English
Language : 

OPA2652 Datasheet, PDF (12/15 Pages) Burr-Brown (TI) – Dual, 700MHz, Voltage-Feedback OPERATIONAL AMPLIFIER
BANDWIDTH VS GAIN: NON-INVERTING OPERATION
Voltage feedback op amps exhibit decreasing closed-loop
bandwidth as the signal gain is increased. In theory, this
relationship is described by the Gain Bandwidth Product
(GBP) shown in the specifications. Ideally, dividing GBP by
the non-inverting signal gain (also called the Noise Gain, or
NG) will predict the closed-loop bandwidth. In practice, this
only holds true when the phase margin approaches 90°, as it
does in high gain configurations. At low gains (increased
feedback factor), most amplifiers will exhibit a wider band-
width and lower phase margin. The OPA2652 is compen-
sated to give a flat response in a non-inverting gain of 1
(Figure 1). This results in a typical gain of +1 bandwidth of
700MHz, far exceeding that predicted by dividing the
200MHz GBP by NG = 1. Increasing the gain will cause the
phase margin to approach 90° and the bandwidth to more
closely approach the predicted value of (GBP/NG). At a gain
of +5, the 45MHz bandwidth shown in the Typical Specifi-
cations is close to that predicted using this simple formula.
INVERTING AMPLIFIER OPERATION
Since the OPA2652 is a general purpose, wideband volt-
age feedback op amp, all of the familiar op amp applica-
tion circuits are available to the designer. Inverting opera-
tion is one of the more common requirements and offers
several performance benefits. Figure 2 shows a typical
inverting configuration.
In the inverting configuration, three key design consider-
ation must be noted. The first is that the gain resistor (RG)
becomes part of the signal channel input impedance. If input
impedance matching is desired (which is beneficial when-
ever the signal is coupled through a cable, twisted pair, long
PC board trace or other transmission line conductor), RG
may be set equal to the required termination value and RF
adjusted to give the desired gain. This is the simplest
approach and results in optimum bandwidth and noise per-
formance. However, at low inverting gains, the resultant
feedback resistor value can present a significant load to the
amplifier output. For an inverting gain of –1, setting RG to
50Ω for input matching eliminates the need for RM but
requires a 50Ω feedback resistor. This has the interesting
advantage that the noise gain becomes equal to 2 for a 50Ω
source impedance—the same as the non-inverting circuits
considered above. However, the amplifier output will now
see the 50Ω feedback resistor in parallel with the external
load. In general, the feedback resistor should be limited to
the 200Ω to 1.5kΩ range. In this case, it is preferable to
increase both the RF and RG values as shown in Figure 2, and
then achieve the input matching impedance with a third
resistor (RM) to ground. The total input impedance becomes
the parallel combination of RG and RM.
The second major consideration, touched on in the previous
paragraph, is that the signal source impedance becomes part
of the noise gain equation and influences the bandwidth. For
the example in Figure 2, the RM value combines in parallel
with the external 50Ω source impedance, yielding an effec-
tive driving impedance of 50Ω || 57.6Ω = 26.8Ω. This
impedance is added in series with RG for calculating the noise
gain (NG). The resultant NG is 1.94 for Figure 2, (an ideal 0Ω
source would cause NG = 2.00).
The third important consideration in inverting amplifier
design is setting the bias current cancellation resistor on the
non-inverting input (RB). If this resistor is set equal to the
total DC resistance looking out of the inverting node, the
output DC error, due to the input bias currents, will be
reduced to (Input Offset Current) • RF. If the 50Ω source
impedance is DC-coupled in Figure 2, the total resistance to
ground on the inverting input will be 429Ω. Combining this
in parallel with the feedback resistor gives 208Ω, which is
close to the RB = 205Ω used in Figure 2. To reduce the
additional high frequency noise introduced by this resistor,
it is sometimes bypassed with a capacitor. As long as RB
<300Ω, the capacitor is not required since its total noise
contribution will be much less than that of the op amp’s
input noise voltage.
OUTPUT CURRENT AND VOLTAGE
The OPA2652 specifications in the spec table, though famil-
iar in the industry, consider voltage and current limits sepa-
rately. In many applications, it is the voltage • current, or V-
I product, which is more relevant to circuit operation. Refer
to the “Output Voltage and Current Limitations” plot in the
Typical Performance Curves. The X and Y axes of this graph
show the zero-voltage output current limit and the zero-
current output voltage limit, respectively. The four quadrants
give a more detailed view of the OPA2652’s output drive
capabilities, noting that the graph is bounded by a “Safe
Operating Area” of 1W maximum internal power dissipation
(500mW for each channel). Superimposing resistor load
lines onto the plot shows that the OPA2652 can drive ±2.2V
into 50Ω or ±2.5V into 100Ω without exceeding the output
capabilities, or the 1W dissipation boundary line.
To maintain maximum output stage linearity, no output
short-circuit protection is provided. This will not normally
be a problem since most applications include a series match-
ing resistor at the output that will limit the internal power
dissipation if the output side of this resistor is shorted to
ground. However, shorting the output pin directly to the
adjacent positive power supply pin will, in most cases,
destroy the amplifier. Including a small series resistor (5Ω)
in the power supply line will protect against this. Always
place the 0.1µF decoupling capacitor directly on the supply
pins.
DRIVING CAPACITIVE LOADS
One of the most demanding and yet very common load
conditions for an op amp is capacitive loading. Often, the
capacitive load is the input of an A/D converter—including
additional external capacitance which may be recommended
to improve A/D linearity. A high speed amplifier like the
OPA2652 can be very susceptible to decreased stability and
closed-loop response peaking when a capacitive load is
®
OPA2652
12