hf/rf ZONE Products for the week of January 19, 2004
Texas Instruments Says . . .
OPA695: Wideband Current-Feedback Amplifier For Video
And Wireless
Texas Instruments Incorporated (TI) announced a low-power, current-feedback amplifier from the company's Burr-Brown product line, which offers unmatched bandwidth at low and high gains. The device is designed for applications in video signal processing, cable modems, wireless communications and test instrumentation.
At higher gains (G = +8), the OPA695 offers 4300V/usec slew rate and 450MHz bandwidth, providing a low-power solution for Intermediate Frequency (IF) amplifier applications. Capable of single supply +5V to +12V operation, the OPA695 gives an improved output intercept at lower power than existing dedicated IF amplifiers. At lower gains (G = +2), popular for video line driving, the OPA695 offers 1.4GHz bandwidth with 2500V/usec slew rate, covering the highest speed video requirements.
"The OPA695 establishes a new performance level for wideband current feedback amplifiers. No other competitive device can match the bandwidth at a gain of +2, nor hold a high bandwidth to higher gains as well as the OPA695," said Michael Steffes, strategic marketing manager for TI's high-speed signal processing group. "Wireless communication applications will benefit from this low power, low cost, high dynamic range IF amplifier, while high-resolution RGB applications will enjoy exceptional pulse fidelity through the highest pixel update rates."
The low 12.9mA supply current is precisely trimmed at +25C. This trim,
along with low quiescent current shift over a wide temperature range, gives
low system power over temperature. Fully specified over both the commercial
(0 C to +70 C) and industrial temperature (-40 C to +85 C) ranges, the OPA695
shows minimal drop-off in performance at the temperature extremes. Using
the optional disable control pin, the OPA695 supply current may be further
reduced to less than 170uA. This power-saving feature, along with exceptional
single +5V operation and ultra-small SOT23-6 packaging, will be attractive
to wideband amplifier requirements in portable applications.
analogZONE Says . . .
I seem to have spent part of this last week looking at data sheets where some specifications are downright unbelievable, or they are completely missing (wonder why?), so it is such a pleasure to pick up a Burr-Brown op amp data sheet where nothing is left out, and where everything is spelled out over full temperature ranges. This is how it is with the data sheet for the OPA695 which is specified for both split ±5 V rails and for a single +5 V rail. The part can be operated with a rail spread up to 12 V total.
Getting bandwidth and slew rate performance without throwing current at the solution is not an easy semiconductor design challenge but with a typical 12.9 mA on split rails and 11.4 mA on a single rail, TI (Burr-Brown) shows that it knows what it takes to design small transistors that can work. The higher performance, of course, is on split rails, where video design engineers like to work because a single design can be used with any analog format, and the RF design engineer -- who is more used to working with a single rail -- may be tempted to take the higher performance by changing his preferences.
The bipolar rail performance at a gain of +2 shows a typical small-signal bandwidth of 1400 MHz, and a 0.2 dB gain flatness bandwidth of 320 MHz, very believable and very usable for high-performance video. Slew rate is not quoted for this lower, line driver type of gain.
At a higher gain of +8, a more likely type of operation for an IF application, the small-signal bandwidth is a typical 450 MHz -- as is the large signal bandwidth -- and the slew rate is an impressive 2900 V/µs, typical. If the configuration is set for an inverting -8 gain that number improves to a typical 4300 V/µs. Very impressive. The noise figures of an op amp compared to a fixed-gain RF amplifier are a little dreary, but provided there is sufficient gain before a stage using an op amp the mathematics show that the poorer performance is immaterial. What is to be gained, however, is much better matching, better load isolation and higher 3rd-order intercepts. The input voltage noise is not the lowest available but it is good at 1.8 nV/rtHz while the 3rd-order intercept is better than 30 dBm to over 100 MHz. The output swing with bipolar supplies of ±5 V can be up to ±4.2 V
To further show that they are serious about RF designers using the OPA695 the company has put some RF characteristics into the data sheet: With a gain of -8 (and split rails) S11 is better than 20 dB (a VSWR of about 1.2:1) up to about 100 MHz; with a gain of +8 that improves to about 400 MHz. S22 is better than 20 dB out to about 120 MHz but can be improved with a suitable trim capacitor.
Distortion numbers are quoted at 10 MHz but there enough Figures provided that you can estimate the numbers at your frequency of interest.
All-in-all the OPA695 offers extremely good performance for video and would make an ideal IF amplifier, LO buffer, and such for frequencies certainly up to 100 MHz and, with careful design, higher. It is also a shoe-in for designs previously using the OPA685. The premium asked for the performance given is quite low.
The OPA695 is in production in SO-8 and will be available February 2004 in SOT-23-6, priced at $1.35 in 1000-piece lots.