AD22105
PRODUCT DESCRIPTION
The AD22105 is a single supply semiconductor thermostat
switch that utilizes a unique circuit architecture to realize the
combined functions of a temperature sensor, setpoint comparator,
and output stage all in one integrated circuit. By using one
external resistor, the AD22105 can be programmed to switch at
any temperature selected by the system designer in the range of
–40°C to +150°C. The internal comparator is designed to switch
very accurately as the ambient temperature rises past the
setpoint temperature. When the ambient temperature falls, the
comparator relaxes its output at a somewhat lower temperature
than that at which it originally switched. The difference between
the “switch” and “unswitch” temperatures, known as the hysteresis,
is designed to be nominally 4°C.
THE SETPOINT RESISTOR
added to the value found in the specifications table. For example,
consider using the AD22105 programmed to switch at +125°C.
Figure 4 indicates that at +125°C, the additional error is
approximately –0.2°C/% of R
SET
. If a 1% resistor (of exactly
correct nominal value) is chosen, then the additional error could
be –0.2°C/%
×
1% or –0.2°C. If the closest standard resistor
value is 0.6% away from the calculated value, then the total
error would be 0.6% for the nominal value and 1% for the
tolerance or (1.006)
×
(1.10) or 1.01606 (about 1.6%). This
could lead to an additional setpoint error as high as 0.32°C.
For additional accuracy considerations, the thermal drift of the
setpoint resistor can be taken into account. For example, con-
sider that the drift of the metal film resistor is 100 ppm/°C.
Since this drift is usually referred to +25°C, the setpoint resistor
can be in error by an additional 100 ppm/°C
×
(125°C – 25°C) or
1%. Using a setpoint temperature of 125°C as discussed above,
this error source would add an additional –0.2°C (for positive drift)
making the overall setpoint error potentially –0.52°C higher than
the original accuracy error.
Initial tolerance and thermal drift effects of the setpoint resistor
can be combined and calculated by using the following
equation:
R
MAX
=
R
NOM
×(1+ ε)×
(
1+
T
C
×(T
SET
– 25°C)
)
where:
R
MAX
is the worst case value that the setpoint resistor can be at
T
SET
,
R
NOM
is the standard resistor with a value closest to the desired
R
SET
,
ε
is the 25°C tolerance of the chosen resistor (usually 1%,
5%, or 10%),
T
C
is the temperature coefficient of the available resistor,
T
SET
is the desired setpoint temperature.
Once calculated, R
MAX
may be compared to the desired R
SET
from Equation 1. Continuing the example from above, the
required value of R
SET
at a T
SET
of 125°C is 5.566 kΩ. If the
nearest standard resistor value is 5.600 kΩ, then its worst case
maximum value at 125°C could be 5.713 kΩ. Again this is
+2.6% higher than R
SET
leading to a total additional error of
–0.52°C beyond that given by the specifications table.
THE HYSTERESIS AND SELF-HEATING
The setpoint resistor is determined by the equation:
R
SET
=
39
MΩ
°C
– 90.3
kΩ
T
SET
(°C)+ 281.6°C
Eq. 1
The setpoint resistor should be connected directly between the
R
SET
pin (Pin 6) and the GND pin (Pin 3). If a ground plane is
used, the resistor may be connected directly to this plane at the
closest available point.
The setpoint resistor, R
SET
, can be of nearly any resistor type,
but its initial tolerance and thermal drift will affect the accuracy
of the programmed switching temperature. For most applications,
a 1% metal-film resistor will provide the best tradeoff between
cost and accuracy. Calculations for computing an error budget
can be found in the section “Effect
of Resistor Tolerance and
Thermal Drift on Setpoint Accuracy.”
Once R
SET
has been calculated, it may be found that the calcu-
lated value does not agree with readily available standard
resistors of the chosen tolerance. In order to achieve an R
SET
value as close as possible to the calculated value, a compound
resistor can be constructed by connecting two resistors in series
or in parallel. To conserve cost, one moderately precise resistor
and one lower precision resistor can be combined. If the mod-
erately precise resistor provides most of the necessary resistance,
the lower precision resistor can provide a fine adjustment. Con-
sider an example where the closest standard 1% resistor has only
90% of the value required for R
SET
. If a 5% series resistor is
used for the remainder, then its tolerance only adds 5% of 10%
or 0.5% additional error to the combination. Likewise, the 1%
resistor only contributes 90% of 1% or 0.9% error to the combi-
nation. These two contributions are additive resulting in a total
compound resistor tolerance of 1.4%.
EFFECT OF RESISTOR TOLERANCE AND THERMAL
DRIFT ON SETPOINT ACCURACY
Figure 3 shows the typical accuracy error in setpoint temperature
as a function of the programmed setpoint temperature. This
curve assumes an ideal resistor for R
SET
. The graph of Figure 4
may be used to calculate
additional
setpoint error as a function
of resistor tolerance. Note that this curve shows additional error
beyond the initial accuracy error of the part and should be
The actual value of the hysteresis generally has a minor
dependence on the programmed setpoint temperature as shown
in Figure 6. Furthermore, the hysteresis can be affected by self-
heating if the device is driving a heavy load. For example, if the
device is driving a load of 5 mA at an output voltage (given by
Figure 9) of 250 mV, then the additional power dissipation
would be approximately 1.25 mW. With a
θ
JA
of 190°C/W in
free air the internal die temperature could be 0.24°C higher
than ambient leading to an increase of 0.24°C in hysteresis. In
the presence of a heat sink or turbulent environment, the
additional hysteresis will be less.
–6–
REV. 0
AD [ ANALOG DEVICES ]
