than the vertical trace (right) and will carry a higher thermal
power or electrical current for the same temperature rise. 1
Figure 4: Flaw in IPC-2221: the two traces have the same
cross-section, but they cannot have the same temperature with
the same current. For the same current, the left trace must be
cooler.
2.4. Validating the simulation with IPC-2152
2
For a revision of IPC-2221, the IPC task group 1-10b
performed new experiments on current-temperature
correlations. These new results [3,4] are also well reproduced
by our numerical models. Board parameters, materials and
environmental conditions are as described in Sect. 2.1. The
IPC-2152 design rule is not yet published in its final release.
3
Figure 5: Simulation results compared with some new 4
experimental data from [3].
3. New T vs. I correlations
3.1. FR4-based PCBs
The successful reproduction of the old IPC- and DN-Data
and the new experiments encourages us to calculate T(I)
diagrams for other PCB scenarios. The base material is FR4
(k=0.3 W/m-K), board thickness is D=1.6 mm, trace thickness
is t=35 μm (copper with solder resist). The copper layers
extend over the PCB completely. Ambient condition is ‘still
air’ (i.e. free convection) with Ta=20 degC. The thickness and
5
position of additional copper layers is indicated in the insert
of the diagrams and is motivated by typical application
requirements to PCB manufacturers [7]. The parameter of the
curves is the trace width w (in mm). The copper planes in
scenarios 4-7 are symmetric with respect to the mid-plane.
The temperature of the trace on the vertical axis is the
calculated average temperature in the volume of the trace.
There are little deviations from uniformity, which shall not be
discussed here.
Adam, New Correlations Between Electrical Current and … 20th IEEE SEMI-THERM Symposium
Page: Previous 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Next
Last modified: September 9, 2013