Unless otherwise noted, these specifications apply for V
S
= +5 V
DC
and I
LOAD
= +0.5 µA, in the circuit of
Figure 1.
Boldface
limits apply for the specified T
A
= T
J
= T
MIN
to T
MAX
; all other limits T
A
= T
J
= +25˚C, unless otherwise noted.
Parameter
Conditions
Typical
Accuracy
(Note 6)
Nonlinearity (Note 7)
Sensor Gain
(Average Slope)
Output Resistance
Line Regulation
(Note 8)
Quiescent Current
(Note 9)
Change of Quiescent
Current (Note 9)
Temperature Coefficient of
Quiescent Current
Long Term Stability (Note 10)
T
J
= 125˚C, for
1000 hours
Note 1:
Absolute Maximum Ratings indicate limits beyond which damage to the device may occur. DC and AC electrical specifications do not apply when operating
the device beyond its rated operating conditions.
Note 2:
See AN-450 “Surface Mounting Methods and Their Effect on Product Reliability” or the section titled “Surface Mount” found in a current National Semicon-
ductor Linear Data Book for other methods of soldering surface mount devices.
Note 3:
Human body model, 100 pF discharged through a 1.5 kΩ resistor. Machine model, 200 pF discharged directly into each pin.
Note 4:
Thermal resistance of the SOT-23 package is specified without a heat sink, junction to ambient.
Note 5:
Limits are guaranteed to National’s AOQL (Average Outgoing Quality Level).
Note 6:
Accuracy is defined as the error between the output voltage and 10mv/˚C times the device’s case temperature plus 500 mV, at specified conditions of volt-
age, current, and temperature (expressed in ˚C).
Note 7:
Nonlinearity is defined as the deviation of the output-voltage-versus-temperature curve from the best-fit straight line, over the device’s rated temperature
range.
Note 8:
Regulation is measured at constant junction temperature, using pulse testing with a low duty cycle. Changes in output due to heating effects can be com-
puted by multiplying the internal dissipation by the thermal resistance.
Note 9:
Quiescent current is defined in the circuit of
Figure 1
.
Note 10:
For best long-term stability, any precision circuit will give best results if the unit is aged at a warm temperature, and/or temperature cycled for at least 46
hours before long-term life test begins. This is especially true when a small (Surface-Mount) part is wave-soldered; allow time for stress relaxation to occur. The ma-
jority of the drift will occur in the first 1000 hours at elevated temperatures. The drift after 1000 hours will not continue at the first 1000 hour rate.
LM50B
Limit
(Note 5)
LM50C
Typical
Limit
(Note 5)
Units
(Limit)
T
A
= +25˚C
T
A
= T
MAX
T
A
= T
MIN
±
2.0
±
3.0
+3.0, −3.5
±
0.8
+9.7
+10.3
2000
+4.5V
≤
V
S
≤
+10V
+4.5V
≤
V
S
≤
+10V
+4.5V
≤
V
S
≤
+10V
+1.0
4000
2000
±
3.0
±
4.0
±
4.0
±
0.8
+9.7
+10.3
4000
˚C (max)
˚C (max)
˚C (max)
˚C (max)
mV/˚C (min)
mV/˚C (max)
Ω
(max)
mV/V (max)
mV/V (max)
µA (max)
µA (max)
µA (max)
µA/˚C
˚C
±
0.8
±
1.2
130
180
2.0
+2.0
±
0.8
±
1.2
130
180
2.0
±
0.08
±
0.08
www.national.com
2
Typical Performance Characteristics
circuit board as shown in
Figure 2.
Thermal Resistance
Junction to Air
To generate these curves the LM50 was mounted to a printed
Thermal Time Constant
Thermal Response in Still Air
with Heat Sink (
Figure 2
)
DS012030-22
DS012030-21
DS012030-23
Thermal Response
in Stirred Oil Bath
with Heat Sink
Start-Up Voltage
vs Temperature
Thermal Response in Still
Air without a Heat Sink
DS012030-25
DS012030-24
DS012030-26
Quiescent Current vs
Temperature (
Figure 1
)
Accuracy vs Temperature
Noise Voltage
DS012030-28
DS012030-27
DS012030-29
3
www.national.com
Typical Performance Characteristics
circuit board as shown in
Figure 2.
(Continued)
Supply Voltage
vs Supply Current
To generate these curves the LM50 was mounted to a printed
Start-Up Response
DS012030-31
DS012030-30
as Humiseal and epoxy paints or dips are often used to en-
sure that moisture cannot corrode the LM50 or its connec-
tions.
Temperature Rise of LM50 Due to Self-Heating
(Thermal Resistance,
θ
JA
)
SOT-23
no heat sink
*
Still air
Moving air
450˚C/W
SOT-23
small heat fin
**
260˚C/W
180˚C/W
*
Part soldered to 30 gauge wire.
**
Heat sink used is
1
⁄
2
" square printed circuit board with 2 oz. foil with part at-
tached as shown in
Figure 2.
DS012030-19
FIGURE 2. Printed Circuit Board Used
for Heat Sink to Generate All Curves.
1
⁄
2
" Square Printed Circuit Board
with 2 oz. Foil or Similar
2.0 Capacitive Loads
1.0 Mounting
The LM50 can be applied easily in the same way as other
integrated-circuit temperature sensors. It can be glued or ce-
mented to a surface and its temperature will be within about
0.2˚C of the surface temperature.
This presumes that the ambient air temperature is almost the
same as the surface temperature; if the air temperature were
much higher or lower than the surface temperature, the ac-
tual temperature of the LM50 die would be at an intermediate
temperature between the surface temperature and the air
temperature.
To ensure good thermal conductivity the backside of the
LM50 die is directly attached to the GND pin. The lands and
traces to the LM50 will, of course, be part of the printed cir-
cuit board, which is the object whose temperature is being
measured. These printed circuit board lands and traces will
not cause the LM50s temperature to deviate from the de-
sired temperature.
Alternatively, the LM50 can be mounted inside a sealed-end
metal tube, and can then be dipped into a bath or screwed
into a threaded hole in a tank. As with any IC, the LM50 and
accompanying wiring and circuits must be kept insulated and
dry, to avoid leakage and corrosion. This is especially true if
the circuit may operate at cold temperatures where conden-
sation can occur. Printed-circuit coatings and varnishes such
www.national.com
4
DS012030-7
FIGURE 3. LM50 No Decoupling Required
for Capacitive Load
DS012030-8
FIGURE 4. LM50C with Filter for Noisy Environment
The LM50 handles capacitive loading very well. Without any
special precautions, the LM50 can drive any capacitive load.
The LM50 has a nominal 2 kΩ output impedance (as can be
seen in the block diagram). The temperature coefficient of
the output resistors is around 1300 ppm/˚C. Taking into ac-
count this temperature coefficient and the initial tolerance of
the resistors the output impedance of the LM50 will not ex-
ceed 4 kΩ. In an extremely noisy environment it may be nec-
essary to add some filtering to minimize noise pickup. It is
