Sie sind auf Seite 1von 5

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/4076525

Common mode stability in fully differential


voltage feedback CMOS amplifiers

Conference Paper · January 2004


DOI: 10.1109/ICECS.2003.1302033 · Source: IEEE Xplore

CITATIONS READS

11 725

4 authors, including:

Cristoforo Marzocca
Politecnico di Bari
122 PUBLICATIONS 1,566 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Yield modelling using the clustering effect View project

4DMPET View project

All content following this page was uploaded by Cristoforo Marzocca on 16 January 2015.

The user has requested enhancement of the downloaded file.


COMMON MODE STABILITY IN FULLY DIFFERENTIAL VOLTAGE
FEEDBACK CMOS AMPLIFIERS
Arturo Tauro*, Cristoforo Marzocca*, Francesco Corsi*, Antonio Di Giandomenico**
* Dipartimento di Elettrotecnica ed Elettronica, Politecnico di Bari, Italy
** Infineon Technologies, Villach, Austria

ABSTRACT may cause the set up of an erroneous d.c. operating


point at the amplifier power on.
In this study consideration is made of the common Fig.1 shows the schematic of a differential
mode (CM) behavior of two-stage fully differential voltage amplifier implemented with a fully
amplifiers in voltage feedback connection. In differential Opamp.
particular, interaction between internal CMFB One of the feedback loops, internal to the
circuitry and the external feedback network is Opamp, includes the CMFB circuitry, while the
investigated in order to highlight a possible other is made up of the external passive feedback
operating point instability condition, due to the network and the CM signal path of the Opamp.
presence of a positive feedback loop. While the former achieves negative feedback, the
Two alternative CMFB topologies are latter is positive for a typical two-stage amplifier,
compared, one of which is affected by instability, and may cause instability.
with development of constraints on the feedback
network parameters to be fulfilled in order to avoid R2
instability. Comparison was made by circuit
+ -
simulations with reference to a low voltage, deep V’in
1 R1 3 5
Vout=V’in
sub-micron CMOS technology (0.12 µm) 2 R1 4 - + 6

implementation of a differential amplifier.


R2

Figure 1 – Fully-Differential voltage feedback Opamp.


1. INTRODUCTION
In order to distinguish between the two loops,
the former is referred to as ‘internal’, while the
Fully differential circuit structures are at present
latter is ‘external’. According to this definition, Tint
widely employed in low voltage (LV) applications,
and Tex represent corresponding loop gains while
due to their improved immunity to noise, larger
V’icm, Vicm and Vocm are CM voltages at nodes 1-2,
output swings and reduced distortion.
3-4 and 5-6 respectively, as indicated in Fig.1.
However, these benefits are obtained at the
expense of greater complexity, since differential
2. ANALYZING THE CMFB LOOP
amplifiers usually require internal common mode
feedback (CMFB) loop in order to stabilize their
The problem can be studied in its essential aspects
d.c. voltages at both internal and output nodes.
with reference to simple Miller Opamp architecture,
Although a great amount of work is currently
although analysis can easily be extended to more
underway which addresses this problem either by
complex situations including both telescopic and
optimizing the CMFB loop in terms of linearity,
folded cascode topologies, circuits or opamps. An
speed, power consumption, and so on [1,2], or by
initial CMFB topology is analyzed here which sets
trying to control the CM without employing any
up CM control by regulating gate voltage across
additional circuitry [3,4], no general solution is
M0, as shown in Fig.2. In this figure, ACMFB, which
presently available since the variety of different
is negative, represents the voltage gain of the
situations arising in practical applications requires
CMFB circuit, while M0 is in the saturation region.
special approaches.
Apparently, the amplifier has a non-inverting
This study shows that two distinct CMFB loops
common-mode gain, Acm, resulting as it does from
are generally active in a differential voltage
cascade connection of two inverting stages. As a
amplifier with internal CMFB circuitry [5] and
consequence, as has already been pointed out, a
identify the conditions for which their interaction
positive feedback effect arises when the external

0-7803-8163-7/03/$17.00 © 2003 IEEE ICECS-2003


288
differential feedback network is connected which
M7 M0/2
may affect overall amplifier stability. In fact, Vr2
A C MFB

R1
denoting the feedback factor by: f = , the
R1 + R 2
external CM loop gain, Tex = Acm·f, must be kept VoCM

smaller than unity so as to avoid instability. M1


ViCM

M3

M0
VoCM
A C MFB
M5
Vr1
M7 M8
Vr2

M1 M2
Vin- Vin+

Vout- Vout+
Figure 3 – CM Half circuit of the amplifier.

M3 M4

 R  R
Vr1
M5 M6 Vocm =  1 + 2  Vicm − 2 ⋅ VREF (2)
 R1  R1

Graphical representation of this relationship,


Figure 2 – Miller Opamp employing a CMFB topology with k= R2 / R1, is shown in Fig.4 (solid lines). The
regulating the gate voltage across M0. slope of these characteristics is 1/f. The dashed
curve refers to the static input-output CM
A cm0 characteristic of the Opamp.
In particular: A cm = , where: Acm0 is the CM
1 + Tint
The equilibrium condition must satisfy the
gain without the CMFB circuit and Tint is the
constraints imposed by both characteristics and
internal CM loop gain. Defining the impedances at
corresponds to the intersection of the two curves of
the drain of the cascode and output transistors by rp1
Fig.4.
and rp2 respectively, Tint can be calculated from the
Owing to the non-linearity of Acm, the
CM half circuit of the amplifier shown in Fig.3:
equilibrium condition may not be univocally
identified, also depending on the values assumed by
g m0
Tint = A CMFB rp1g m3 rp2 (1) the two resistors R1 and R2.
2
In particular, a range [0, k*] exists, such that three
distinct intersection points occur when 0 < k < k*,
As Acm0 has a low value due to the high CMRR of while a single equilibrium point exists for k > k*.
the Opamp, Acm will normally be very small.
For example, assuming VREF = 0.7 V and VDD =
However, with reference to the CM loop in
1.5 V leads to a value of k* = 1.67 (obtained by
Fig.2, an increase of the CM signal Vicm, at the
input of the amplifier, may lead the transistor M0 to setting Vocm = VDD and Vicm = VDD-VSG-VSDsat in eq.
operate in its triode region, in which case the circuit (2)). Thus even a simple, widely employed, voltage
assumes a nonlinear behavior since the M0 output buffer (k = 1) may be affected by the described
resistance is no longer constant. Reducing output operating point ambiguity.
resistance of the tail current source will cause a Analysis of Fig.4 appears to indicate that for k
decrease in CMRR and an increase in Acm0, with within the interval [0, k*], points 1 and 3
Tint also dropping owing to its proportionality to correspond to stable equilibrium conditions, since
gm0. As a result, Acm increases along with Tex. they refer to low values of Acm, while point 2 is
The amplifier d.c. operating point is determined by unstable, since it occurs in correspondence with a
the equilibrium condition resulting from the high Acm value. This is also apparent from
simultaneous action of both internal and external evaluation of the slope of the dashed line at point 2.
feedback loops. Under these conditions, in fact, Acm> 1/f or,
If V’icm = VREF = cost., the expression of the equivalently, Tex > 1.
static characteristic of the external feedback In order to check for the validity of the above
network is simply: conclusions, circuit simulations were carried out
with reference to a low voltage deep sub-micron

289
obtain Vicm = 1.1V which falls outside the CM input
Vocm
dynamic range of the Opamp, corresponding
VDD (Fig.5) approximately to the interval [0, 0.9V].
3 Simulation (Fig.6) confirms that the circuit leaves
the operating point 3 for V’icm = 0.3V, obtained by
1
2 evaluating VREF in eq. (2) for Vicm = 0.9V and Vocm
VREF
= VDD.
k=0
k=k*
0
VREF Vicm

Figure 4 – Static input-output CM characteristic of


the open loop Opamp (dashed line). Eq. (2) has been
plotted (solid lines) for two values of k: k=0 and a
generic value k*.

CMOS technology (0.12 µm) implementation of a


differential amplifier. Values for VDD and VREF are
respectively 1.5V and 0.7 V. A voltage buffer
configuration has been chosen: R1 = R2 = R, with
CMFB circuit assumed to have a unity gain. Figure 6 – Transient CM output voltage of the
feedback amplifier (solid line) and input signal V’icm
The operating point ambiguity is apparent from (dotted line).
analysis of the figure 5, which shows the CM static
input-output characteristics of the open loop
Opamp. 3. ALTERNATIVE CMFB TOPOLOGY

A dissimilar CMFB topology is obtained by


controlling the gate voltage across transistors M5-
M6, rather than the Vgs of M0. Fig.7 shows the CM
half circuit of the Miller Opamp of Fig.2 which
employs the above CMFB topology.
Evaluation of the internal CM loop gain Tint for
the circuit in Fig.7 leads to the following
expression:

Tint = A CMFB g m5 rp1g m3 rp2 (3)

In contrast to the previous case, Tint is now


Figure 5 – CM static characteristics of the open loop
independent from gm0, guaranteeing that Acm and,
Opamp. The dotted line corresponds to eq. (2) for k = 1.
consequently Tex, are within a relatively low value
A graphical estimation of k* from Fig.5, gives a even when high CM input voltage is applied which
value of 1.92, not too far from that obtained by reduces gm0. The value of ACMFB can be chosen so
manual analysis. as to obtain the desired value of Acm, even though it
Figure 6 shows the transient CM output voltage must also satisfy stability constraints referred to the
of the feedback amplifier corresponding to an input internal CM loop.
triangular signal V’icm. The output signal waveform
can be derived from the static characteristic in Reduction of Acm over a wider range of Vicm
Fig.5, since input signal frequency is small enough well above VREF results in the elimination of the
to be considered as a static stimulation. For V’icm instability condition since no multiple intersection
approximately equal to 1.2 V, the d.c. amplifier points will be allowed in the plots of Fig.4.
output sets to: Vocm = VDD (point 3 in Fig.4). Circuit simulations were again carried out under
Once in this erroneus operating condition, the the same conditions as in Figs. 5 and 6, but with the
circuit is unable to exit even when V’icm settles to adoption of the second CMFB network, with results
VREF. In fact, setting Vocm = VDD in equation (2), we presented in Fig.8 and Fig.9.

290
M7 M0/2
Vr2 Vr1

VoCM

M1
ViCM

M3

M5
A C MFB
Figure 8 – CM static characteristics of the open loop
Opamp employing the alternative CMFB network. The
dotted line corresponds to eq. (2) for k = 1.

Figure 7 – CM half circuit of the amplifier employing


the alternative CMFB topology.

Fig.8 shows that a single equilibrium point now


exists for Vicm = 0.7 V. This result is also confirmed
by transient analysis (Fig.9) since, in this case, the
circuit does not enter an erroneous operating
condition when stimulated with a full amplitude
triangular waveform.
However, for high values of the input signal, the
output CM voltage Vocm drifts from its correct
value, due to variation of the CM gain Acm, as is Figure 9 – Transient CM output voltage of the feedback
clearly in evidence looking at the slope of the solid amplifier (solid line) and input signal V’icm (dotted line).
line in Fig.8.
5. REFERENCES

4. CONCLUSIONS [1] J. F. Duque-Carrillo, “Continuous-Time Common-


Mode Feedback Networks for Fully-Differential
This study shows that two distinct feedback loops Amplifiers: a Comparative Study,” Circuits and
are generally operating in a differential voltage Systems, 1993., ISCAS ’93, 1993 IEEE
amplifier with internal CMFB circuitry. The International Symposium on, vol. 2. pp. 1267-1270,
conditions under which their interaction does not 3-6 May 1993.
[2] P. M. VanPeteghem, J. F. Duque-Carrillo, “A
allow for univocal determination of the d.c.
General Description of Common-Mode Feedback
operating point of the circuit have been identified. in Fully-Differential Amplifiers,” Circuits and
In particular, we related the presence of instability Systems, 1990., IEEE International Symposium on,
to the drop of the internal CM gain Tint for certain vol.. 4. pp. 320–312, 1-3 May 1990.
values of the input CM signal, which, in turn, [3] P. D. Walker and M. M. Green, “An Approach to
increases the value of Acm and Tex. Fully Differential Circuit Design without Common-
When comparison of the two different Mode Feedback,” IEEE Trans. Circuits Syst., vol.
topologies of CMFB networks was carried out, only 43, pp. 752-762, Nov. 1996.
one solution was shown to be affected by [4] G. Nicollini, F. Moretti and M. Conti, “High-
instability. The validity of this analysis was Frequency Fully Differential Filter Using
Operational Amplifiers Without Common-Mode
checked by simulation with reference to deep sub-
Feedback,” IEEE J. Solid-State Circuits, vol. 24.
micron CMOS technology implementation. pp. 803-813, Jun. 1989.
[5] P.J. Hurst, “Determination of Stability Using
Return Ratios in Balanced Fully Differential
Feedback Circuits,” IEEE Trans. Circuits Syst., vol.
42, pp. 805-817, Dec. 1995.

291
View publication stats

Das könnte Ihnen auch gefallen