Thermal Resistance Model for ACPL-K34T
The diagram for measurement is shown in Figure 23. Here, one die is heated first and the temperatures of all the dice are
recorded after thermal equilibrium is reached. Then, the second die is heated and all the dice temperatures are recorded.
With the known ambient temperature, the die junction temperature and power dissipation, the thermal resistance can
be calculated. The thermal resistance calculation can be cast in matrix form. This yields a 2 by 2 matrix for our case of
two heat sources.
1
8
2 Die1: Die 2:
7
LED Detector
3
6
4
5
Figure 23. Diagram of ACPL-K34T for measurement
R11 R12
P1
∆T1
R21 R22 • P2 = ∆T2
R11: Thermal Resistance of Die1 due to heating of Die1 (°C/W)
R12: Thermal Resistance of Die1 due to heating of Die2 (°C/W)
R21: Thermal Resistance of Die2 due to heating of Die1 (°C/W)
R22: Thermal Resistance of Die2 due to heating of Die2 (°C/W)
P1: Power dissipation of Die1 (W)
P2: Power dissipation of Die2 (W)
T1: Junction temperature of Die1 due to heat from all dice (°C)
T2: Junction temperature of Die2 due to heat from all dice (°C)
TA: Ambient temperature (˚C)
∆T1: Temperature difference between Die1 junction and ambient (˚C)
∆T2: Temperature deference between Die2 junction and ambient (°C)
T1 = (R11 × P1 + R12 × P2) + TA ------------------(1)
T2 = (R21 × P1 + R22 × P2) + TA ------------------(2)
Measurement is done on both low and high conductivity boards as shown below:
Layout
76mm
Measurement data
Low conductivity board:
R11=191 ˚C/W
R12=R21= 68.5˚C/W
R22=77˚C/W
High conductivity board:
R11=155 ˚C/W
R12=R21= 64˚C/W
R22=41˚C/W
Note that the above thermal resistance R11, R12, R21 and R22 can be improved by increasing the ground plane/copper
area.
14