Meeting safety and RoHS is relatively straightforward. For now, we’ll look at thermal and then EMC.
Effective Cooling of Enclosed Electronics
Achieving proper or adequate thermal dissipation is affected by many variables such as:
- Total watts dissipation
- Watt density per square inch
- Type, volume and material of enclosure
- Type dissipation, conduction, convection, forced air
- Solar Loading (ODUs or Outdoor units)
While we developed and produced conducted and convection enclosures, the majority are forced air (containing fans or blowers). To approach proper cooling the pure scientist would say “We need to do an FEA” (Finite Element Analysis). While FEA works well in some instances, it fails to yield accurate outcomes for forced air electronic enclosures.
Flow FEA uses predominantly fluid principles. If, for instance, we analyze how plastic might fill a mold, FEA works relatively well. Plastic flows into a mold like toothpaste. It has a know mass and viscosity, etc. Air, on the other hand, has no mass. Well, technically it does but its mass in forced air cooling FEA analysis doesn’t provide much value. Force Airflow is finicky, squirrely and doesn’t behave in obvious ways. It’s speed and direction are affected by too many things:
- Pressure (negative and positive)
- Having to change directions
- Square corners
- Chamber volume
- Inlet and outlet size and location
Let’s take a moment and look at just one issue. You put a heat sink on a BGA for example. The data sheet graphs watt dissipation effectiveness given a characteristic of air parading through the pins/fins of the heat sink. That’s all well and good, but . . . Airflow doesn’t like heat sinks. Moving air sees them as obstructions. Heat sinks cause pressure and directional changes resulting in added turbulence. Air prefers the path of least resistance – to go around or over heat sinks, not through them. Not very effective for cooling electronics. Over the years we have developed numerous strategies to positively effect airflow for improved cooling.
EMC – Dealing With Emissions and Immunity
This begins with two primary questions:
- What is your clock speed? Leading to what’s the ¼ wavelength?
- What is the enclosure material and process?
These two basic questions will tell most of the story. More specific questions about the board design will lead to what design direction needs to be taken. It is more cost effective to manage radiation at the board level then to contain it on a global (enclosure) level. For example, can traces be buried between ground planes or add a small localized EMI containment can soldered to the board? When that won’t meet EMC requirements we have to resort to enclosure containment. This sets in play a different set of design rules. We begin with enclosure part geometry, but this may not do all that is needed. Then, having to add grounding materials to the enclosure (beryllium clips, conductive elastomers or fabric over foam, etc.) quickly adds cost. In the case of plastic enclosures, resin selection, and coatings are often the direction. Again, at additional cost.
Depending upon the enclosure materials and processes, we’ve developed inexpensive techniques to circumvent many of these cost adders. We’ll present what can be done early in the design process of your next custom electronic enclosure.