APPLICATION NOTES FOR TGS5042

an ISO9001 company

Application Notes for CO Detectors using TGS5042

The TGS5042 is a battery-operable
electrochemical CO sensor which is
provided with individual sensitivity
data printed on the sensors housing,
allowing users to eliminate the
process of calibration using CO
gas. This document offers example
circuits and important technical
advice for design and manufacture
of detectors.

Page

Sensor Marking.................................................................................2
Circuit Design

Basic Circuit.................................................................................2

Op-Amp Selection..........................................................3

Baseline Design of Sensor Output......................................................4
Microprocessor.......................................................................4

Anti-polarization Circuit........................................................4

Amplification Circuit..............................................................6

Amplification Factor (gain)................................................................6

Leak Prevention Circuit...................................................................7

Electric Noise Prevention....................................................................7

Temperature Compensation Circuit..........................................7

Self Diagnosis Circuit.....................................................7

Proposed Circuit Using Vcc=3V......................................................9

Proposed Circuit for 5000ppm CO Exposure Test........................10
PCB and Housing Design

Position Dependency of the Sensor..............................................12

Thermistor Location.....................................................................12

Housing Design for Quick Response..............................................12

Sensor Lead Configuration.............................................................12
Calibration

Calibration Using CO Gas..................................................13

Calibration Using Individual Sensor Data.....................................13

Temperature Compensation.........................................13

Calculation of CO Concentration...........................................14
Manufacturing Process

Handling and Storage of Sensors..........................................14

PCB Assembly.....................................................14

Sensor Assembly..........................................................15

F i n a l A s s e m b l y. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5

Gas Test.......................................................................15

Storage of Finished Products..............................................15
Packaging.................................................................16
Quality Control......................................................................................17
Expected Performance......................................................................................17
Frequently Asked Questions..............................................................................17
IMPORTANT NOTE: OPERATING CONDITIONS IN WHICH FIGARO SENSORS ARE USED WILL VARY
WITH EACH CUSTOMERS SPECIFIC APPLICATIONS. FIGARO STRONGLY RECOMMENDS CONSULTING OUR TECHNICAL STAFF BEFORE DEPLOYING FIGARO SENSORS IN YOUR APPLICATION AND,
IN PARTICULAR, WHEN CUSTOMERS TARGET GASES ARE NOT LISTED HEREIN. FIGARO CANNOT
ASSUME ANY RESPONSIBILITY FOR ANY USE OF ITS SENSORS IN A PRODUCT OR APPLICATION FOR
WHICH SENSOR HAS NOT BEEN SPECIFICALLY TESTED BY FIGARO.

TGS5042 is a UL recognized component in accordance with the

requirements of UL2034. Please note that component recognition testing
has confirmed long term stability in 15ppm of CO; other characteristics
shown in this brochure have not been confirmed by UL as part of component
recognition.

Revised 04/13

APPLICATION NOTES FOR TGS5042

1. Sensor Marking

One dimensional bar code

The TGS5042 comes with a sticker attached to

the sensor housing (see Fig. 1) which contains
individual sensor data:

FIGARO
TGS5042

One dimensional bar code

xxxx
where: xxx = [Link] nA/ppm

041124

1827

Lot No.

Sensitivity to CO (nA/ppm)
(Ex.1827 = 1.827nA/ ppm)

This bar code indicates the sensors sensitivity

(slope) in numeric value as determined in
Figaros factory by measuring the sensor s
output in 300ppm of CO. This value is also
printed below the bar code--please note that
three decimal places should be added to the
sensitivity reading (e.g. 1827 should be read as
1.827 nA/ppm).

Fig. 1 - Sensor markings

C1

Sensor lot number is printed to the left of the

sensitivity data in YYMMDD format.

1F
R1

2. Circuit Design
2-1 Basic circuit
TGS5042 is a fuel cell type electrochemical sensor
with two electrodes, with sensor output current
changing linearly with CO concentration. To use
the sensor for CO detection, it is necessary to
convert sensor current to output voltage. There
are two conversion methods:
2-1a Sensor current type (see Fig. 2a)
This method directly converts sensor current into
voltage according to the following equation:

1M
IC
+

Working
TGS5042

Vout

Counter
Vout = I x R1
Fig. 2a - Basic circuit for
amplifying sensor current

Vout = Is x R1
where: Is = Sensor output current
2-1b Load resistor type (see Fig.2b)
This method measures generated voltage across
a fixed load resistor which is connected to the
sensor electrodes. In this case, sensor output is
expressed by the following equation:
Vout = Is x R2

Working
TGS5042

R2
100k

Vout

Counter
Vout = I x R2
Fig. 2b - Basic circuit with
fixed load resistor

where: Is = Sensor output current

The load resistor type circuit does not include an
op-amp. An op-amp is used to amplify voltage
since sensor current is very small, in which case
Revised 04/13

the circuits of the sensor current type and load

resistor type become very similar. However,
large differences in characteristics can be seen
2

APPLICATION NOTES FOR TGS5042

between amplifying current and using a fixed
load resistor.
1) Response time to CO is slower when using
a fixed load resistor (see Fig. 3). When using a
fixed load resistor, the larger the value of the
load resistor, the slower the response time. In
addition, the expected output voltage may not
be obtained if a larger than 5.6k load resistor
value should be selected.
2) When amplifying current, an additional load
resistor or FET is required for anti-polarization
of the sensor.
Note 1: Please pay attention to sensor polarity.
Although the sensors package is physically
similar to that of a dry battery, the sensors
polarity is opposite to that of a dry battery.
Note 2: When voltage is applied to the sensor
output terminal, the sensor may be damaged.
Voltage applied to the sensor should be strictly
limited to less than 10mV.

3.5

Op-Amp type
Resitor type (5.6k_)
Resitor type (1.0k_)

CO400ppm
3

2.5

CO 150ppm
2

CO 70ppm

Vout (V)
1.5

CO 30ppm

0.5

0
0

200

400

600

800

1000

1200

1400

Time (sec.)

Figure 3 - Sensor response curve in various

basic circuits (amplified 3.13M times)

Figure 4 - Internal equivalent circuit when using an op-amp

2-2 Op-amp selection

When using a fixed load resistor, in most cases
an op-amp is required to amplify the sensors
small output voltage.
Rail-to-Rail Op-Amps such as the following
are recommended for both basic conversion
circuits:
AD708, AD8698, OP07 (Analog Devices),
TLC272(TI), OPA177(BB), MCP6042 (Microchip),
MCP616 (Microchip), OPA2355 (TI)
For circuits employing an op-amp, Figure 4
shows the internal equivalent circuit which can
be expected according to Figaros test results.
Component values are estimated as follows,
including variation in sensor characteristics and
temperature/humidity dependency:
Rs: 0~1k
Rs+Rp: 10k~
C: 0.8mF~10mF
Off-set voltage of an op-amp (current to voltage
converter) affects the base line level of sensor
output. This is because a small current reaches
the sensor due to the offset voltage of the opamp, causing a voltage shift according to the
Revised 04/13

C1
VR2

R2
100k

Working

IC

TGS5042

Vout

Counter

VR1

VR1: Variable resistor for zero adjustment

VR2: Variable resistor for span adjustment

Figure 5 - Circuit example for zero adjustment

internal DC resistance of the sensor (Rs+Rp).

The voltage shift drives an increase and decrease
from default base line level due to the value ()
offset voltage for the op-amp. To prevent such
a voltage shift, the above listed op-amps are
recommended.
Please note for circuits employing an op-amp,
when a fixed load resistor is used for antipolarization, only an op-amp with small leak
current (such as AD708 , OPA2355) can be used.
3

APPLICATION NOTES FOR TGS5042

When an op-amp with large leak current is used,
offset voltage may fall outside of the adjustable
range by leak current. (see 2-5 Anti-polarization
Circuit).
To obtain high accuracy in an analog circuit, an
op-amp with a zero adjustment function, such
as LF356 (National), is recommended. In Figure
5, a typical zero-span adjustable circuit using
LF356 is shown.

C1
1F
R2
1.2M

Working

IC
AD708

R1
1k

TGS5042

Vout

Counter

2-3 Baseline design of sensor output

Since sensor output may have a negative value
due to the offset voltage of the op-amp, the
baseline of sensor output should be set >1V. If
an anti-polarization circuit (using JFET), and a
self diagnosis circuit (see Items 2-6 and 2-12) are
added, the base line of sensor output should be
set >2V.

Figure 6 - Anti-polarization circuit using

a fixed resistor

C1
1F
R2

2-4 Microprocessor
Incorporating a microprocessor into a circuit
offers several advantages. The complex
calculation of alarm concentration based on
COHb and temperature compensation can be
carried out by the microprocessor, simplifying
circuit design. By recording sensor sensitivity
data from the sensors bar code label in the
microprocessor and by using the microprocessor
to compensate for offset voltage, the calibration
process for detector production can be greatly
simplified. In addition, useful values such as
a maximum CO concentration and a sensor
output timing chart can be recorded in the
microprocessor as additional functions.

Vcc

Working

1.2M

SW

Revised 04/13

AD708

TGS5042

Vout

Counter

Figure 7 - Anti-polarization circuit using

an external switch

C1
1F

From the viewpoint of signal resolution, the

recommended specification of a microprocessor
is 10 bit or higher.
2-5 Anti-polarization Circuit
When the sensor is stored without connection
between the working electrode (W) and counter
electrode (C), polarization will occur between the
electrodes. When a polarized sensor is connected
to an operating circuit, it takes a long time to
stabilize sensor output (refer to 3-5 Influence of
Storage in TGS5042 Technical Information).

IC

R2
1.2M

Working
TGS5042

JFET
(N-Channel)

IC

Vout

Counter

1.5V or more of baseline is required

for baseline to operate JFET

Figure 8 - Anti-polarization circuit using

an N-channel JFET

APPLICATION NOTES FOR TGS5042

To avoid polarization during storage, it is
necessary to keep a short-circuit (or <1k
resistance) between the electrodes. In this
manner, measures such as a timed warm-up
period for stabilizing the sensor are not required
in circuit design.
In a load resistor type circuit, it is not necessary
to add additional parts for anti-polarization
since the circuit already contains a load resistor
between the sensor electrodes.

C1

2) Fixed resistor (Fig. 6)

As a simple method, a fixed load resistor can be
used. In this case, it is necessary to use an opamp with very low leak current (such as AD708
or OPA2355). Op-amps with large leak current
may cause offset voltage to fall outside of the
adjustable range.
To avoid polarization on the sensor during
power-off, Figaro recommends that both of
electrodes of the sensor be shorted by using
a JFET (see Item 2-5.1 above). Alternatively,
adding a fixed resistor (10~100k) between
both electrodes will have a similar effect for
anti-polarization. However, since this effect
will be smaller than in case of using a JFET, the
recovery time after powering on may be longer.
Furthermore, resistors between both sensor
electrodes may cause a large shift from the base
amplified signal.
3) External switch (Fig. 7)
This method is normally used for CO analyzers
where there is an external switch to control
power on/off. By using an external switch, the
connection between the W and C electrodes can
be controlled, whether open or short (<1k)
circuit.
Revised 04/13

R2
1.2M

Working

TGS5042

IC

Vout

JFET
(P-Channel)

Counter

Figure 9 - Anti-polarization circuit using

a P-channel JFET

A sensor current type circuit can use three

methods for anti-polarization:
1) JFET (Figs. 8 and 9) - Figaros recommendation
This method is normally used in CO alarms
without an external switch. In this case a JFET
(Field Effect Transistor) is recommended.
Junction type transistors are not recommended
due to their large leak current. MOSFETs cannot
be used.

1F

Vcc

C1
1F
R2
3.0V
1.2M

Counter

TGS5042

JFET
(N-Channel)

3.0V of baseline is set

to operate JFET.
IC

Vout

Working
Vout decreases by a CO
concentration increases.

Figure 10 - Circuit for extending effective

voltage range using N-channel JFET

It is necessary to choose a P-channel or N-channel

FET, depending on the required effective output
range, gain, and the variation range of Vcc.
Figures 8 and 9 show the basic circuit for each
type.
The advantage of a P-channel FET is a wider
effective voltage range. However P-channel
FETs have higher cost, weakness against
applied voltage fluctuation, and they require
higher operating voltage than 5V (making them
unsuitable for applications which need <5V).
Recommended P-channel FETs are J177, J270
(Fair Child), and 2SJ103 (Toshiba).
N-channel JFETs are widely available in many
models, have lower cost, and are more stable
to voltage fluctuation of power supply. Their
disadvantage is their narrower effective voltage
range. However, one option to extend their
effective voltage range is to shift baseline to Vcc
as shown in Fig.10. In the example circuit, output
5

APPLICATION NOTES FOR TGS5042

voltage decreases with a CO concentration
increase. The recommended N-channel JFETs
are J201, PN4117 (Fair Child), and 2SK117
(Toshiba).

C1
22F
R2
100k
V1= - (I xR1)

2-6 Amplification Circuit (see Figs. 11 and 12)

In a load resistor type circuit, voltage can be
amplified by using either an inverted or noninverted amplifier. Depending on the method,
the gain is slightly different. Please note that
direction of W and C electrodes are opposite
between non-inverting and inverting amplifier
circuit.

TGS5042

Revised 04/13

1k
I

IC
AD708

R1
1k

Vout

Counter
Vout = I x R1 x (R2/R3)

Figure 11 - Inverting amplifier circuit

C1
22F
R2
100k
V1= I x R1
+
Counter

TGS5042

R1
1k

IC

R3
1k

Vout

Working
Vout = I x R1 x (1+R2/R3)

2-7 Amplification Factor (gain)

It is necessary to decide gain by selecting Vcc,
JFET and the op-amp in terms of sensor output
range, target gas concentration range, and
required accuracy.

To meet EN50291 requirements, a CO alarm

must be able to distinguish the output signal
(voltage) between 30ppm and 50ppm of CO.
For obtaining an appropriate gain value which
can meet with this requirement, the full scale of
CO concentration and baseline of output signal
in a CO alarm should be decided before starting

Working

Non-inverting amplification has lower leak

current from the sensor to the op-amp since the
working electrode is connected to the positive (+)
terminal of the op-amp. To prevent leak current,
setting a voltage follower between sensor output
and the op-amp is recommended (refer to Sec. 2-8
Leak Current Prevention Circuit).

For example, a gain of at least 2.5 million is

calculated to be necessary for the following
example CO detector:

- 5V Vcc

- 10 bit microprocessor

- Accuracy: 20% of reading

- Detection range: 0 ~ 750ppm
To extend the detection range, it is necessary
to increase Vcc or to use a microprocessor with
higher resolution. For conformity to EN50291,
Figaro recommends at least a 10bit A/D converter
with 1.2M to 1.4M times as a gain value.

R3

Figure 12 - Non-inverting amplifier circuit

C1
22F
R2

V1= - (I xR1)

100k
IC 1
+

Working
TGS5042

Current to Op-Amp
will be shut out.
R1
1k

R3
1k

IC 2
+
Vout

Counter
Vout = I x R1 x (R2/R3)

Figure 13 - Leak current measure

using a voltage follower

circuit design. For example, if a circuit voltage

(Vcc) would be 5V, the appropriate gain value
is obtained as follows:

- full scale of CO conc.: 1000ppm

- baseline of sensor output (in 0ppm CO): +1V
- max CO sensitivity of TGS5042: 2.4nA/ppm
- temp dependency of TGS5042: I(60C)/I(20C)=1.2

Gain: 1.38*106 -times =

[5V-1V] / [2.4*1.2 * 10-9(V/ppm) * 1000(ppm)]

In the above conditions, since a 10bit A/D

converter can distinguish the difference in sensor
6

APPLICATION NOTES FOR TGS5042

output between 30ppm and 50ppm CO, Figaro
recommends using at least a 10bit A/D converter
and 1.2M to 1.4M-times as a gain value in the
circuit.
2-8 Leak Current Prevention Circuit (Fig. 13)
In a fixed load resistor circuit, a non-inverting
amplifier circuit is recommended for simplicity.
For further countermeasures, adding a voltage
follower circuit between sensor output and the
op-amp is recommended.

To prevent external electrical noise, setting a

noise filter inside the detector, making a noise
prevention circuit pattern, and averaging the
output signal are recommended.
In addition to reduce electrical noise due to
radio waves, adding fixed resistors (R3 and R4
as shown in Figure 14) on the input terminal
of the op-amp is also effective. Please not that
the capacitor between the output and input
terminals (C2 on Figure 14) should not be added
when a CMOS type op-amp is used.
2-10 Temperature Compensation Circuit
Temperature compensation can be done in one
of two ways:
Revised 04/13

Figure 14 - Electrical noise prevention circuit

Temperature compensation
circuit

C5
+

C2
+

T22uF/16V
R7

T2.2uF/16V
R2

IC1A

AD708JN

R4

10k
R3

10k

AD708JN
+
-

IC1B

R8

10k

A
+

IC2

C3
3

39k

06P 10k
VR2

R6

1.5V

R9

10k at 20C
B: 3380

10.0k

2SK117

110k

RT

VR3
50k

0.1u
2

12k

DINS4

310k

VCC

9.1k

CN1
MB5P-90S

D1
C1
REG1

1
2
3
4
5

680JT

VCC

Tr 3

BATT1
UM3XI

1
2

VR1
R5 67W 5k R1

Three options to prevent electrical noise:

1) Use an electrical noise filter
2) Use a voltage follower (see Sec. 2-8 Leak Current
Prevention)
3) Build up an RC network with a resistor and
capacitor in place of the feedback resistor of the
op-amp. The recommended time constant (T)
of RC circuit for power input is 2~2.5 seconds
since the value in the basic recommended circuit
is 2.2 seconds.

T33u/16V

2-9 Electrical Noise Prevention

Since sensor impedance is 10 or less, the
sensor itself will not be a source of electric
noise. However, the sensor is easily influenced
by external electric noise since sensor output is
very small in both basic circuits. Therefore, it
is necessary to incorporate measures into the
circuit pattern and power supply to eliminate
electrical noise.

NJU7031D
6
C4

0.1u
VCC

RT: NTSA0XH103FE1B0 (Murata: Axial type)

VR1, VR2: Zero adjust
VR3: Span Adjust

Figure 15 - Temperature compensation circuit

1) Input the thermistor s signal into the

microprocessor and compensate it with the
temperature compensation table which is also
stored inside microprocessor (refer to Item
3-Calibration for details of the compensation process,
and refer to Appendix 1 for temperature compensation
factors).
2) Make an analog circuit with a negative
temperature coefficient (NTC) thermistor and
resistors in place of the feedback resistor of IC2
(see Fig.15)
2-11 Self Diagnosis Circuit (patented by Figaro)
Sensitivity to CO would be lost in case
several failure modes were to occur, such as
wire breakage, short circuit, or in case the
sensors water reservoir were to dry up. By
using Figaros patented self diagnosis circuit,
7

APPLICATION NOTES FOR TGS5042

malfunctions involving loss of CO sensitivity
can be detected.

the sensor from the circuit so that self test may

be conducted without activating an alarm.

Please note that this method cannot detect CO

sensitivity loss caused by lack of gas diffusion
when dust or water droplets cover the pin
holes for gas diffusion. In addition, slight loss
of CO sensitivity cannot be detected by selfdiagnosis.

2) Apply a minute current to the sensor

Activate a transistor (TR1) and apply about
1A (absolute maximum rating: 5A), which
simulates exposure to about 1000ppm CO. This
current should be applied for 2~5 seconds. If the
sensor is normal, sensor current will be output
and then quickly recover to its base level.

Depending on the users circuit design, factors

for self diagnosis such as current value, self
diagnosis period, measurement timing, and
voltage range for judgement may vary. Therefore,
it is recommended that experimentation with the
users circuit be conducted for fine tuning these
factors in self diagnosis.

3) Reconnect the sensor to the circuit

1~2 seconds after self diagnosis current is
terminated (TR3 is reset), activate TR2 and
reconnect the sensor to the circuit. Then current
applied to the sensor will be discharged.
4) Self diagnosis determination is carried out
Approx. 5 seconds after the sensor is reconnected
to the circuit, if the sensor output falls within the
range of 2.3 ~ 3.8V (normal output expected in
1000ppm of CO), which corresponds to 0.4 to
2A of sensor current, the sensor can be judged
to have normal CO sensitivity. If the sensor

Figs. 16, 17, and 18 show examples of the circuit,

timing chart, and process chart.
The basic steps of self-diagnosis are:
1) Temporarily cut the sensor off from the circuit
Activate a transistor (TR2) to temporarily isolate
Self test current control

Self test drive control

Tr 1
Microprocessor port2
H-Z on
0V off

2SK982
2M

R10

Microprocessor port1

C2

Vcc on
0V off

R11

1M

0.1

A
1
2
BATT1
UM3XI
Tr 3
2SK117

VCC1
C3
MCP6042

IC1

0.1
8

1
5

1.0k

0.1
2

2.0V
3.0k

VCC1
C3

680JT

33/50V
REG1

C1

VR1
R5 67W 10k R6

CN1
MB5P-90S

T1/16V

C4

1
2
3
4
5

D1
DINS4

1.2M

MCP6042

+
-

IC1
4

VCC

5V

R2
Tr 2
2SK117

Figure 16 - Self diagnosis circuit

Revised 04/13

APPLICATION NOTES FOR TGS5042

output is 3.8V, the sensor is judged to be short
circuited. If the sensor output is <2.3V, the sensor
is judged to be open circuited. Please refer to
Fig. 19 (see following page) for the Vout pattern
corresponding to each of these cases.
Note: The above Vout range is valid only
for the circuit shown in Fig.16. When gain
of amplification and/or measuring timing is
different, the Vout range would be changed.
Sensor output recovers to its initial level about
1 minute after the sensor judgement in the
step 4 above. The total self diagnosis time for
Steps 1-4 is 1~2 minutes. Please note that the
larger or longer current is applied to the sensor,
the longer it takes to complete self diagnosis.
The self diagnosis process should be carried
out periodically to ensure that the sensor has
sufficient CO sensitivity to afford protection.

H-Z (on)

0V (off)
approx. 1 min.
VCC (on)

Self test
current
control

0V (off)
approx. 5 sec.

Gain: 0.69*106 -times =

[3V-1V] / [2.4*1.2 * 10-9(V/ppm) * 1000(ppm)]

This small gain value may not meet EN50291

requirements due to the small signal resolution
Revised 04/13

0.01 - 1sec.

0.01 - 1sec.

Measuring point
5 - 7 sec. after changing to 0V for drive port

Figure 17 - Self diagnosis timing chart

Power on (Normal)

Drive control: H-Z (on)

Current control: 0V (off)

Normal mode

Drive control: H-Z (on)

Current control: 0V (off)

The recommended self diagnosis interval for

the circuit in Fig.16 is 180 seconds or more.
To shorten the interval, it is recommended to
minimize the current applied to the sensor (less
current/shorter duration). However, the smaller
the current, the more difficult to distinguish
between normal sensors and abnormal sensors.
Users should conduct a verification test using
their actual circuit.

Normal

Drive
control

Note:
Please restart normal operation mode when
sensor output recovers to its initial level after
the self diagnosis operation. The interval
between self diagnosis operations should be set
considering the recovery period for the sensor.
If current is applied to the sensor before it can
recover to its initial level, the sensor may be
damaged due to overcharging.

2-12 Proposed circuit using Vcc=3V

In case that the Vcc is +3V, the maximum gain
value can obtain 0.7M-times under conditions
which is described in previous section.

Self test
for sensor

Normal

Self test ?

NO

Yes
Drive control: H-Z (on) to 0V (off)
0.1 - 1 sec.
Current control: 0V (off) to VCC (on)
5 sec.
Current control: VCC (on) to 0V (off)
0.1 - 1 sec.
Drive control: 0V (off) to H-Z (on)
5 - 7 sec.
Measure the output voltage (Vout)

Yes
Vout = 2.3 - 3.8V?
NO
Trouble signal

Figure 18 - Self diagnosis flow chart

APPLICATION NOTES FOR TGS5042

Drive : H-Z
0V
Current: VCC
0V

5 - 7 sec.
5 sec.

Vout/V

4
Normal range
2.3 - 3.8 V

3
2
1
0

10

20

30
40
Time/sec.

50

60

Normal sensor
Drive : H-Z
0V
Current: VCC
0V

Drive : H-Z
0V
Current: VCC
0V

Vout/V

Vout/V

4
Normal range
2.3 ~ 3.8V

3
2

2
1

1
0

Normal range
2.3 ~ 3.8V

10

20

30

40

50

60

10

20

50

60

50

60

Drive : H-Z
0V
Current: VCC
0V

Drive : H-Z
0V
Current: VCC
0V

Vout/V

Vout/V

40

No sensitivity

Open circuit

Normal range
2.3 ~ 3.8V

3
2

Normal range
2.3 ~ 3.8V

3
2
1

1
0

30

Time/sec.

Time/sec.

10

20

30

40

50

60

10

20

30

40

Time/sec.

Time/sec.

Water/electrolyte dried up

Short circuit

Figure 19 - Self diagnosis Vout response patterns

between 30ppm and 50ppm CO. As a solution,

Figaro suggests the following three types of
circuits:

during the powering-on period. As a result, the

acceptable detection range of sensor output falls
between 0 and 3V.

1) Gain switching circuit (Fig. 20)

The gain value of a second amplifier (IC2) is
switched by the micro controller according to
the output signal of the micro controller. By
setting a large gain value while the sensor is in
low CO concentration range, adequate detection
resolution is achieved even if Vcc is 3.0V.

3) Change-baseline circuit (Fig. 21a)

In the circuit shown in Figure x, the baseline
and gain values can be changed according to
application of voltage in V1 and V2. As shown in
Figure 21b, when Vcc of 3.0V and 0V are applied
to V1 and V2 respectively during a short time,
the baseline of sensor output changes to lower
voltage. Gain value is also increased before
applying V1 and V2. Thus, by controlling both
V1 and V2, a detector can achieve high detection
resolution when the detector is exposed to a low
concentration of CO.

2) Opposite-baseline circuit (Fig. 10)

In this circuit, the sensor is set up in the opposite
direction (i.e. the working electrode is connected
to the positive input terminal of the op-amp
rather than the negative). Since the voltage
between the source and gate of the JFET is fixed
at 3.0V only during power-on, the JFET can
make a short circuit for anti-polarization except
Revised 04/13

2-13 Proposed circuit for 5000ppm CO exposure test

Most electrochemical CO sensors show very
long recovery time after high CO gas exposure.
10

APPLICATION NOTES FOR TGS5042

Figure 20 - Gain switching circuit

As a result, detectors cannot meet the 5000ppm

CO exposure test of EN50291 (5.3.6) without
some countermeasure. TGS5042 displays faster
recovery time than other types of electrochemical
sensors, so it may be possible to pass the test.
However, if the amplified signal of the sensor
exceeds Vcc (max output level of op-amp) in
5000ppm CO due to large gain, the sensor output
in air after exposure of 5000ppm CO would drop
below the baseline for a while (undershoot),
temporarily causing reduced sensitivity to CO
which may lead to failure to meet ES50291.
As countermeasure to the phenomenon of
undershoot, the overflow current of the sensor
should be canceled during the period when
the amplified sensor signal exceeds Vcc. For
example, by installing four diodes in the
amplifier circuit (see Figure 22), the overflow

Revised 04/13

Figure 21a - Change baseline circuit

Figure 21b -Change of baseline sensor output

using Vcc-3.0V on V1 and Vcc-0V on V2

Figure 22 - Basic circuit condition using a diode

11

APPLICATION NOTES FOR TGS5042

current goes back to the sensor through the
diodes during the period when the amplified
signal exceeds 2.8V.
Alternatively, switching the gain of the amplified
output signal is also effective. Figure 23 shows
a sample circuit diagram which can change the
gain by using a FET.
A similar function can also done by Changebaseline circuit as described in Item 2-12.3
above. In both of these cases, Figaro recommends
a gain value of 100,000 to 200,000-times when the
amplified output signal is higher than Vcc.

Figure 23 - Circuit for changing gain using a FET

3. PCB and Housing Design

3-1 Position Dependency of the Sensor
TGS5042 has no position dependency in normal
usage such as in residential CO detectors.
However, for applications where ambient
temperature can change drastically and suddenly
to less than -20C, it is recommended that the
sensor should be placed in a vertical position
with the working electrode upward. If the
sensor is positioned horizontally or vertically
with the working electrode down, the sensor
may be structurally damaged by large volume
expansion if water in the reservoir freezes.
3-2 Thermistor Location
It is recommended that a thermistor is located
as near to the sensor as possible in order to
accurately measure ambient temperature around
the sensor.
3-3 Housing Design for Quick Response
For applications where quick response is
required, such as for simple CO analyzers, it is
recommended that the gas inlet of the sensor be
located at the detector slits or opening. It is also
recommended to make a small compartment
with slits in at least two sides (see Fig.24).
3-4 Sensor Lead Configuration
There are two lead configurations for TGS5042.
The best version will depend on the user s
application.
Revised 04/13

1) Sensor compartment

2) Slits

Figure 24 - Sensor compartment design

12

APPLICATION NOTES FOR TGS5042

TGS5042-A00: Sensor with SUS lead pins
The hard pins enables the sensor to be mounted
directly to the PCB, simplifying the assembly
process.
TGS5042-B00: Sensor with flexible nickel ribbon
The flexible nickel ribbon allows for a variety of
methods for connection to the PCB. This type
is also suitable for insertion into a socket. The
flexible nickel ribbon may be broken by strong
mechanical shock, drop, or vibration, so it is
recommended that the sensor body be affixed
onto the PCB by using two-sided tape or wire,
for example.
4. Sensor Calibration
4-1 Calibration Using CO Gas
1) After powering the circuit, wait 5 minutes to
stabilize sensor output in clean air (see Note 1)
2) Measure sensor output in clean air (V0) (See
Note 2)
3) Inject C1ppm of CO gas
4) After stabilizing sensor output (e.g. 3 to 4 min),
measure sensor output (V1)
5) Calculate sensor sensitivity from V0 and
V1 values:
= (V1-V0) / C1
Using this method, accuracy of 5% can be
obtained for display readings. Please note that
temperature should be in the range of 20C2C
during the calibration process since the sensor
has dependency on temperature.
Note 1: In principle, due to the nature of
electrochemical cells, pre-heating before
calibration is not required. However, in actual
manufacturing, it is recommended to wait 5~10
min. before calibration to stabilize sensor output
in the circuit.
Note 2: If CO gas is present during the zero
adjustment process, a correct zero adjustment
cannot be carried out. A detector should be
checked in advance to verify that it generates

Revised 04/13

output corresponding to CO a concentration less

than 10ppm after subtracting detector output
without sensor.
4-2 Calibration Using Individual Sensor Data
Using individual data printed on sensor, which
is measured at Figaro factory before shipping,
can considerably simplify the calibration
process. Though the expected accuracy of 15%
accuracy in this method is less than that for using
CO gas, this method can achieve significant
reduction in handling costs while achieving
acceptable accuracy.
4-2-1 Input sensitivity data into microprocessor
Sensor data from the label can be read into the
microprocessor in one of two ways:
1) Manually input the user readable value on the
label, located beneath the one dimensional bar
code (this value is nA/ppm and contains three
decimal places).
2) Using a barcode reader which can read Code128, read the one dimensional barcode and input
directly to the microprocessor (this value is nA/
ppm and contains three decimal places).
4-2-2 Compensation of offset voltage (zero
adjustment)
To compensate for offset voltage which is created
by the sensor and operational amplifier, measure
the offset voltage (V0) in clean air (0ppm of CO)
and write into an EEPROM or a microprocessor.
This value should be read from the finished
detector (after installation of sensor, op-amp,
etc.).
To obtain higher accuracy, keep ambient
temperature in a range of 2010C and be sure
that the ambient air is completely free of CO.
4-3 Temperature Compensation
There are two methods for temperature
compensation:
1) Using a microprocessor
In case of using a microprocessor, it is necessary

13

APPLICATION NOTES FOR TGS5042

to read the thermistor output and write it into
the microprocessor. Inside the microprocessor,
temperature compensation is carried out by
using the compensation coefficiency table shown
in Appendix 1. Temperature compensated CO
sensitivity (t) is calculated by the following
equation:
t= /CF
where: CF = compensation coefficient at certain
temperature
2) Without using a microprocessor
In case of not using a microprocessor, this
process can be eliminated.
4-4 Calculation of CO Concentration
CO concentration can be calculated by using
sensor output (Vout), offset voltage (V0),
temperature compensated CO sensitivity (t), and
gain (A) in the following formula:
C = (Vout-Vo)/A/t

Equation 1

30

NJU7034

10

Display reading (ppm)

-10

-20

-30
-20

-10

10

20

30

Fig.26 shows basic flow chart of signal processing.

5. Manufacturing Process
5-1 Handling and storage of sensors
Prior to usage, sensors should be stored in
Figaros original sealed bag under conditions
of 5~30C, 30~80%RH, and avoiding dew
condensation for a maximum period of 6
months. Do NOT use a moisture proof bag (such
as an aluminum coated bag) for prevention of
dew condensation. When the sensor is shipped
from Figaro, the sensor electrodes are short
Revised 04/13

40

50

60

70

Temperature (C)

Figure 25 - Temperature dependency of offset voltage

for op-amps
Power ON
Alarm dalay againt
polarity of sensor

Generate trouble
signal

Yes

Sensor trouble
detection
No

Depending on the op-amp, offset voltage has

a large temperature dependency as shown
in Fig.25. To compensate the temperature
dependency, it is recommended to make a table
of offset voltage at different temperatures Vo(T)
in the microprocessor, and Vo in Equation 1
should be replaced to Vo(T) in Equation (1).
C = (Vout-Vo)/A/t Equation 1
Actual gain (A) should be measured instead
of calculated or specified at a theoretical value
since such value may not be obtained in actual
measurement.

AD708

20

Sensor output
sampling
Thermistor
output sampling
CO calculation
Display CO
concentration
Convert to COHb
concentration

Activate CO alarm

Yes

Alarm
determination
No
Suppress CO alarm

Figure 26 - Signal processing

flowchart

circuited. Please maintain this condition during

storage. If stored in open circuit condition,
sensor polarization will occur, requiring a long
warm up period for stabilization of output.
5-2 PCB assembly
Flux should be sufficiently dried before sensors
are assembled onto the PCB to avoid any
contamination of the sensor by flux vapors.
14

APPLICATION NOTES FOR TGS5042

No.

Soldering Material
Model

Composition

Melting Temp

Company

Model

H63E

Sn/37Pb

183C

Tamura Kaken Corp

ULF-20
(Cl free)

H63E

Sn/37Pb

183C

Asahi Chemical Research

Laboratory Co., Ltd.

AGF-780R

H63A

Sn/37Pb

183C

Asahi Chemical Research

Laboratory Co., Ltd.

AGF-550BK

LLS219AB18
(Pb free)

Sn/3.0Ag/0.5CuNi/
Ge

Soldus line: 271C

Liquidus line: 221C

Tamura Kaken Corp

EC-19S-8

Company

Flux

Solder Coat
Co. Ltd.

Table 1 - Wave soldering materials

5-3 Sensor assembly

The sensor electrodes of model -B00 are shortcircuited by the Nickel ribbon when the sensors
are shipped. Before mounting the sensors on a
PCB, the Nickel ribbon of model -B00 should
be cut. Model -A00 is shipped in an opencircuited condition, so the sensor output should
be stabilized by shorting the sensor (refer to FAQ
for advice on short-circuiting). Both models -A00
and -B00 can be directly soldered onto a PCB.
In case the B00 model is inserted into a socket,
it is recommended to cut Ni ribbon as short as
possible from the welded part. Do NOT peel
away the Nickel ribbon with strong force at the
working electrode side. The mechanical force
may damage the sensor electrode.
The metal ribbon of the -B00 model is for
electrical connection and therefore should not
be used to affix the sensor to a PCB. For securing
the sensor to a PCB and to prevent disconnection
of the Ni leads, the sensor should be attached to
a PCB using wire, two-sided tape, etc.
Recommended conditions for manual
soldering:
Temperature of soldering copper head: 380C
Period: < 10 sec.

Figaro has confirmed that wave soldering can

be done by using the materials shown in Table
1. When different materials will be used, a test
should be conducted before production starts
to see if there would be any influence to sensor
characteristics.
5-4 Final assembly
Revised 04/13

Avoid any shock or vibration which may be

caused by air driven tools. This may cause
breakage of the sensors lead wires or other
physical damage to the sensor.
5-5 Gas test
Test all finished products in the target gas
under normal operating conditions. Keep the
atmospheric conditions in the chamber stable,
utilizing a user-defined standard test condition
which is based on applicable performance
standards and on anticipated usage for detectors.
Remove any traces of smoke, adhesives, gases,
or solvents from the chamber.
Do NOT use Nitrogen balanced CO gas. Oxygen
molecules are required for the reaction of the
sensor with CO (refer to Sec. 2-Operation Principle
of TGS5042 Technical Info).
During exposure to a mixture of CO and N2,
the sensor reacts to CO by consuming oxygen
molecules inside sensor. After consuming all the
oxygen molecules inside sensor, the sensor will
not react to CO.
Dry/bottled CO gas can be used since the
sensors humidity dependency is very small.
NOTE: Without testing after final assembly,
detectors have no guarantee of accuracy or
reliability.
5-6 Storage of Finished Products
Detectors should be stored in a clean air
environment at room temperature. Avoid
storage in dirty or contaminated environments.
Avoid storage in extremely low humidity-15

APPLICATION NOTES FOR TGS5042

sensor life may be shortened. Please refer to Sec.
6-Notes in TGS5042 Technical Info for additional
information.
5-7 Packaging
When a plastic clam shell is used for packaging
of a CO detector, a small charcoal filter should
be placed inside to avoid potential influence by
organic vapors generated from the package.
Never expose the sensor to a vacuum. Sudden
exposure to a vacuum may temporarily damage
the sensor.

Acceptance of sensor

Read out
sensitivity data

Acceptance of components

PCB Assembly

Repair

NG

Circuit test
OK
Input sensitivity data to
EEPROM
Sensor Assembly
Zero adjust

6. Quality Control
1) A sample of finished products from each
production lot should be tested to confirm alarm
concentration. Check whether these samples are
acceptable for shipment and maintain a record
of these tests.
2) Periodically sample a certain number
of finished products to confirm the alarm
concentration under extreme conditions (e.g.
-10C or 40C/85%RH) and maintain a record
of these tests.
3) Periodically sample a certain number of
completed products periodically to confirm their
long-term characteristics and maintain a record
of such tests.
7. Expected Performance
Considering sensor variation as well as the
tolerances of electric components such as the opamp and thermistor, display accuracy of 20%
can be expected when individual bar code data
of TGS5042 is used for calibration. This level
of accuracy may not be obtained if low quality
components are used. For higher accuracy, gas
calibration for each sensor is recommended.
8. Frequently Asked Questions
Q: What approvals does TGS5042 have?
A: The sensor has received UL2034 component
recognition.
Q. How long is the expected life of TGS5042?
A: Unlike some electrochemical sensors which
are short-lived, the expected sensor life is more
Revised 04/13

Final Assembly
(PCB,Casing,etc.)
Re-calibration
(Zero, Gas)

Gas test

NG

Function check

OK
Packaging
Storage
Shipping

Figure 27 - Manufacturing flow chart

than 7years under normal operating condition. The

expected output change is only 3% after 7 years on
average in normal air.
Q: Is it true that the accuracy of two-electrode
electrochemical CO sensors is less than that of threeelectrode type sensors?
A: While this may be true for sensors whose electrode
potentials are unstable, the TGS5042 exhibits good
accuracy. With an optimized sensor structure and
electrodes, TGS5042 maintains very stable electrode
potentials. As a result, the sensor shows excellent long
term stability, drifting by only 2% over 2000 days.
Q: Where does CO gas enter into the sensor?
A: There are three pin holes in the working electrode
which act as a gas inlet. Refer to Figure 1 on page 2
of the TGS5042 Technical Information.
Q: Does the sensor comply to RoHS restrictions?
A: Yes.
Q: How long does it take to stabilize sensor output
16

APPLICATION NOTES FOR TGS5042

after storage when the sensor electrodes are opencurcuited?
A: According to Figaros test results, short-circuiting
the sensor for one hour should be sufficient to stabilize
sensor output after as much as 6 months storage in
fresh air. Nevertheless, the period to reset the sensors
polarization depends on the type(s) of gases existing
in ambient air, ambient conditions, storage period,
and the customers circuit. For achieving the best
results, Figaro recommends that users conduct tests
at their production facility to determine the optimal
stabilization period for sensor output in air.
Q: Are any special precautions needed to use
TGS5042 for a simple CO analyzer?
A: To obtain quick response, the gas inlet of sensor
of the sensor should be located nearest to the detector
slits or opening. For this application, TGS5042-B00 is
recommended. Please cut the Nickel ribbon as short
as possible and connect an insulated copper wire by
soldering. In addition, it is recommended to make
small compartment with slits at least in two sides
around the sensor. Refer to Sec. 3-3 Housing Design
for Quick Response.
Q: How can electric noise be prevented?
A: Since sensor impedance is 10 or less, the sensor
itself will not be a source of electric noise. However,

the sensor is easily influenced by external electric

noise since sensor output current is very small.
Power-supply noise should be minimized, an antielectrical noise circuit pattern should be made, and
a CR circuit should be used to prevent influence of
electrical noise.
If incoming noise is too large to be prevented by
the above measures, these additional steps should
be taken:
- coat the detector case with a copper board
- use anti-EMI material
- place metal mesh around the electric circuit
Figaro USA Inc. and the manufacturer, Figaro
Engineering Inc. (together referred to as Figaro)
reserve the right to make changes without notice
to any products herein to improve reliability,
functioning or design. Information contained in
this document is believed to be reliable. However,
Figaro does not assume any liability arising out of the
application or use of any product or circuit described
herein; neither does it convey any license under its
patent rights, nor the rights of others.
Figaros products are not authorized for use as critical
components in life support applications wherein a
failure or malfunction of the products may result in
injury or threat to life.

FIGARO GROUP

Revised 04/13

HEAD OFFICE

OVERSEAS

Figaro Engineering Inc.

1-5-11 Senba-nishi
Mino, Osaka 562-8505 JAPAN
Tel.: (81) 72-728-2561
Fax: (81) 72-728-0467
email: figaro@[Link]
[Link]

Figaro USA Inc.

121 S. Wilke Rd. Suite 300
Arlington Heights, IL 60005 USA
Tel.: (1) 847-832-1701
Fax.: (1) 847-832-1705
email: figarousa@[Link]

17

APPLICATION NOTES FOR TGS5042

Appendix 1 - Temperature Compensation Coefficients

Revised 04/13

Temp
(C)

CF
(I/Io)

Temp
(C)

CF
(I/Io)

Temp
(C)

CF
(I/Io)

-40

0.453

0.844

40

1.105

-39

0.463

0.852

41

1.109

-38

0.473

0.861

42

1.112

-37

0.483

0.870

43

1.115

-36

0.493

0.878

44

1.118

-35

0.503

0.887

45

1.121

-34

0.523

0.895

46

1.124

-33

0.513

0.903

47

1.126

-32

0.534

0.911

48

1.128

-31

0.544

0.919

49

1.130

-30

0.554

10

0.927

50

1.132

-29

0.564

11

0.935

51

1.134

-28

0.574

12

0.943

52

1.135

-27

0.584

13

0.950

53

1.136

-26

0.594

14

0.958

54

1.137

-25

0.605

15

0.965

55

1.138

-24

0.615

16

0.972

56

1.139

-23

0.625

17

0.980

57

1.139

-22

0.635

18

0.987

58

1.139

-21

0.645

19

0.994

59

1.139

-20

0.655

20

1.000

60

1.139

-19

0.664

21

1.007

61

1.139

-18

0.674

22

1.013

62

1.139

-17

0.684

23

1.020

63

1.139

-16

0.694

24

1.026

64

1.139

-15

0.704

25

1.032

65

1.139

-14

0.714

26

1.038

66

1.139

-13

0.723

27

1.044

67

1.139

-12

0.733

28

1.050

68

1.139

-11

0.742

29

1.055

69

1.139

-10

0.752

30

1.060

70

1.139

-9

0.761

31

1.066

-8

0.771

32

1.071

-7

0.780

33

1.076

-6

0.789

34

1.080

-5

0.799

35

1.085

-4

0.808

36

1.089

-3

0.817

37

1.094

-2

0.826

38

1.098

-1

0.835

39

1.101

18