ANADIGICS expanded its position as an RF IC solution leader with the introduction of its first silicon device, the ACS2301 dual-frequency synthesizer. Designed using the company's expertise in high-frequency design, this dual-frequency synthesizer was designed to support cable modem, set-top box, television (TV), and cable telephony tuner applications.

Ron Michels, Vice President, Cable & Broadcast Segment, ANADIGICS, Inc. observes that: "ANADIGICS is committed to identifying, developing, and exploiting new process technologies to provide leading RF solutions for our customers. The development of silicon products broadens our product portfolio and provides new opportunities in the RF communications markets."

The ACS2301 is manufactured using a CMOS process, which has fewer steps than a BiCMOS process, translating to lower costs for the end user. Michels explains: "Most dual-synthesizer solutions for tuner applications are manufactured using BiCMOS processes. We wanted to provide a low-cost, highly integrated, easy-to-use solution, so we chose a CMOS process and applied our IC design expertise to develop this highly integrated new device."

Designed and 100% RF tested at the ANADIGICS facility in Warren, NJ, the devices are being manufactured by the Taiwan Semiconductor Manufacturing Company Ltd (TSMC) in Taiwan.

The synthesizer is designed for use in dual-conversion tuners which handle two intermediate frequencies (IF) and have two on-board voltage-controlled oscillators (VCOs). The ACS2301 is equipped with two fully-programmable RF dividers at 2.0 and 1.1 GHz. The device divides each VCO frequency by a known division ratio and compares it to a stable reference frequency. It then generates a DC tuning voltage that the VCOs use to provide very stable, low-noise RF local oscillator (LO) signals.

I remember sitting in a conference room at Anadigics a few years ago and asking "the question" about whether processes other than the GaAs might be used by the company in the future. Generally, companies that only manufacture in gallium arsenide have an almost religious fervor about the process, something that has always disturbed me, as all enthusiasts disturb me; Anadigics' answer to the question was perfectly straightforward, in that it was willing to consider any process that made sense. And that commonsense certainly holds true for this sort of part; the stability and cost of silicon for this RF area is virtually a no-brainer. Managing to produce a satisfactory performance with CMOS reduces the costs even further.