
7
Thermal resistances corresponding to other copper areas
can be obtained from Figure 20 or by calculation using
Equation 2. R
θ
JA
is defined as the natural log of the area
times a coefficient added to a constant. The area, in square
inches is the top copper area including the gate and source
pads.
ln
×
–
=
While Equation 2 describes the thermal resistance of a
single die, several devices are offered with two die in the
TSSOP-8 package. The dual die TSSOP-8 package
introduces an additional thermal component, thermal
coupling resistance,
R
θβ
. Equation 3 describes
R
θβ
as a
function of the top copper mounting pad area.
R
θβ
63.46
15.08
Area
ln
–
=
The thermal coupling resistance vs copper area is also
graphically depicted in Figure 20. It is important to note the
thermal resistance (R
θ
JA
) and thermal coupling resistance
(
R
θβ
) are equivalent for both die. For example at 0.1 square
inches of copper:
R
θ
JA1
= R
θ
JA2
= 173
o
C/W
R
θβ
1
=
R
θβ
2
= 98
o
C/W
T
J1
and T
J2
define the junction temperature of the
respective die. Similarly, P
1
and P
2
define the power
dissipated in each die. The steady state junction
temperature can be calculated using Equation 4 for die 1
and Equation 5 for die 2.
Example: To calculate the junction temperature of each die
when die 2 is dissipating 0.5W and die 1 is dissipating 0W.
The ambient temperature is 70
o
C and the package is
mounted to a top copper area of 0.1 square inches per die.
Use Equation 4 to calculate T
J1
and Equation 5 to calculate
T
J2
.
.
T
J1
= (0W)(173
o
C/W) + (0.5W)(98
o
C/W) + 70
o
C
T
J1
= 119
o
C
T
J2
= (0.5W)(173
o
C/W) + (0W)(98
o
C/W) + 70
o
C
T
J2
= 156.5
o
C
The transient thermal impedance (Z
θ
JA
) is also affected by
varied top copper board area. Figure 21 shows the effect of
copper pad area on single pulse transient thermal
impedance. Each trace represents a copper pad area in
square inches corresponding to the descending list in the
graph. Spice and SABER thermal models are provided for
each of the listed pad areas.
Copper pad area has no perceivable effect on transient
thermal impedance for pulse widths less than 100ms. For
pulse widths less than 100ms the transient thermal
impedance is determined by the die and package. Therefore,
CTHERM1 through CTHERM5 and RTHERM1 through
RTHERM5 remain constant for each of the thermal models. A
listing of the model component values is available in Table 1.
(EQ. 2)
R
θ
JA
138.68
14.95
Area
(
)
0
0.001
50
100
150
200
250
300
0.01
0.1
1
R
θ
,
θ
J
(
o
C
AREA, TOP COPPER AREA (in
2
) PER DIE
199
o
C/W - 0.017in
2
230
o
C/W - 0.0022in
2
IGURE 20. THERMAL RESISTANCE vs MOUNTING PAD AREA
R
θ
JA
= 138.68- 14.95 *
ln
(AREA)
R
θβ
= 63.46 - 15.08 *
ln
(AREA)
(EQ. 3)
(
)
×
(EQ. 4)
TJ1
P1R
θ
JA
P2
R
θβ
TA
+
+
=
(EQ. 5)
TJ2
P2R
θ
JA
P1
R
θβ
TA
+
+
=
FIGURE 21. THERMAL IMPEDANCE vs MOUNTING PAD AREA
50
100
150
200
0
10
-1
10
0
10
1
10
2
10
3
t, RECTANGULAR PULSE DURATION (s)
Z
θ
J
,
COPPER BOARD AREA - DESCENDING ORDER
0.02 in
2
0.14 in
2
0.26 in
2
0.38 in
2
0.50 in
2
I
o
C
ITF87008DQT