
4-2
DSP56007/D 
MOTOROLA
Design Considerations
Thermal Design Considerations
The thermal performance of plastic packages is more dependent on the temperature 
of the PCB to which the package is mounted. Again, if the estimations obtained from 
R
θ
JA
 do not satisfactorily answer whether the thermal performance is adequate, a 
system level model may be appropriate. 
A complicating factor is the existence of three common ways for determining the 
junction-to-case thermal resistance in plastic packages: 
To minimize temperature variation across the surface, the thermal resistance 
is measured from the junction to the outside surface of the package (case) 
closest to the chip mounting area when that surface has a proper heat sink. 
To define a value approximately equal to a junction-to-board thermal 
resistance, the thermal resistance is measured from the junction to where the 
leads are attached to the case. 
If the temperature of the package case (T
T
) is determined by a thermocouple, 
the thermal resistance is computed using the value obtained by the equation 
(T
J
 – T
T
)/ P
D
. 
As noted above, the junction-to-case thermal resistances quoted in this data sheet are 
determined using the first definition. From a practical standpoint, that value is also 
suitable for determining the junction temperature from a case thermocouple reading 
in forced convection environments. In natural convection, using the junction-to-case 
thermal resistance to estimate junction temperature from a thermocouple reading on 
the case of the package will estimate a junction temperature slightly hotter than 
actual temperature. Hence, the new thermal metric, Thermal Characterization 
Parameter or 
Ψ
JT
, has been defined to be (T
J
 – T
T
)/ P
D
. This value gives a better 
estimate of the junction temperature in natural convection when using the surface 
temperature of the package. Remember that surface temperature readings of 
packages are subject to significant errors caused by inadequate attachment of the 
sensor to the surface and to errors caused by heat loss to the sensor. The 
recommended technique is to attach a 40-gauge thermocouple wire and bead to the 
top center of the package with thermally conductive epoxy.