A seemingly simple circuit with only two components, an op-amp and a feedback resistor, is commonly used to convert a current to a voltage. This circuit is familiar as a photodetector or impedance-meter current-sense amplifier. The circuit is shown below, where the input capacitance of the op-amp, Ci is shown externally for analysis. Op-amps have typically around Ci = 10 pF across their inputs.
This kind of circuit is found in impedance meters, to convert the unknown current, ix , through the device-under-test (DUT) to a voltage, vix . The feedback resistance, RR in such an instrument is range-switched to cover a wide range of currents over a wide range of DUT impedances.
To keep the circuit seemingly simple, let the op-amp be very ordinary – a voltage-feedback op-amp with a single dominant pole at fG . G is the forward path gain of the feedback loop and H is the feedback path, in accordance with control theory nomenclature; no A for G and β for H as is commonly found in active-circuits textbooks because BJTs have an over-riding claim to the use of β . Gain magnitude decreases with frequency to a gain of one at fT from a high quasistatic (0+ Hz) gain of G0 . The op-amp is the forward path of the feedback loop with a voltage gain of
where G0 is the open-loop op-amp quasistatic gain magnitude and fG = ½ x π x τG is the open-loop bandwidth. In the dynamic or frequency-dependent factor of G that depends on s , it is sometimes easier to work with the math by using pole or zero time constants such as τG instead of pole or zero frequencies.
Two simplifications of this op-amp transfer function or gain can be made. The first is often closely approximated in practice and is what makes op-amps “operational”: high gain. By approximating the quasistatic gain, G0 , as infinite, or G0 → ∞, the first op-amp simplification occurs. Divide numerator and denominator of G by 1/G0 , then let G0 approach infinity;
where τT = 1/2 x π x fT = τG /G0 . This is the response of an infinite-G0 , finite-fT op-amp. The graph below is a frequency response plot of an op-amp having an open-loop bandwidth of fbw (ol) extended to zero as a pole at the origin by infinite open-loop quasistatic gain. The open-loop gain is G and closed-loop gain includes the feedback.
The second op-amp simplification is to let its frequency response be unlimited by letting the bandwidth of G become infinite – an “infinitely fast” op-amp. As τT decreases (or fT increases), the gain plot shifts to the right (or upward), which increases gain. A faster amplifier has more quasistatic gain, higher quasistatic loop gain, GH0 , higher precision, and can be less stable in a circuit with slow poles. An op-amp with infinite G0 and infinite fT is an ideal op-amp and is the op-amp assumed in the commonplace inverting and noninverting gain formulas for op-amps.
This second approximation of infinite bandwidth is not as practical as infinite gain and if assumed can sometimes result in an oscillating amplifier. Consequently, to understand seemingly simple op-amp circuits, it is sometimes – nay, usually – necessary to include the single pole in the gain derivation, then assess it for amplifier dynamic behavior. Unless the op-amp is being used in a slow, high-precision application, infinite G0 is usually valid in practice – but not infinite fT .
The closed-loop amplifier is a transimpedance (current in, voltage out) amplifier with transfer function,
The rational factor is the closed-loop feedback formula, with voltage gain vix /v – . Ti is the ix -to-v – transfer function before the input to the feedback loop. It is not in the loop but is entangled with it. Ci forms an input impedance with RR ;
The feedback part of the amplifier is a voltage amplifier that has input, v – = ix x Zin .
Because a voltage divider in one direction can be H and is Ti in the other direction, it is a good practice to always include Ti (and for similar reasons, To ) in the general feedback formula , so that it becomes
This feedback formula corresponds to the general block diagram shown below where x can be voltage or current.
For the noninverting op-amp configuration, Ti = 1 as is To , but there are circuits (such as this transresistance amplifier) for which it is important to recognize that circuit elements involved in the loop also have a pre-loop or post-loop effect on gain.
H is the feedback-path transfer function with the same Ci , RR as in Ti but as a voltage divider from output to error voltage, v – :
The feedback error-summing (the circle with Σ in the block diagram) is accomplished in the circuit by the superposition of the divider and input voltages: the input quantity, ix through Ti adds to the feedback quantity, the output of H . The feedback error-summer of the closed-loop formula subtracts the input voltage from the fed-back voltage, but because the voltage divider adds (does not invert), the negative sign in H corrects for the non-inversion so that the formula remains correct and consistent with the block diagram.
GH = G x H is the loop gain and is of interest because it determines loop stability. G , H , and Ti , when substituted into the feedback formula for a single-pole, infinite-G0 op-amp result in the closed-loop transimpedance of
Although RR is a resistance, the frequency-dependent Ci and op-amp make the closed-loop response an impedance which simplifies to
At s = 0+ Hz, the gain is an inverting transresistance, –RR , as expected. Frequency effects appear with the poles in the denominator. The amplifier has a two-pole (quadratic) response, where the poles are
In Part 2, we will examine the consequences for circuit behavior of Zm (s ).