Lab 10
Lab 10
A. Theory
In this experiment, we built the folded-cascode differential amplifier shown in Figure 1. This amplifier operates
similarly to the amplifiers that was investigated in Labs 8 and 9. This differential amplifier has an advantage
of having a larger transconductance gain than the last two amplifiers.
B. Procedure
For this experiment we constructed a folded-cascode differential amplifier, shown in Figure 1 (a), using two
ALD1106 nMOS transistor arrays and two ALD1107 pMOS transistor arrays. To measure the voltage transfer
characteristics, we held Vc at 1V, 2V and 3V , V2 at 2V and swept V1 from 1.8V - 2.2V while measuring Vout .
We also held Vc at 3.9V , V2 at 1V, 2V and 3V and swept V1 from 0V - 5V while measuring Vout .
FIG. 1: (a) A simple MOS folded-cascode differential amplifier. (b) A simple bias circuit for generating Vbn
and Vbp .
Figure 2 shows the voltage transfer characteristics of the amplifier as we sweep V1 , hold Vc at 1V, 2V and 3V
and hold V2 at 2V . We observe kinks in the region of high gain. The size of the kinks decreases with increasing
Vc . The start of the kink occurs at higher Vout as Vc increases.
Figure 3 shows the voltage transfer characteristics of the amplifier as we sweep V1 , hold Vc at 3.9V and hold
V2 at 1V, 2V and 3V . As expected, the output voltage rises dramatically at the set point voltages just as the
case from Labs 8 and 9. In addition, the output voltage goes from rail to rail.
2
4.5
3.5
3
[V]
2.5
out
V
1.5
0.5 V , V = 1V
out c
V , V = 2V
out c
0 Vout, Vc = 3V
1.75 1.8 1.85 1.9 1.95 2 2.05 2.1 2.15 2.2 2.25
V [V]
1
FIG. 2: The voltage transfer characteristics between Vout and V1 of a simple MOS folded-cascode differential
amplifier.
A. Theory
The slope of the voltage transfer characteristic between Vout and Vdm in the high-gain region is the differential-
mode gain for this MOS amplifier.
The reciprocal of the slope of the current-voltage characteristic between Iout and Vout (when Vdm is held
constant) in the shallow region of the characteristic is the incremental output resistance for this MOS amplifier.
The slope of the current-voltage characteristics between Iout and Vdm in the high-gain region is the incremental
transconductance gain for this MOS amplifier.
The differential-mode gain can also be calculated by multiplying the incremental output resistance by the
incremental transconductance gain, as shown in Equation 1.
B. Procedure
For this experiment, we set V2 to 2V , Vc to 1V, 2V and 3V , set the differential-mode input voltage to zero
and swept Vout from 0V - 5V while measuring Iout . We fit a straight line to the shallow region of the data which
corresponds to incremental output conductance, whose reciprocal is the incremental output resistance. Then,
3
4.5
3.5
3
[V]
2.5
out
V
1.5
0.5 Vout, V2 = 1V
Vout, V2 = 2V
0 Vout, V2 = 3V
FIG. 3: The voltage transfer characteristics between Vout and V1 of a simple MOS folded-cascode differential
amplifier.
we fixed the output voltage at 1V and measured the output current as we swept V1 around V2. We then fit a
straight line to this data to find the incremental transconductance. We then multiplied the incremental output
resistance and incremental transconductance to calculate the differential-mode voltage gain.
Figure 4 shows Iout as a function of Vo ut for Vc = 1V, 2V and 3V with V2 = 2V . The slope of the fit in the
shallow region gives an incremental output conductance. Taking the reciprocal of this gives an output resistance
of 5.18 × 106 Ω for Vc = 1V , 6.57 × 106 Ω for Vc = 2V and 8.38 × 106 Ω for Vc = 3V . Just as with the voltage
transfer characteristics from Experiment 1, we observe kinks in the data. The kinks in these datasets occur at
higher Vout for higher Vc , just like Experiment 1.
Figure 5 shows Iout as a function of Vdm for Vc = 3.9 and V2 = 2V . The slope of the best-fit line in the
linear region of the data gives an incremental transconductance of 5.78 × 10−5 Ω1 . Multiplying the incremental
transconductance with the output resistance give a differential-mode voltage gain of 300 for Vc = 1V , 380 for
Vc = 2V and 485 for Vc = 3V . This is about 3.5 times higher than the differential-mode gain obtained from
the simple differential amplifier from Lab 8 and about the same as the differential-mode gain obtained from
the current-mirror differential amplifier from Lab 9. The output current has limiting values of 7.9 × 10−6 A and
−9.3 × 10−6 A.
4
x 10
−6 Output Current − Output Voltage Characteristics
6
Vout Linear Fit, Vc = 1V
V Linear Fit, V = 2V
5.5 out c
V Linear Fit, V = 3V
out c
Vout Experimental, Vc = 1V
5
Vout Experimental, Vc = 2V
Vout Experimental, Vc = 3V
4.5
4
[A]
3.5
out
I
2.5
1.5
1
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
V [V]
out
FIG. 4: The output current - output voltage characteristics between Vout and V1 of a simple MOS
folded-cascode differential amplifier.
A. Theory
We can create a unity-gain follower using our amplifier by connecting Vout to our inverting input, V2 . The
step response of the unity-gain follower can be found by measuring the device’s response to a square wave input.
In prelab 9, we found that the relation in time for Vout and Vin for a small-amplitude input to be
dVout
τ + Vout = Vin (2)
dt
where
C
τ= . (3)
Gm
This means that Vout will exponentially
approach Vin . We found that for a large-amplitude input, Vout will
change linearly in time at a rate of ICb , until it transitions into a small-amplitude difference to Vin where it
follows the earlier describes exponential relation.
B. Procedure
For this experiment, we set up the amplifier as a unity gain follower by connecting the output voltage to
the inverting input V2 . We then loaded the circuit at the output with a 1nF capacitor. We tested the circuit
5
0.8
0.6
0.4
0.2
[A]
0
out
I
−0.2
−0.4
−0.6
−0.8
Linear Fit
Experimental
−1
−0.4 −0.3 −0.2 −0.1 0 0.1 0.2 0.3 0.4
V [V]
dm
FIG. 5: The output current - differential-mode voltage characteristics of a simple MOS folded-cascode
differential amplifier.
response with both a small amplitude step and a large amplitude step. For the small amplitude step, we used a
2.3 kHz square wave with a peak-to-peak amplitude of 85mV and an offset of 2V as the input of the circuit. For
the large amplitude step we used a 3.2 kHz, 2V peak-to-peak amplitude square wave with an offset of 2V. For
each of these scenarios, we measured the input signal and output response. We then obtained the time constants
for the up and down step of the small signal response and the slew rate of the large amplitude response’s up
and down step.
Figure 6 shows the small amplitude step response of the unity gain follower circuit. After the exponential rise
of fall, the amplifier exhibits approximately linear behavior. The response for both the up and down step are
approximately symmetrical. The time constants for the up and down step are 2.64 × 10−5 s and 2.50 × 10−5 s.
The calculated time constant from the capacitance and differential-mode transconductance is 1.74 × 10−5 which
is adequately close to the experimental values.
Figure 7 shows the large amplitude step response of the unity gain follower. The qualitative behavior shows
that the circuit slews for most of the time until it rails. The response is adequately symmetrical. The slew rates
for the up and down steps are 7.8 × 103 V /s and 9.4 × 103 V /s, respectively. The calculated slew rates from the
load capacitance and the limiting values of the out put currents are 7.9 × 103 V /s and 9.3 × 103 V /s for the up
and down respectively. These match up to the measured slew rates well.
6
Unity−Gain Follower −
Small Amplitude Step Response
0.06
0.04
0.02
Voltage [V]
−0.02
−0.04
FIG. 6: The small amplitude step response to a simple MOS folded-cascode differential amplifier in a
unity-gain follower configuration.
7
Unity−Gain Follower −
Large Amplitude Step Response
0.5
Voltage [V]
−0.5
−1
0 1 2 3 4 5 6
Time [s] x 10
−4
FIG. 7: The large amplitude step response to a simple MOS folded-cascode differential amplifier in a
unity-gain follower configuration.