High Speed IC Op Amps


In the earliest years of IC op amps, everyone was using essentially the same NPN bipolar process, and speed was severely limited because of the slow PNP transistors available. An early scheme to partially get around the PNP bottleneck was the LM118/218/318, designed by Bob Dobkin at National Semiconductor in 1971 (Reference 63: Bob Dobkin, "LM118 Op Amp Slews 70 V/μs," National Semiconductor LB-17, September 1971 (The high slew rate LM118 IC op amp)). ADI produced their own version of this op amp, the AD518, designed by Dave Kress. Although these amplifiers did achieve much higher slew rate and bandwidth, they did not settle fast, nor were they well-suited to driving low impedance loads.

In the early seventies, just about the only truly fast IC process was owned by Harris Semiconductor. This dielectrically isolated process produced equal speed NPN and PNPs, and the Harris HA2500 series became popular for fast settling characteristics. In 1973 ADI released the fast AD509 op amp, a screened Harris part (Reference 64: "AD509: Fast Op Amp 2μs to 0.01%," Analog Dialogue, Vol. 7, No. 1, pp. 10 (The AD509 fast IC op amp)).

Until junction isolated CB processes came on board, the dielectrically isolated parts were to dominate high speed applications. There were however, notable exceptions to this general rule. The AD744, designed by Scott Wurcer, was introduced in 1988 (Reference 65: Kristen Dinsmore, "Fast, Accurate BIFET Op Amp Settles to 0.01% in 900ns Max," Analog Dialogue, Vol. 21, No. 2, 1988, pp. 14-15). Although this op amp still used a basic NPN process, it took advantage of ion-implanted P-channel FETs for the input differential stage, and could settle quickly and cleanly, reaching a 900ns settling time to 0.01%.

ADI introduced a high speed 36V CB process in 1988 (see References 57 and 58, again), and with it, a host of fast IC op amps. Among these were a high speed voltage feedback group, the AD840 series, and the AD846 current feedback op amp, all designed by Wyn Palmer. Many other very successful op amps were to soon follow in this series, using the CB process. Notable among them were the unity-gain stable AD847 and externally compensated AD829, also designed by Wyn Palmer. Later on, the AD811 designed by Dave Whitney, was among the first high performance current feedback op amps available on the CB process, achieving very low video distortion specifications while driving 75Ω cables (Reference 66: Dave Whitney and Walt Jung, “Applying a High-Performance Video Operational Amplifier," Analog Dialogue, Vol. 26, No. 1 1992, pp. 10-13 (The AD811 high-speed, high output video IC op amp and applications)).

Frank Goodenough's Op Amp Reporting for Electronic Design

A notable documentation source on these high speed op amp developments was an Electronic Design series, by analog editor Frank Goodenough (References 67-71: Frank Goodenough, “Focus Monolithic IC Op Amps: Faster and More Precise,” Electronic Design, February 4, 1988 pp. 127-132 (A survey of high-speed IC op amps); Frank Goodenough, “A Slew of New High-Performance Op Amps Shatters Speed Limits,” Electronic Design, March 3, 1988 pp. 29-32 (High-speed IC op amps and processes, including the ADI CB process); Frank Goodenough, “New Processes, Designs Boost IC Op Amp Speeds,” Electronic Design, April 12, 1990 pp. 45-48, 52-56 (Survey of high-speed IC op amps with emphasis on CB processes and current feedback architectures); Frank Goodenough, “Linear ICs Attain 8-GHz NPNs, 4-GHz PNPs," Electronic Design, December 19, 1991 pp. 35-37, 40, 43-45 (The high-speed ‘UHF-1’ DI process as used by Harris); Frank Goodenough, “New Processes to Spawn Next-Generation Analog, Mixed-Signal, Power ICs,” Electronic Design, January 2, 1992 pp. 59-62, 64-66, 68, 70 (Survey of high-speed process developments)).

The CB process was just the beginning of ADI high speed IC op amps, and within less than a decade a further jump in performance was produced. This was the 12V XFCB process, introduced in 1993 (References 72 and 73: Frank Goodenough, “Wideband IC Op Amps Reach New Bandwidth Highs,” Electronic Design, September 2, 1993 pp. 48, 50, 52, 56, 58, 63, 64-66 (The ADI XFCB process, and a survey of other high-speed processes) and Frank Goodenough, “IC Op Amps Combine Low Cost and Performance,” Electronic Design, September 2, 1993 pp. 39-40, 42, 46 (Survey of high-speed processes, ADI AD8001 XFCB process IC op amp and AD96XX series IC op amps)). This produced such key parts as the AD8001, designed by Scott Wurcer (Reference 74: “Fast Op Amp: High Performance, Low Power, Low Cost," Analog Dialogue, Vol. 28, No. 2, 1994 pp. 11 (The AD8001 high-speed IC op amp, first on the XFCB process)). The AD8001 set new performance standards, hitting a bandwidth of 800MHz on ±5V supplies, and achieving very low video distortion.

Frank Goodenough's op amp articles continued to provide a valuable source of IC performance, as well as historical references, through the late 1990's, including other op amp categories as well (References 75-78: Frank Goodenough, “Silicon, Analog Processes Becoming More Sophisticated,” Electronic Design, May 2, 1994 pp. 97,98, 100-103 (Survey of complementary and mixed signal compatible processes)Frank Goodenough, “Rail-Rail In-and-Out IC Op Amps Run Off 2.7V,” Electronic Design, May 16, 1994 pp. 51,52, 54, 56, 60, 64,-65 (Survey of rail-rail, low voltage single-supply design techniques and IC op amp devices); Frank Goodenough, “Single-Supply Op Amps Come of Age,” Electronic Design Analog Applications Issue, November 7, 1994 pp. 52-61 (Survey of rail-rail design techniques and IC op amp devices); Frank Goodenough, “Op Amp, ADC Topple Old Price/Performance Marks,” Electronic Design, December 5, 1994 pp. 75, 76, 78, 80, 81, 84, 87 (New product highlight of high-speed, low power AD8011 IC op amp and AD878 ADC)). He passed away in February of 1998, and was fittingly memorialized by Roger Allan of Electronic Design (Reference 79: Roger Allan, "Frank Goodenough: 1925-1998," Electronic Design, March 9, 1998 pp. 6).
Share this article :

0 comments:

Post a Comment

Please wait for approval of your comment .......