TruTherm should only be enabled when measuring the tem-
perature of a transistor integrated as shown in the processor
of Figure 3, because Equation 5 only applies to this topology.
3.1.2 Calculating Total System Accuracy
The voltage seen by the LM95241 also includes the IFRS volt-
age drop of the series resistance. The non-ideality factor, η,
is the only other parameter not accounted for and depends
on the diode that is used for measurement. Since ΔVBE is
proportional to both η and T, the variations in η cannot be
distinguished from variations in temperature. Since the non-
ideality factor is not controlled by the temperature sensor, it
will directly add to the inaccuracy of the sensor. For the Pen-
tium D processor on 65nm process, Intel specifies a +4.06%/
−0.89% variation in η from part to part when the processor
diode is measured by a circuit that assumes diode equation,
Equation 4, as true. As an example, assume a temperature
sensor has an accuracy specification of ±1.25°C at a temper-
ature of 65 °C (338 Kelvin) and the processor diode has a
non-ideality variation of +4.06%/−0.89%. The resulting sys-
tem accuracy of the processor temperature being sensed will
be:
TACC = + 1.25°C + (+4.06% of 338 K) = +14.97 °C
and
TACC = - 1.25°C + (−0.89% of 338 K) = −4.26 °C
TrueTherm technology uses the transistor equation, Equation
5, resulting in a non-ideality spread that truly reflects the pro-
cess variation which is very small. The transistor equation
non-ideality spread is ±0.4% for the Pentium D processor on
65nm process. The resulting accuracy when using TruTherm
technology improves to:
TACC = ±1.25°C + (±0.4% of 338 K) = ± 2.60 °C
The next error term to be discussed is that due to the series
resistance of the thermal diode and printed circuit board
traces. The thermal diode series resistance is specified on
most processor data sheets. For the Pentium D processor on
65 nm process, this is specified at 4.52Ω typical. The
LM95241 accommodates the typical series resistance of the
Pentium D processor on 65nm process. The error that is not
accounted for is the spread of the Pentium's series resistance,
that is 2.79Ω to 6.24Ω or ±1.73Ω. The equation to calculate
the temperature error due to series resistance (TER) for the
LM95241 is simply:
(6)
Solving Equation 6 for RPCB equal to ±1.73Ω results in the
additional error due to the spread in the series resistance of
±1.07°C. The spread in error cannot be canceled out, as it
would require measuring each individual thermal diode de-
vice. This is quite difficult and impractical in a large volume
production environment.
Equation 6 can also be used to calculate the additional error
caused by series resistance on the printed circuit board. Since
the variation of the PCB series resistance is minimal, the bulk
of the error term is always positive and can simply be can-
celled out by subtracting it from the output readings of the
LM95241.
Processor Family Transistor Equation
nD, non-ideality
Series
R
min typ max
Intel processor on 65nm
process
0.997 1.001 1.005 4.52 Ω
Processor Family Diode Equation ηD,
non-ideality
Series
R
min typ max
Pentium III CPUID 67h 1 1.0065 1.0125
Pentium III CPUID 68h/
PGA370Socket/
Celeron
1.0057 1.008 1.0125
Pentium 4, 423 pin 0.9933 1.0045 1.0368
Pentium 4, 478 pin 0.9933 1.0045 1.0368
Pentium 4 on 0.13
micron process,
2-3.06GHz
1.0011 1.0021 1.0030 3.64 Ω
Pentium 4 on 90 nm
process
1.0083 1.011 1.023 3.33 Ω
Pentium on 65 nm
porcess
1.000 1.009 1.050 4.52 Ω
Pentium M Processor
(Centrino)
1.0015
1
1.0022
0
1.0028
9
3.06 Ω
MMBT3904 1.003
AMD Athlon MP model 6 1.002 1.008 1.016
AMD Athlon 64 1.008 1.008 1.096
AMD Opteron 1.008 1.008 1.096
AMD Sempron 1.0026
1
0.93 Ω
3.1.3 Compensating for Different Non-Ideality
In order to compensate for the errors introduced by non-ide-
ality, the temperature sensor is calibrated for a particular
processor. National Semiconductor temperature sensors are
always calibrated to the typical non-ideality and series resis-
tance of a given processor type. The LM95241 is calibrated
for two non-ideality factors and series resistance values thus
supporting the MMBT3904 transistor and the Intel processor
on 65nm or 90nm process without the requirement for addi-
tional trims. For most accurate measurements TruTherm
mode should be turned on when measuring the Intel proces-
sor on the 65nm or 90nm process to minimize the error
introduced by the false non-ideality spread (see 3.1.1 Diode
Non-Ideality Factor Effect on Accuracy). When a temperature
sensor calibrated for a particular processor type is used with
a different processor type, additional errors are introduced.
Temperature errors associated with non-ideality of different
processor types may be reduced in a specific temperature
range of concern through use of software calibration. Typical
Non-ideality specification differences cause a gain variation
of the transfer function, therefore the center of the tempera-
ture range of interest should be the target temperature for
calibration purposes. The following equation can be used to
calculate the temperature correction factor (TCF) required to
compensate for a target non-ideality differing from that sup-
ported by the LM95241.
TCF = [(ηS−ηProcessor) ÷ ηS] × (TCR+ 273 K) (7)
where
•ηS = LM95241 non-ideality for accuracy specification
•ηT = target thermal diode typical non-ideality
•TCR = center of the temperature range of interest in °C
The correction factor of Equation 7 should be directly added
to the temperature reading produced by the LM95233. For
example when using the LM95241, with the 3904 mode se-
www.national.com 18
LM95241