400 C. If the substrate is a metal-cored PCB, which is designed to dissipate heat, the peak temperature might be 500 C.
At the peak temperature, it usually takes 3 to 8 seconds for the flux to activate and the solder to reflow. Though the thermode might be heated to 500 C, the joint itself will only reach typical reflow temperatures.
Once the solder reflows, energy to the thermode is stopped and it begins to cool down. The cooling process can be shortened by blowing air on the assembly. The cooling period is approximately 2 seconds. When the solder has solidified, the cylinder lifts the thermode away, releasing the parts.
According the Heesch, the three most important process parameters are time, temperature and pressure. "If you do not have enough heating time, you will not reflow the solder," he says. "If you have too much pressure, you will push the solder out of the joint, which may cause bridging."
Design Considerations
Besides the thermode, the most important component of the hot-bar soldering system is the fixture. The fixture aligns, holds and supports the parts during the process. It consists of a static-safe, heat-resistant thermoplastic nest on an aluminum base. Pins help to locate the flex over the substrate. In some cases, the fixture can be equipped with a vacuum system to pull the flex down on the substrate and keep it still before, during and after soldering.
"It's helpful if tooling holes can be designed into the flex and the substrate so that we can use locating pins in the fixture," says Hoover. "The fixture has to be ‘dead-nuts-on,' so the pads on the flex align perfectly with the pads on the substrate."
Some hot-bar soldering systems can be equipped with a video camera to help the operator load and align the parts. This feature is typically used in conjunction with an adjustable fixture, which enables the operator to fine-tune the position of the parts.
When designing an assembly for hot-bar soldering, engineers should provide clearance for the thermode on either side of the joint, as well as below the joint. "You want clearance around the pads where you're connecting the flex, especially on the underside of the substrate," advises Hoover.
If a component is blocking access to the joint, a space can be cut out of the thermode to provide clearance.
Engineers should also avoid concentrations of metal near the joint. These can act as heat sinks. "Any kind of metallization or gold tracing near the pads can suck heat away from the joint. [As a result,] the pads may not get up to the correct temperature," says Hoover.
If heat sinks can't be avoided, an auxiliary heating system can be built into the fixture to make up for any lost heat.
Hot-bar soldering can be done with any flux chemistry and solder alloy, including lead-free alloys. "What's more important is the amount of solder on the pads of the substrate," says Hoover.
After printing, the solder deposit should cover approximately 40 percent of the pad. After reflow, the average height of the solder on the pads should be 10 microns for small-pitch applications and 50 microns for large-pitch applications. As the distance between pads decreases, the importance of finding the right solder volume and thermode size increases, says Hoover.
Ideally, the pads on the flex circuit should be narrower than the pads on the PCB. This provides space for the solder to flow on either side of the flex pad and helps prevent problems with bridging.
Process Alternatives
Flex circuits do not have to be soldered to a PCB. Instead, they can be bonded to the board with a heat-activated, electrically conductive adhesive film, or they can be attached with a pin-and-socket connector.
A hot-bar soldering system can be configured to bond parts with adhesive film. Bonding is similar to hot-bar soldering in that the parts are held together under constant pressure. However, the heating profile is different. With soldering, the temperature rises, holds steady and cools down. With adhesive bonding, the temperature is held steady, with no ramp-up or cool-down period.
The adhesive method is typically used for flex circuits with fine-pitch leads, according to Tom Woznicki, president of Flex Circuit Design Co. (San Jose, CA) and editor of Flex Circuit News. Hot-bar soldering is good for leads with a pitch greater than 15 mils. Adhesive bonding is better for leads with a pitch less than 15 mils.
Using solder or conductive adhesive to attach flex circuits directly to PCBs has several benefits, says Woznicki. Engineers save the cost of a connector, and they don't have to allocate space for it in the assembly. Moreover, a direct connection can be more reliable than using a connector.
However, the process has some disadvantages too. It adds another step to the assembly process, and the solder joints can crack if the flex is pulled. In addition, rework can be difficult.
"Removing a flex circuit from a PCB will almost always ruin the flex," says Woznicki. "If you're using a simple flex jumper that's cheap, it's probably worth taking the risk. But if you have an expensive flex circuit or a complex flex assembly, I would explore every other option before using direct attachment."
