Winning the clone war

The difference between an emulated component and a replicated component may seem insignificant, but making the wrong choice can cause production delays or even affect long-term reliability When your critical component has passed ‘end-of-life’ and reached ‘end-of-existence,’ a little advance research and planning can make a big difference.

The dreaded day is finally at hand. Your dwindling supply of a crucial semiconductor will soon be used up. The original manufacturer has stopped making the component, there is no excess stock anywhere and grey market brokers offering potentially counterfeit or sub-standard devices are not an option. Various re-creation solutions are available. Unfortunately some options can result in production delays and product failure.

Understanding the choices begins with a vocabulary review: emulation produces a functional duplicate of the original component; replication produces a schematic and process duplicate of the original component.

So whats the difference? An emulated component is designed to mimic the function of the original device and is usually produced without the authorisation of the original manufacturer. A replicated component is a complete re-creation of the component as it was manufactured by the original manufacturer and is produced with the intellectual property of the original manufacturer. Replication can also include production and yield enhancements to the original design.

The difference between emulation and replication may seem insignificant, but in practice it is critical. With emulation, the final product may function as expected at first, however, if some infrequently used parameter of the original device is omitted, it can affect long-term performance. This means companies that choose emulation often experience halts and delays in production.

Semiconductor manufacturer and distributor, Rochester Electronics, specialises in addressing this problem with extension-of-life solutions, making is continued semiconductors available on a long-term basis.

Rochester’s semiconductor replication process (SRP) produces components that are physically and electrically identical to the original devices and are 100 per cent guaranteed by Rochester to meet or exceed the specifications of the original manufacturer. This is achieved by working under the approval and licensing of the original manufacturer. After replication, Rochester can manufacture a continuous supply of silicon wafers and finished devices.

Where available, Rochesters design and technology group work with the original manufacturers design and manufacturing intellectual property (IP). The team can also work from the last existing functional part, re-creating the test program, load-board, and in some cases, the packaging tools to re-introduce the original packaging solution. Products can be tested to standard, commercial, industrial, military, or space level manufacturing flows.

The approach to replication is determined by the available resources from the original manufacturer and the requirements of the customer.

Re-fabrication uses the original manufacturers design, tooling and masks and the same technology, to create new components. Reverse-engineering is where Rochester builds components from scratch when the design archive is no longer available. Process migration involves re-targeting a device to a similar process by translating the original design archive to modern tools and processes. Alternatively, Rochester can re-design a component including parameter modifications or enhancements, such as RoHS modifications or an upgrade from commercial to military specifications. Finally, process re-creation is possible in cases where the design archive is complete and a family of products is needed. Rochester can re-create the original fab process at a new foundry, maintaining the original designs.

The difference between ’emulation’ and ‘replication’ is more than just words. A true, 100 per cent guaranteed replication can make a significant improvement to the bottom line.