
MOTOROLA
Appendix G. 66-MHz Electrical Characteristics
G-5
Thermal Characteristics
Historically, the thermal resistance has frequently been expressed as the sum of a junction
to case thermal resistance and a case to ambient thermal resistance:
RθJA = RθJC + RθCA
where:
RθJA = junction to ambient thermal resistance (°C/W)
RθJC = junction to case thermal resistance (°C/W)
RθJA = case to ambient thermal resistance (°C/W)
RθJC is device related and cannot be influenced. The user controls the thermal environment
to change the case to ambient thermal resistance, RθCA. For instance, the air flow can be
changed around the device, add a heat sink, change the mounting arrangement on printed
circuit board, or change the thermal dissipation on the printed circuit board surrounding the
device. This description is most useful for ceramic packages with heat sinks where about
90% of the heat flow is through the case to the heat sink to ambient. For most packages, a
better model is required.
The simplest thermal model of a package which has demonstrated reasonable accuracy
(about 20 percent) is a two resistor model consisting of a junction to board and a junction
to case thermal resistance. The junction to case covers the situation where a heat sink will
be used or where a substantial amount of heat is dissipated from the top of the package. The
junction to board thermal resistance describes the thermal performance when most of the
heat is conducted to the printed circuit board. It has been observed that the thermal
performance of most plastic packages and especially PBGA packages is strongly dependent
on the board. temperature.
If the board temperature is known, an estimate of the junction temperature in the
environment can be made using the following equation:
TJ = TB + (RθJB x PD)
where:
TB = board temperature (°C)
RθJB = package junction to board resistance (°C/W)
PD = power dissipation in package ()
If the board temperature is known and the heat loss from the package case to the air can be
ignored, acceptable predictions of junction temperature can be made. For this method to
work, the board and board mounting must be similar to the test board used to determine the
junction to board thermal resistance, namely a 2s2p (board with a power and a ground
plane) and vias attaching the thermal balls to the ground plane.
When the board temperature is not known, a thermal simulation of the application is
needed. The simple two-resistor model can be used with the thermal simulation of the