Datasheet

=
V
I
V
O
a
1+
R
G
R
F
1+
Z(s)
R +R
F I
1+
R
G
R
F
=
a ´ NG
1+
Z(s)
R +R
F I
NG´
V
O
R
G
V
I
R
I
Z I(S)
ERR
a
R
F
I
ERR
= LoopGain
R +R
F I
´ NG
Z(s)
NG=NoiseGain=1+
R
G
R
F
R =530 NG R-
F
´
I
OPA2673
SBOS382F –JUNE 2008–REVISED MAY 2010 www.ti.com
Figure 82 shows the small-signal frequency response Developing the transfer function for the circuit of
analysis circuit for the OPA2673. Figure 82 gives Equation 13:
(13)
This formula is written in a loop-gain analysis format,
where the errors arising from a non-infinite open-loop
gain are shown in the denominator. If Z(s) is infinite
over all frequencies, the denominator of Equation 13
reduces to 1 and the ideal desired signal gain shown
in the numerator is achieved. The fraction in the
denominator of Equation 13 determines the frequency
response. Equation 14 shows this as the loop-gain
equation:
Figure 82. Current-Feedback Transfer Function
(14)
Analysis Circuit
If 20log(R
F
+ NG × R
I
) is drawn on top of the
open-loop transimpedance plot, the difference
The key elements of this current-feedback op amp
between the two would be the loop gain at a given
model are:
frequency. Eventually, Z(s) rolls off to equal the
a = buffer gain from the noninverting input to the
denominator of Equation 14, at which point the loop
inverting input
gain has reduced to 1 (and the curves have
R
I
= buffer output impedance
intersected). This point of equality is where the
I
ERR
= feedback error current signal
amplifier closed-loop frequency response given by
Equation 13 starts to roll off, and is exactly analogous
Z(s) = frequency-dependent open-loop
to the frequency at which the noise gain equals the
transimpedance gain from I
ERR
to V
O
open-loop voltage gain for a voltage-feedback op
amp. The difference here is that the total impedance
in the denominator of Equation 14 may be controlled
(12)
somewhat separately from the desired signal gain (or
The buffer gain is typically very close to 1.00V/V and
NG). The OPA2673 is internally compensated to give
is normally neglected from signal gain considerations.
a maximally flat frequency response for R
F
= 402Ω at
This gain, however, sets the CMRR for a single op
NG = 4V/V on ±6V supplies. Evaluating the
amp differential amplifier configuration. For a buffer
denominator of Equation 14 (which is the feedback
gain of a < 1.0, the CMRR = –20 × log(1 – a)dB.
transimpedance) gives an optimal target of 530Ω. As
the signal gain changes, the contribution of the NG ×
R
I
, the buffer output impedance, is a critical portion of
R
I
term in the feedback transimpedance changes, but
the bandwidth control equation. The OPA2673
the total can be held constant by adjusting R
F
.
inverting output impedance is typically 32Ω.
Equation 15 gives an approximate equation for
A current-feedback op amp senses an error current in
optimum RF over signal gain:
the inverting node (as opposed to a differential input
(15)
error voltage for a voltage-feedback op amp) and
passes this on to the output through an internal
As the desired signal gain increases, this equation
frequency-dependent transimpedance gain. The
eventually suggests a negative R
F
. A somewhat
Typical Characteristics show this open-loop
subjective limit to this adjustment can also be set by
transimpedance response, which is analogous to the
holding R
G
to a minimum value of 20Ω. Lower values
open-loop voltage gain curve for a voltage-feedback
load both the buffer stage at the input and the output
op amp.
stage if R
F
gets too low—actually decreasing the
bandwidth. Figure 83 shows the recommended R
F
versus N
G
for ±6V operation. The values for R
F
versus gain shown here are approximately equal to
the values used to generate the Typical
26 Submit Documentation Feedback Copyright © 2008–2010, Texas Instruments Incorporated
Product Folder Link(s): OPA2673