Interposer with assembled chip (A). Various cross-sections through the embedding PCB showing the embedded chip (B), a resistor (C) and a capacitor (D).
December: Logic Embedding
Which logic components can generally be embedded into the PCB?
Nearly all logic components can be embedded into the PCB.
- Semiconductors based on low-k dielectrics as substrate material
- Chips in which the pitch of the electrical contacts < 80 µm
- Optical components (for obvious reasons)
Under what conditions does it make sense to embed passive components?
It always makes sense to embed passive components when they support the function of the embedded component, such as in the case of block capacitors or terminating resistors.
What advantages does embedding offer?
The embedding of components offers many advantages. The most obvious advantage is the saving of space on the assembly side because both the IC and the direct layout on the top side can be relocated to an inner layer. Due to the possibility of routing to layers situated both above and below, in certain conditions it may also be possible to reduce the complexity or layer count of the PCB. Furthermore, the heat dissipation of the IC also improves as the power loss is now no longer distributed via the soldering area and to a slight degree via convection but, instead, three-dimensionally with significantly improved heat transfers inside the PCB. The electrical behaviour may also exhibit a positive change due to the embedding of the IC. Shielding of the electromagnetic radiation is already ensured due to the technology used by the copper layers above the chip.
If required, IP protection can also be realised. The application can be implemented so that certain IC connections can only be accessed from outside by means of extremely complex preparation of the PCB.
The advantages of the embedding technology should be assessed individually and viewed in relation to a conventional solution. This is the only way to reach a meaningful decision regarding the optimum assembly and connection technology for the application.
Can packaged semiconductors also be embedded?
Yes, in principle. However, it is important to remember that the embedding of packages removes a number of application benefits compared with the embedding of semiconductor chips. Furthermore, the process is technically more complex and thus also incurs higher costs.
To summarise, SCHWEIZER advises against the embedding of package components for series production.
What limits apply today in terms of chip size and the number of electrical contacts for a chip?
As a matter of principle, it is technically feasible to embed both very small and simple chips as well as large and highly complex chips.
Due to the current cost structure and available production equipment for series production, there are, however, limits for the embedding of semiconductors that apply individually to each chip technology. The largest semiconductors to be embedded have an area of approx. 15 to 20 mm² and a number of contact pads ranging between approx. 150 and 200.
How do semiconductor chips have to be reworked so that they can be embedded?
This depends on the application and the assembly and connection technology employed. For components already provided with gold stud bumps, prototypes can be set up without any reworking. There are various options for use in series production such as nickel-gold under bump metallurgy (UBM), solder balls or even copper pillars with solder cap.
What challenges does the supply chain face?
The use of embedding technology demands openness from the partners involved in the supply chain in order to adapt the traditional business model of the electrical industry. Usually the manufacturer of the electronics purchases the PCB, active and passive components as well as additional components. These are then assembled in an SMT soldering process to form an electronic module.
In the embedding model, a workable business model has to be drawn up between the manufacturers of the PCBs and the semiconductors in close cooperation with the customer. The business model has to regulate the purchase of wafers, their reworking and mounting as a flip chip on an interposer, the electrical test, the joint qualification process and the quality requirements. In addition to adherence to the specifications within the series, automotive and industry customers also demand a carefully devised quality concept and, in the event of returns, clear responsibilities from the partners involved.
What is the difference between i² Board® (i² = integrated interposer) and µ² Pack® (µ² = µ-thin / µ-pitch)?
In PCB embedding technologies, a distinction is made between two fundamental approaches. Firstly the “System in PCB” approach, in which components are integrated into the main circuit board. In the “System in Package” approach, a semiconductor package is produced with the aid of PCB and embedding technologies which is then placed on a main circuit board. The i² Board® is one of the “System in PCB” solutions while the µ² Pack® adopts the “System in Package” approach.