AN38

Application Note 38
Basic introduction to the use of the ZNI1000
Nickel Temperature Sensor
Howard Mason, September 2003
This application note describes the ZNI1000
temperature sensor, which uses a nickel
element connected between 2 pins of a SOT23
package to measure temperature accurately.
Nickel has a very precisely known resistance
versus temperature characteristic and this is
expressed as a polynomial. The element is
designed to be 1000⍀ at 0°C. By measuring the
resistance, temperature can be measured to
within ± 2°C down to -55°C and to within ± 1.5°C
up to +150°C.
The element is connected between pins 1 and 2
of a standard SOT23 package. Pin 3 is not
connected electrically, but has very good
thermal contact with the element. This means
it can be soldered to the object whose
temperature is being measured, thus following
changing temperatures quickly and accurately.
The resistance can be expressed as follows:
RT = R0 * (1 + A*T + B*T2 + C*T4 + D*T6)
where
RT is the resistance at temperature T
R0 is the resistance at 0°C (1000⍀)
A = 5.485 * 10-3
B = 6.650 * 10-6
C = 2.805 * 10-11
D = -2.000 * 10-17
Clearly, for most applications, the sixth order
term can be omitted. For example, even at
+150°C, it only decreases the resistance, which
is almost 2000⍀ at that temperature, by 0.2⍀.
The curve of resistance versus temperature is
shown in Figure 1.
Figure 1. Resistance of ZNI1000 versus temperature in degrees Celsius
ISSUE 1 - SEPTEMBER 2003
AN 38 - 1
SEMICONDUCTORS
In addition to the polynomial, there is a
specification of the deviation in degrees.
For temperatures from -55°C to 0°C, the
deviation in °C is:
Error ° = ± (0.4 + (0.028 * |T|) °C.
Note the modulus signs around the T, so that
the two error terms still ADD despite T being
negative - they do NOT cancel.
For temperatures from 0°C to +150°C, the
deviation in °C is:
Error ° = ± (0.4 + (0.007 * T) °C Again
the two error terms ADD.
At first sight enabling a readout of temperature
versus resistance measured appears to require
a look-up table, but a technique can be used
which "linearises" the output and makes
readout much simpler.
If the element is put into a bridge circuit which
is driven by a constant current then, as the
bridge resistance increases, the applied bridge
voltage increases and this applies a non-linear
term to the Vout versus Resistance equation
which causes the higher order terms to be
partially canceled out.
The bridge circuit is shown in Figure 3.
This produces a Deviation graph as shown in
Figure 2.
Figure 2. ZNI1000 Error in Degrees Celsius versus Operating Temperature
ISSUE 1 - SEPTEMBER 2003
SEMICONDUCTORS
AN 38 - 2
The constant current chosen is shown as 6mA
in the circuit, but is strictly 6.0623mA to give
exactly 1 volt output at +100°C. The 500⍀
trimpot is adjusted so that the output is exactly
1 volt at +100°C and the DVM will then read
with ± 2°C over the range -55°C to +150°C.
The error introduced by this circuit by
assuming that Vout is linearly proportional to
temperature is only -0.8°C at -55°C and only
+0.2°C at +150°C, obviously being zero by
definition at 0°C and zero by calibration at
+100°C. The worst case error between 0°C and
+100°C is 0.033°C in addition to the error due to
the deviation formula above, which gives 1.1°C
at 100°C, so within the 0°C to +100°C range all
the errors only add up to ± 1.2°C.
It should be noted from the spec that the
maximum continuous current is specified as
4mA. This is because it is necessary to make the
nickel element very thin to obtain 1000⍀
resistance as nickel has a very low resistivity. In
practice the user should keep the current as low
as possible because of the thermal resistance of
a SOT23 package. This will not be appreciable if
the unit is soldered down to a metal tab, as this
will always ensure that the nickel element is at
the correct temperature. However, if the unit is
used in free air, the thermal resistance will be
about 200°C/Watt, so the current should really
be kept down to 2mA to keep the errors due to
self-heating to below 1°C.
Figure 3. ZnI1000 and ZMR500 used with a DVM as a Thermometer
ISSUE 1 - SEPTEMBER 2003
AN 38 - 3
SEMICONDUCTORS
© Zetex plc 2003
Americas
Asia Pacific
Zetex GmbH
Streitfeldstraße 19
D-81673 München
Zetex Inc
700 Veterans Memorial Hwy
Hauppauge, NY 11788
Germany
Telefon: (49) 89 45 49 49 0
Fax: (49) 89 45 49 49 49
[email protected]
USA
Telephone: (1) 631 360 2222
Fax: (1) 631 360 8222
[email protected]
Zetex (Asia) Ltd
3701-04 Metroplaza Tower 1
Hing Fong Road
Kwai Fong
Hong Kong
Telephone: (852) 26100 611
Fax: (852) 24250 494
[email protected]
Europe
Zetex plc
Fields New Road
Chadderton
Oldham, OL9 8NP
United Kingdom
Telephone (44) 161 622 4444
Fax: (44) 161 622 4446
[email protected]
These offices are supported by agents and distributors in major countries world-wide.
This publication is issued to provide outline information only which (unless agreed by the Company in writing) may not be used,
applied or reproduced for any purpose or form part of any order or contract or be regarded as a representation relating to the products
or services concerned. The Company reserves the right to alter without notice the specification, design, price or conditions of supply of
any product or service.
For the latest product information, log on to www.zetex.com
ISSUE 1 - SEPTEMBER 2003
SEMICONDUCTORS
AN 38 - 4