Datasheet

17
Thermal Model
(Discussion applies to ACPL-3130, ACPL-J313 and ACNW3130)
Figure 32. Thermal Model.
The steady state thermal model for the ACPL-3130 is
shown in Figure 32. The thermal resistance values given
in this model can be used to calculate the temperatures
at each node for a given operating condition. As shown
by the model, all heat generated ows through q
CA
which
raises the case temperature T
C
accordingly. The value of
q
CA
depends on the conditions of the board design and is,
therefore, determined by the designer. The value of q
CA
=
83°C/W was obtained from thermal measurements using a
2.5 x 2.5 inch PC board, with small traces (no ground plane),
a single ACPL-3130 soldered into the center of the board
and still air. The absolute maximum power dissipation
derating specications assume a q
CA
value of 83°C/W.
From the thermal mode in Figure 32 the LED and detector
IC junction temperatures can be expressed as:
T
JE
= LED junction temperature
T
JD
= detector IC junction temperature
T
C
= case temperature measured at the center of the
package bottom
q
LC
= LED-to-case thermal resistance
q
LD
= LED-to-detector thermal resistance
q
DC
= detector-to-case thermal resistance
q
CA
= case-to-ambient thermal resistance
*q
CA
will depend on the board design and the placement
of the part.
Inserting the values for q
LC
and q
DC
shown in Figure 32
gives:
For example, given P
E
= 45 mW, P
O
= 250 mW, T
A
= 70°C
and q
CA
= 83°C/W:
T
JE
and T
JD
should be limited to 125°C based on the
board layout and part placement (q
CA
) specic to the
application
�
LD
= 442 °C/W
T
JE
T
JD
θ
LC
= 467 °C/W θ
DC
= 126 °C/W
θ
CA
= 83 °C/W*
T
C
T
A
θ
( )
( )
( )( )
ACA
LDDCLC
DCLC
DCADCLDLCEJE
Tθ
θθθ
θθ
Pθθθ||θPT ++
++
+++=
( )
(
ACALCLDDCDCA
LDDCLC
DCLC
EJD
Tθθθ||θPθ
θθθ
θ
θ
PT +++++
++
=
)
•
•
•
•
•
•
( ) ( )
ACADCAEJE
TθC/W57PθC/W256PT ++°++°=
( ) ( )
ACADCAEJD
TθC/W111PθC/W57PT ++°++°=
•
•
•
•
C120C70C/W140250C/W33945mW
TC/W140PC/W339PT
ADEJE
°=°+°+°=
+°+°=
mW
C70C/W194250C/W14045
TC/W194PC/W140PT
ADEJD
°+°+°=
+°+°=
mWmW
• •
•
•
•
•
•
•