M2.

Signal conditioning circuits, biopotential

amplifiers, and analog-to-digital conversion
M2.1. Why signal conditioning circuits
M2.2. First look at Operational Amplifiers (OAs)
M2.3. Inverting, follower and non-inverting OA configurations
M2.4. Common-mode voltage and difference-mode voltage
M2.5. Common-Mode Rejection Ratio (CMRR)
M2.6. Difference amplifiers
M2.7. Instrumentation amplifiers
M2.8. The Wheatstone bridge
M2.9. Current sources and transimpedance amplifiers
M2.10. The Schmitt trigger (Hysteresis)
M2.11. Square wave oscillator
M2.12. Active filters
M2.13. Analog integration and differentiation with OPAMPs
M2.14. Biopotential amplifiers
M2.15. …

BIS 2024-2025 M2.a: 1/38

 M2.1. Why signal conditioning circuits?

Stimulus

conversion into amplification vin #

sensor voltage signal + range shift filtering ADC

Vref #
Signal conditioning

vin

ADC input range

ADC input must be a voltage signal within a limited amplitude range; for instance from 0 V to
5 V. The task of the signal conditioning circuitry is to adapt sensor’s signal to that voltage
range so that maximum effective resolution can be achieved for a specified stimulus range.
BIS 2024-2025 M2.a: 2/38
 M2.2. First look at Operational Amplifiers (OAs)

An Operational Amplifier (OPAMP or, simply, OA) is a circuit able to largely amplify an
input differential voltage. It is also characterized by a very large input resistance and a very
small output resistance.

(external appearance of an
OPAMP; PDIP package)

Circuit symbol for an OPAMP with the five principal terminals.

BIS 2024-2025 M2.a: 3/38

 (source: “Electronic Circuits and Applications”, S. D. Senturia and B. D. Wedlock)
Linear amplification only occurs for a very limited
range of differential voltages at the input. Outside that
range the output saturates (either at ~VCC or at –VCC)

Circuit model for an ideal OPMAP

M2.a: 4/38
 iout

Note that the output energy in an OPAMP comes from the power supply and not
from the input.

BIS 2024-2025 M2.a: 5/38

Since an OPAMP has a very high differential gain, it can be used as a comparator: a
small voltage difference at input terminals will either saturate the output voltage at the
negative supply voltage or at the positive supply voltage.

A OPAMP used as comparator for detecting zero- A comparator can detect transitions through a
crossings. threshold voltage set with a voltage divider
(source: “Electronic Circuits and Applications”, S. D. Senturia and B. D. Wedlock) M2.a: 6/38
The linear gain factor of an OPAMP, 𝐴 , is always very large (>10000) but it has an
uncertain value. In order to use an OPAMP as a linear amplifier, negative feedback is
employed thus sacrificing potential gain but obtaining accuracy.

𝑣𝑂𝑈𝑇 = 𝑣+ − 𝑣− 𝐴

𝑣𝑂𝑈𝑇 = 𝑣+ − 𝐵𝑣𝑂𝑈𝑇 𝐴
𝑣+
+ 𝑣𝑂𝑈𝑇 1 + 𝐵𝐴 = 𝐴𝑣+
𝑣𝑂𝑈𝑇
𝑣− 𝐴𝑣+ 𝑣+ (implies gain
- 𝑣𝑂𝑈𝑇 = =ቚ
1 + 𝐵𝐴 𝐴≫1 𝐵 since B ≤1)

and,
B
𝑣+
= 𝑣+ − 𝑣− 𝐴 = 𝐴𝑣+ − 𝐴𝑣−
𝐵

Feedback network (B≤1) 1

𝐵 +𝐴 =ቚ 𝑣+
(The feedback network is a circuit that 𝑣− = 𝑣+ 𝐴≫1
virtually, 𝑣+ − 𝑣− = 0
produces accurate fraction of 𝑣𝑂𝑈𝑇 ) 𝐴
(key fact for easily analyzing OPAMP
circuits with negative feedback)
BIS 2024-2025 M2.a: 7/38
 M2.3. Inverting, follower and non-inverting OA configurations
Inverting amplifier

𝑖− = 0 → 𝑖𝑆 = 𝑖𝐹

𝑣𝑠 − 𝑣− 𝑣𝑠 − 𝑣− 𝑣𝑠
𝑖𝑆 = =ቚ =
(infinite input resistance) 𝑅𝑆 𝑣− =𝑣+ =0 𝑅𝑆 𝑅𝑆

𝑣− − 𝑣0 0 − 𝑣0 −𝑣0
𝑖𝐹 = =ቚ =
𝑅𝐹 𝑣− =𝑣+ =0 𝑅𝐹 𝑅𝐹

𝑣𝑆 −𝑣0 𝑅𝐹
= → 𝑣0 = − 𝑣
𝑅𝑆 𝑅𝐹 𝑅𝑆 𝑆
BIS 2024-2025 M2.a: 8/38
 Follower

Input impedance is very high (ideally infinite) and output impedance is very low (ideally
zero). This circuit is useful for “copying” the voltage from a network without disturbing it.

BIS 2024-2025 M2.a: 9/38

A follower is, for instance, used in the Arduino UNO board for “driving” a LED (Light
Emitting Diode) without drawing excessive current from the corresponding microcontroller
(μC) output (“SCK” wire=pin 18 of the μC and pin 13 of the Arduino board).

(http://arduino.cc/en/uploads/Main/Arduino_Uno_Rev3-schematic.pdf)
LED

μC

The microcontroller output would be able to drive by itself the LED (current ~ 5 mA). The follower is added here so as to
make the SCK line similar to the other output lines in terms of maximum current capability (~ 20 mA)

BIS 2024-2025 M2.a: 10/38

BIS 2024-2025
Non-inverting amplifier

(exercise: demonstrate)

M2.a: 11/38

(source: “Electronic Circuits and Applications”, S. D. Senturia and B. D. Wedlock)

BIS 2024-2025
Analog addition and subtraction

(exercises: demonstrate)

M2.a: 12/38

(source: “Electronic Circuits and Applications”, S. D. Senturia and B. D. Wedlock)

 Warning
Negative feedback does not guarantee linear operation; saturation of an OPAMP is always
possible.
Therefore, it is necessary to numerically verify that |𝑣0 |<|VSATURATION| in every presumably
linear circuit application. When saturation occurs the virtual equality (𝑣+ − 𝑣− = 0) is not valid.

BIS 2024-2025 M2.a: 13/38

 Another warning: OPAMP ground is not the external ground

The reference ground of the

OPAMP is the common voltage
(average) of the two power
supply voltages.

(two OPAMPs in a single integrated circuit

share the power supply terminals)

(source: “Electronic Circuits and Applications”, S. D. Senturia and B. D. Wedlock)

BIS 2024-2025 M2.a: 14/38

An example to illustrate the reference ground concept….

5V
In this case the reference ground of the
OPAMP is not

but

2.5 V

0V
that is, in case v+-v- is zero then the
output voltage will be + 2.5 V with respect
to the external ground.

BIS 2024-2025 M2.a: 15/38

 M2.4. Common-mode voltage and difference-mode voltage
Some signals – and particularly signals from sensors – do not come as voltages referenced
to ground (“single-ended”) but as differential voltages (two terminals) with an added
common voltage with respect to ground.

+ +
VD/2

V+ VC = common-mode voltage = (V+ + V-)/2

VD
+ VD = difference-mode voltage = V+ - V-

VD/2
VC
-

V-

BIS 2024-2025 M2.a: 16/38

 +
VD/2= VU/2

VU
+
VU ≡
VC=VU/2 VD/2= VU/2

Single-ended voltage 0V
(to ground)
Same voltage signal represented as a differential
signal with a common-mode voltage.

(However, typically VC>>VD)

BIS 2024-2025 M2.a: 17/38

 M2.5. Common-Mode Rejection Ratio (CMRR)

Common-Mode Rejection Ratio can be defined for any circuit that amplifies a

(source: “Electronic Circuits and Applications”, S. D. Senturia and B. D. Wedlock)

voltage difference, such as an OPAMP

BIS 2024-2025 M2.a: 18/34

 M2.6. Difference amplifiers
It is possible to build multiple circuits based on OPAMPs that are capable of accurately
amplifying a differential voltage signal rejecting its common-mode voltage (i.e high
(source: “Adquisición y distribución de señales” by R Pallás Areny)

CMRR). Here it is shown one of the simplest configurations containing a single OPAMP:

Common-mode gain Difference-mode gain

if then:

(However, this is not easy to accurately achieve

as resistors have tolerances and exhibit drift)
(In addition, as a negative aspect, this circuit has
limited input resistance) M2.a: 19/38
 +
VD/2

+
VC VD/2

Observe that in the previous difference amplifier the ground reference is defined by an
additional terminal; it is not defined from the OPAMP power supply terminals.

BIS 2024-2025 M2.a: 20/38

 M2.7. Instrumentation amplifiers
An instrumentation amplifier is a difference amplifier characterized by exhibiting very high-
input impedance, very-high CMRR (particularly at low frequencies) and accurate
difference-mode gain.

Implementation of an instrumentation amplifier

with three OPAMPs
(exercise: obtain common-mode gain and difference gain
assuming (R5/R4)=(R7/R6) and R1=R3)

Currently, an instrumentation amplifier (IA) can be

purchased as a single integrated circuit
component. This option is generally preferable in
terms of performance, however, it can be
significantly more expensive than a custom
developed IAs with low cost OPAMPs.

BIS 2024-2025 M2.a: 21/38

 M2.8. The Wheatstone bridge

An important group of resistive sensors has a linear output of this type:

R = R0 (1 + x ) = R0 (1 +  s ) R − R0 R
x =s = = −1
where s is the stimulus. R0 R0

In principle, output from these sensors could be read by injecting current and measuring
voltage. However, x is very small and, therefore, the overall resolution of the system
would be poor (i.e. small voltage variations over a large fixed voltage).

A simple voltage divider would not help in terms of resolution and would make things
more difficult to calculate x

 V0 
RA   RA
Vs  V0   VS − V0 
Vo R =   RA x= −1
 VS − V0  R0
R = R0 (1 + x )

BIS 2024-2025 M2.a: 22/38

In addition, it needs to be taken into account that all resistive sensors will exhibit a
dependence on temperature:

Thermal coefficient

R = R0 T (1 + x +  T ) = R0 T (1 +  s +  T )
0 0

where x may be much smaller than  T

So, there is a need for a technique for magnifying x and for cancelling the impact of the
temperature changes. And that’s what the Wheatstone bridge does…

BIS 2024-2025 M2.a: 23/38

 The Wheatstone bridge provides a differential-mode voltage proportional to x:

Vs R1 R2

- VD +
R4 R3 = R0 (1 + x )

(commonly used representation for

Wheatstone bridge configuration)

constant
 R3 R4  kx k
VD = VS  −  = VS = VS x
 R3 + R2 R4 + R1  k = R1 = R2 (k + 1)(k + 1 + x ) x ( k +1) (k + 1)(k + 1)
R4 R0

If R1,R2,R3 and R4 are made with the same technology and on the same substrate, their
thermal coefficients will be very similar and, therefore, temperature will not have an effect
on the output.
R1 R2 R1 (1 +  T ) R2 (1 +  T )
k= = = =
R4 R0 R4 (1 +  T ) R0 (1 +  T )

Thermal coefficient M2.a: 24/38

 Other configurations for the Wheatstone bridge provide some extra advantages:

Vs
R1 = R0 (1 + x ) R2 = R0
R1 R2
- VD + facilitates k= =
R4 R0
R4 = R0 R3 = R0 (1 + x )
(and increases sensitivity)

(preferred Wheatstone bridge configuration)

Vs
R1 = R0 (1 + x ) R2 = R0 (1 − x )
Linear output
- VD +
R4 = R0 (1 − x ) R3 = R0 (1 + x ) VD = VS x

BIS 2024-2025 M2.a: 25/38

 Strain gauges (galgas extensiométricas) are typical examples of the resistive sensors that have
been discussed in previous slides.

R = R0 (1 + x ) = R0 (1 +  s )

(source: “Adquisición y distribución de señales” by R Pallás Areny)

(image source: http://en.wikipedia.org/wiki/Strain_gauge)

strain gauge

load cell

strain gauge

BIS 2024-2025 M2.a: 26/38

Strain gauges are used in pressure sensors (built within, or attached to, a deformable membrane)

(visit this link for an animated demo on the functioning of strain gauge pressure sensors)
(http://www.rdpe.com/us/hiw-sgpt.htm)

BIS 2024-2025 M2.a: 27/38

About the pressure sensor used in the lab in the past:

10 kPa On-Chip Temperature Compensated Silicon Pressure Sensor by Freescale Semiconductor

BIS 2024-2025 M2.a: 28/38

Let’s go back and overview more stuff that can be done with OPAMPs…

BIS 2024-2025 M2.a: 29/38

 M2.9. Current sources and transimpedance amplifiers
Current sources

Howland current source

(the load is grounded)

iL

Exercise:

Exercise: verify that iL is independent of the

load impedance and equal to VREF/R if
R2R=R3R1

BIS 2024-2025 M2.a: 30/38

 Transimpedance amplifier
(aka current-to-voltage (I/V) converter)

(BPW21)

Photodiode current/voltage characteristic:

The transimpedance amplifier is typically used

to measure current produced by a photodiode

BIS 2024-2025 M2.a: 31/38

 (positive feedback)

(inverting)
M2.10. The Schmitt trigger (Hysteresis)

M2.a: 32/38

(source: “Electronic Circuits and Applications”, S. D. Senturia and B. D. Wedlock)

The Schmitt trigger is less likely than a comparator to bounce back and forth between states if a
waveform with many small ripples (i.e. noise) on it gradually passes through the transition level.

Exercise:
Show that T = (R1/R2)VCC

(non inverting Schmitt trigger)

BIS 2024-2025
 M2.11. Square wave oscillator

The Schmitt trigger is combined with a time delayed negative feedback

frequency is proportional to 1/C

Some sensors produce a change in its capacitance. That, for instance, would be the case of
most humidity sensors. Therefore, a square oscillator represents a good option to get
measurements from these sensors as the output signal is ready to be acquired by digital
systems, without need for an ADC; the digital system only needs to measure frequency.
BIS 2024-2025 M2.a: 34/38
 Exercise:
Calculate frequency (assume
this * is exactly 2.5 V)

BIS 2024-2025 M2.a: 35/38

 M2.12. Active filters
OPAMPs can used to build active filters (i.e. with gain) of high-order (i.e. step frequency
response). We are not covering this topic here because it is quite complex and, besides,
there are automated (and free) software tools that can design those filters for us. For
instance:
ANALOG FILTER WIZARD by Analog Devices:
https://www.analog.com/designtools/en/filterwizard/

BIS 2024-2025 M2.a: 36/38

The ANALOG FILTER WIZARD by Analog Devices has been used here to design a low-pass
filter (4th order) in order to remove the high frequency harmonics from a square signal so that
a sinusoidal signal is generated.

BIS 2024-2025 M2.a: 37/38

 M2.13. Analog integration and differentiation with OPAMPs

(source: “Electronic Circuits and Applications”, S. D. Senturia and B. D. Wedlock)

Integration is used for sensors whose output is in
the form of charge rather than current. That
would be the case of some extremely sensitive
photodetectors. M2.a: 38/38
BIS 2024-2025