NE1617A All information provided in this document is subject to legal disclaimers. © NXP B.V. 2012. All rights reserved.
Product data sheet Rev. 5 — 20 March 2012 13 of 30
NXP Semiconductors
NE1617A
Temperature monitor for microprocessor systems
9. Application design-in information
9.1 Factors affecting accuracy
9.1.1 Remote sensing diode
The NE1617A is designed to work with substrate transistors built into processors’ CPUs
or with discrete transistors. Substrate transistors are generally PNP types with the
collector connected to the substrate. Discrete types can be either a PNP or an NPN
transistor connected as a diode (base shorted to collector). If an NPN transistor is used,
the collector and base are connected to D+ and the emitter to D. If a PNP transistor is
used, the collector and base are connected to D and the emitter to D+. Substrate
transistors are found in a number of CPUs. To reduce the error due to variations in these
substrate and discrete transistors, a number of factors should be taken into consideration:
The ideality factor, n
f
, of the transistor. The ideality factor is a measure of the deviation
of the thermal diode from the ideal behavior. The NE1617A is trimmed for an n
f
value
of 1.008. Equation 2
can be used to calculate the error introduced at a temperature
T C when using a transistor whose n
f
does not equal 1.008. Consult the processor
data sheet for n
f
values.
This value can be written to the offset register and is automatically added to or
subtracted from the temperature measurement.
(2)
Some CPU manufacturers specify the high and low current levels of the substrate
transistors. The I
source
high current level of the NE1617A is 100 A and the low level
current is 10 A.
If a discrete transistor is being used with the NE1617A, the best accuracy is obtained by
choosing devices according to the following criteria:
Base-emitter voltage greater than 0.25 V at 6 mA, at the highest operating
temperature.
Base-emitter voltage less than 0.95 V at 100 mA, at the lowest operating temperature.
Base resistance less than 100 .
Small variation in h
FE
(say 50 to 150) that indicates tight control of V
BE
characteristics.
Transistors such as 2N3904, 2N3906, or equivalents in SOT23 packages are suitable
devices to use. See Table 11
for representative devices.
NE1617A All information provided in this document is subject to legal disclaimers. © NXP B.V. 2012. All rights reserved.
Product data sheet Rev. 5 — 20 March 2012 14 of 30
NXP Semiconductors
NE1617A
Temperature monitor for microprocessor systems
9.1.2 Thermal inertia and self-heating
Accuracy depends on the temperature of the remote-sensing diode and/or the internal
temperature sensor being at the same temperature as that being measured, and a
number of factors can affect this. Ideally, the sensor should be in good thermal contact
with the part of the system being measured, for example, the processor. If it is not, the
thermal inertia caused by the mass of the sensor causes a lag in the response of the
sensor to a temperature change. In the case of the remote sensor, this should not be a
problem, since it is either a substrate transistor in the processor or a small package
device, such as the SOT23, placed in close proximity to it.
The on-chip sensor, however, is often remote from the processor and is only monitoring
the general ambient temperature around the package. The thermal time constant of the
SSOP16 package in still air is about 140 seconds, and if the ambient air temperature
quickly changed by 100 C, it would take about 12 minutes (five time constants) for the
junction temperature of the NE1617A to settle within 1 C of this. In practice, the
NE1617A package is in electrical and therefore thermal contact with a printed-circuit
board and can also be in a forced airflow. How accurately the temperature of the board
and/or the forced airflow reflect the temperature to be measured also affects the accuracy.
Self-heating due to the power dissipated in the NE1617A or the remote sensor causes the
chip temperature of the device or remote sensor to rise above ambient. However, the
current forced through the remote sensor is so small that self-heating is negligible. In the
case of the NE1617A, the worst-case condition occurs when the device is converting at
16 conversions per second while sinking the maximum current of 1 mA at the ALERT
output. In this case, the total power dissipation in the device is about 11 mW. The thermal
resistance, R
th(j-a)
, of the SSOP16 package is about 121 C/W.
In practice, the package has electrical and therefore thermal connection to the printed
circuit board, so the temperature rise due to self-heating is negligible.
Table 11. Representative diodes for temperature sensing
Manufacturer Model number
Rohm UMT3904
Diodes Inc. MMBT3904-7
Philips MMBT3904
ST Micro MMBT3904
ON Semiconductor MMBT3904LT1
Chenmko MMBT3904
Infineon Technologies SMBT3904E6327
Fairchild Semiconductor MMBT3904FSCT
National Semiconductor MMBT3904N623
NE1617A All information provided in this document is subject to legal disclaimers. © NXP B.V. 2012. All rights reserved.
Product data sheet Rev. 5 — 20 March 2012 15 of 30
NXP Semiconductors
NE1617A
Temperature monitor for microprocessor systems
9.1.3 Layout considerations
Digital boards can be electrically noisy environments, and the NE1617A is measuring very
small voltages from the remote sensor, so care must be taken to minimize noise induced
at the sensor inputs. The following precautions should be taken.
1. Place the NE1617A as close as possible to the remote sensing diode. Provided that
the worst noise sources, that is, clock generators, data/address buses, and CRTs, are
avoided, this distance can be four to eight inches.
2. Route the D+ and D tracks close together, in parallel, with grounded guard tracks on
each side. Provide a ground plane under the tracks if possible.
3. Use wide tracks to minimize inductance and reduce noise pickup. 10 mil track
minimum width and spacing is recommended (see Figure 4
).
4. Try to minimize the number of copper/solder joints, which can cause thermocouple
effects. Where copper/solder joints are used, make sure that they are in both the D+
and D path and at the same temperature.
Thermocouple effects should not be a major problem since 1 C corresponds to about
200 V and thermocouple voltages are about 3 V/C of temperature difference.
Unless there are two thermocouples with a big temperature differential between them,
thermocouple voltages should be much less than 200 V.
5. Place a 0.1 F bypass capacitor close to the V
DD
pin. In very noisy environments,
place a 1000 pF input filter capacitor across D+ and D close to the NE1617A.
6. If the distance to the remote sensor is more than eight inches, the use of twisted pair
cable is recommended. This works up to about six feet to 12 feet.
7. For really long distances (up to 100 feet), use shielded twisted pair, such as
Belden #8451 microphone cable. Connect the twisted pair to D+ and D and the
shield to GND close to the NE1617A. Leave the remote end of the shield
unconnected to avoid ground loops.
Because the measurement technique uses switched current sources, excessive cable
and/or filter capacitance can affect the measurement. When using long cables, the filter
capacitor can be reduced or removed.
Cable resistance can also introduce errors. 1 resistance introduces about 1 C error.
Fig 4. Typical arrangement of signal tracks
002aag953
GND
GND
D+
D−
10 mil
10 mil
10 mil
10 mil
10 mil
10 mil
10 mil

NE1617ADS,112

Mfr. #:
Manufacturer:
NXP Semiconductors
Description:
Board Mount Temperature Sensors I2C LOC +/- 2OC AND
Lifecycle:
New from this manufacturer.
Delivery:
DHL FedEx Ups TNT EMS
Payment:
T/T Paypal Visa MoneyGram Western Union

Products related to this Datasheet