Menlo Park, CA--Think you have heat dissipation problems? How's this for a tough equipment environment: electronics jam packed into an enclosure "can" atop a telephone pole somewhere in the tropics. And the can is sealed to keep out insects and contaminants.
That's the challenge engineers at one of Europe's largest telecom equipment providers faced in designing a radio base station for cellular-phone communications. The stations are the links phones connect with, initiating or receiving calls. Key design parameters are small size, for more devices in a given space, and low total cost--meaning low-maintenance operations. With a high level of integration and performance, heat dissipation from such a high-power density base station was a critical element in the design.
Sweat box. Here's the thermal and other considerations for the design:
100-200W power
-
Atmospheric temperatures ranging from -45 to 45C
-
Peak incident solar radiation of 1,100 W/m2
-
100% humidity
-
Resistance to heavy rain (as well as snow, ice) and industrial pollution
-
Tight EMI emissions and equipment security concerns
In order to keep the internal station temperature below approximately 70C under the worst combination of conditions, any heat dissipation mechanism had to provide a low thermal resistance of 0.0075C/W. Several methods were available to the designers to transfer heat from electronic components.
The engineers ruled out effective active solutions, such as fans, Peltier-effect (thermocouple-effect) cells, and refrigeration compressors, to keep power consumption under 200W. Remote site locations also made maintenance of active systems more costly. In addition, any fans could become blocked and deteriorate in open air, Peltier cells were expensive, and compressors occupied too much space.
Designers considered passive heat dissipation including: metal springs, dielectric liquids, phase-change materials (PCM), heat pipes, and interface materials. Metal springs posed electrical insulation and manufacturing problems. Dielectric liquids needed maintenance, including filling and emptying the case, and careful handling. PCMs required a double-walled case, increasing size, weight, and cost. And heat pipes were costly and mechanically complex. The designers thus felt a thermal interface material offered the best solution. But which one?
Testing time. The engineers conducted their own tests to determine thermal resistance and behavior under load for each interface material under consideration. They were forced to do this since manufacturers' data were not normalized to common conditions, and thus impossible to compare.
 |
| Low thermal resistance (shown by small test temperature differences) of HeatPath gel, combined with low compression stresses (shown here), permit proper heat dissipation even at minimal levels of surface pressure. Reduced clamping force lightens stresses on components. |
For the thermal test, they clamped the material between insulated aluminum hot and cold, thermocouple-instrumented plates, compressing it by 30% from its nominal thickness. Hot plate power and temperature were set at typical operational values. When thermal equilibrium was reached, the engineers measured the temperature difference ({DELTA}T) between plates. The smallest difference showed the best heat dissipation (see figure). Designers could then quantify thermal resistance ( degrees C/cm2) using power, {DELTA}T, and sample surface area.
An Instron Tester determined material characteristics (stress and displacement) under compression. Pressure-pad compression was limited to 1 mm/min to avoid spurious reactions from viscoelastic behavior of the elastomeric, conductive materials. Displacement and thermal-induced stresses had to be minimized since, with a large surface area covered by any thermally conductive material, small characteristic changes under pressure could overstress components, pins, and the pc boards on which they are mounted.
The effect of such stresses is increased by large changes in equipment operating temperatures. In addition, any thermal interface material had to be highly conformable and able to conduct heat even at low surface pressures--allowing for changes and differences in the size of interface gaps between components and heat-dissipation hardware.
The designers' verdict: Raychem's HeatPath thermal gel (see Design News 5/18/98, p.60). As one of the team notes, "The gel was chosen both for its low {DELTA}T during thermal conductivity comparison tests and its low compression set and ability to dissipate heat, even at low surface pressures. Gaps were easily and completely filled between components and heat sinks with minimal pressure, avoiding damage to delicate components and lowering stress on printed circuit boards. It could even accommodate large stack-ups of mechanical tolerances, tolerance mismatches, and non-parallel surfaces. Cost was also a factor in the decision."
Other applications
-
Hand-held electronics
-
Portable and desktop PCs
-
Power electronics