The optoelectronic hybrid device has experienced a revitalization over the past few years. While hybrid technology has always been a useful design methodology, recent changes in the economy and in engineering philosophy have made it popular again. In fact, it has become a viable alternative to single-chip devices.
Simply put, a hybrid device is a single substrate that contains one or more individual semiconductor die that comprise a single circuit. This circuit usually serves as a small functional portion of a larger circuit and is addressable as a single device rather than by its individual die. In the case of the optoelectronic hybrid, the hybrid contains die that emit and/or detect light.
The encoder hybrid (Fig. 1) consists of an aperture plate with slots in it, a broad-field light source on one side, and an array of detectors on the other side. As the aperture plate moves relative to the detectors and light passes through the slots, the different detectors receive patterns of signals, which, upon conversion, provide electrical information about the position of the system.
The smaller the slots are, the more slots the aperture can contain, which means higher resolution for the system. While having the detectors close together is essential for high resolution, it also minimizes cross talk between detectors due to stray light. The encoder hybrid has numerous applications, though one of the more common applications is the volume knob, such as the knob on a car stereo (Fig. 2).
Reflective And Receiver Hybrids
In the reflective type of hybrid, at least one die emits light (Fig. 3) while at least one other die detects it. As an object passes in front of the hybrid, the amount of light that reflects off of the object and returns to the detector(s) provides different information about that object.
Depending on the type of die used as well as on the mechanical design of the system, the reflective hybrid can tell how far away the object is, how rough or shiny its surface is, and in some cases what color the object is. The reflective hybrid is very popular in the printer industry for functions such as verifying the correct paper size and type and that it is feeding its paper evenly.
The receiver type of hybrid consists of a detector die, an amplifier die, and the associated capacitors required for noise filtering (Fig. 4). The receiver hybrid is very useful in the telecommunications industry for transferring large volumes of data where high noise immunity and low signal degradation are key.
Why Use A Hybrid?
There are many advantages to using hybrid technology—first among them, size and weight. Since the hybrid usually consists of bare die rather than packaged die, it can be a relatively small device, taking up much less room on a larger printed circuit board (PCB) than prepackaged components. Also, as a direct result of the absence of the bulky packaging materials, the hybrid weighs less than its counterparts.
Continue on next page
Considering optical pitch, the spacing between optical elements is critical in many applications. Eliminating the bulky packaging materials common to prepackaged die allows designers to place the optical elements as closely to each other as possible.
In high-speed circuits, especially in the telecommunications field, circuit paths need to be as short as possible. The use of surface-mount elements in these types of circuits can be detrimental to a circuit. A hybrid allows the functional circuit elements to be as close together as possible, significantly reducing the harmful capacitances and inductances present in all circuits.
For example, the emitter in a reflective hybrid usually consists of aluminum gallium arsenide (AlGaAs) or indium gallium phosphide (InGaP). The detector is usually gallium arsenide (GaAs) or silicon (Si). Even with all prior processing enhancements, it’s still impossible to combine these types of materials on a single die. The hybrid, as the name “hybrid” implies, can combine different types of technologies in a single device.
Hybrid designs allow for modularization, meaning different engineers or even different engineering companies can use their respective expertise to design parts of a circuit. Even in companies where complete vertical integration is possible, it makes sense to have different groups that are highly specialized in one field. Where modularization exists, hybrids fit right in.
When a circuit doesn’t function properly, we can probe key portions of it to try and localize the problem. If a hybrid is performing the function of the problem area, it is simpler to remove the entire hybrid rather than having to try and remove all of the affected elements individually.
Given that the hybrid derives from PCB technology, the time that it takes to prototype a hybrid is the same as prototyping a PCB, which is usually a week or two. This, of course, assumes that the necessary internal components are readily available. A side benefit here is that if the problem is the hybrid, the turnaround time for a redesign is still only a week or two.
Lastly, the cost for tooling a hybrid is very low. Usually, PCB manufacturers require a small setup fee, an artwork fee, and a test board fee. These costs total less than $1000 for some applications.
The Current Engineering Climate
Several different factors are contributing to the resurgence of the hybrid. During the 1980s, 1990s, and even into the 2000s, engineering companies enjoyed a booming economy. The massive influx of money meant that these companies, to a large degree, had a blank check for new designs. However, the financial reality of today has made wasteful spending a thing of the past.
As the economy has changed, so has the corporate mindset. Design teams are constantly under pressure to decrease project costs and get designs right the first time. Hiring new employees for part of a design is not an option, especially if existing companies have that expertise. Design teams are additionally encouraged to design around existing equipment given that new pieces of equipment can mean investments of millions of dollars.
Continue on next page
There also has been an increased focus on time-to-market. In today’s global economy, there is competition from everywhere. Especially in the world of startup companies, beating the competition can mean the success or failure of an entire company. In an effort to reduce timelines, companies are encouraged to work with other companies that may already have existing technology or expertise for parts of their designs.
In recent years, there has also been a push to reduce overhead. It is even more important now for companies to outsource parts of their design or parts of their manufacturing to other companies.
Hybrid Versus IC
Given these economic pressures, the hybrid has found new life. Because the hybrid is a specialized PCB, its tooling costs are very low. By contrast, the cost for tooling an IC on the same scale as the hybrid in complexity is often 25 to 50 times higher.
Tooling a hybrid can take a few days to a few weeks depending on the monetary comfort zone. But even the quickest prototypes are still inexpensive. By contrast, the tooling time for even the simplest IC is several months.
Companies always have to pay for an initial volume of parts when they’re tooling a hybrid or an IC. For a hybrid, the initial volume is in the hundreds. For an IC, it’s in the thousands. As long as the design is good, this isn’t a disadvantage for the IC, given that any initial volume eventually will be incorporated into a product. But if there needs to be a redesign, the tooling and the initial volume of parts are both wasted.
Redesigning, which is a fact of life in any project, is tougher with an IC given the tooling costs, tooling time, and waste of inventory. And if additional design requirements present themselves or if enhancements are necessary, project costs can mount quickly.
For these reasons, there is a tradeoff between the hybrid and the IC. Companies that manufacture ICs tend to be very partial to high volumes, where the IC is unquestionably less expensive than a hybrid. So, for the product that will be high in volume and low in cost, the IC is the right choice. But in products where volume is low and the asking price can be high, the hybrid is the smart alternative.
The amount of time that it takes to create a medical product can be staggering, starting with countless designs and redesigns, followed by all the clinical studies. Finally, the product goes to the U.S. Food and Drug Administration for approvals, which can take months. Hence, medical companies are looking for any way possible to shorten any part of this timeline.
Hybrid designs allow medical companies to prototype their products quickly. Also, given the inexpensive tooling for the hybrid, they can prototype over and over again as needed when new inputs from clinical studies come in.
Furthermore, given the small areas that some medical designs must fit into, hybrids fill an obvious need. Large PCBs definitely have their place in large pieces of medical equipment. However, devices that come in contact with the human body or are implanted in the body need to be as small as possible. The ability to place optical elements accurately and very close together is a natural requirement.
Continue on next page
One final factor is noise immunity. Signals generated by devices that interact with the human body must be very small so there is no possible harm to the patient. Any extra path length between circuit elements can contain noise that overwhelms the desired signal. The hybrid provides an excellent means of minimizing unwanted circuit noise.
Encoder designs are a natural area where hybrid designs thrive. Encoders are becoming smaller and more complex, and they are continuously striving for higher resolutions. How much of this can be achieved heavily depends on how closely spaced the optical elements can be. In the hybrid design, the only impediment to smaller die pitches is how small the die can be. Therefore, the hybrid provides the best possible solution.
Another reason is the pick and place, a.k.a., surface-mount technology. Most companies that design complex circuit boards, or the companies that build them, have dozens of pieces of equipment designed to handle prepackaged die. However, it takes a different expertise to handle bare die. Some companies already have this expertise, so it makes sense to outsource this part of the production.
In addition, a Class 10,000 clean room usually is more than adequate to handle prepackaged materials, though the handling of bare optical die requires Class 1000 or better. The difference in costs for constructing each of these types of rooms is staggering. Another reason is development cost, but it depends on the type of encoder.
The basic encoders on the market are very inexpensive. In the beginning stages of their design, prototyping with hybrids makes sense to keep down design costs and time. But when these designs go into high production, it is more prudent to design an IC.
In the case of the more complex encoders, there tends to be a lot of design effort, and the designs can be very tricky. Once the designs are complete, it usually isn’t necessary to spend the additional money to create IC die. The cost of the product can absorb the higher production cost of the hybrid.
The biggest reason for military designs to use hybrids is that there is no need for high volume. By nature, military designs aren’t commercial, so there is no need to create millions of any piece of equipment. As such, the pressure of price, which usually favors the IC, does not exist.
Security is also an issue. For most countries, IC design and manufacturing companies are offshore. To protect the sensitive nature of military designs, companies prefer the route of the hybrid because it allows them to more closely monitor who has access to the designs.
Another point has to do with older designs. As the military uses a design for decades or longer, it is common for the original manufacturer to come under new ownership or even cease to exist. In those cases, some designs will require reverse engineering and more. Given that hybrid designs were very common in the past, it is easier for new companies to copy the hybrid designs that already exist.
Startups And Smaller Companies
Smaller and/or startup companies can benefit greatly from hybrid designs. It comes down to the two most obvious reasons: time and money. The money for smaller companies usually comes from venture capitalists, who are quite sensitive to how much money is spent and how long it takes for returns to come in. The costs for tooling a hybrid are very low and the prototyping lead times are low as well, so they are an obvious choice for these types of companies.
The question of whether the hybrid or IC is better for designs is not new. But while the IC has its obvious place in certain designs, the hybrid is definitely making a strong comeback. And more and more, companies and industries are starting to take advantage of its numerous benefits.