Everything for Electronics
Posted in: Developing Perspectives (August 2010)

Designing For The Best-Case Scenario

By Bryan Bergeron

Given the abundance of sunlight, I asked why solar panels and batteries weren’t used to power lights at night. I was lead to the roof of the trailer where the owner showed me three solar panels, each rated at 26V @ 4A. The US manufactured panels were apparently still functional, but collapsed for storage and without the matching batteries. The owner said the panels had destroyed two sets of 100 Ah deep-discharge batteries over the past few months, at considerable expense. The solar panels simply overcharged the batteries.

The designers of the solar power system apparently hadn’t considered the best-case scenario of days filled with blue skies and intense, uninterrupted, sunlight for weeks on end. They were probably more concerned with performance on overcast days and, at best, a day or two per week of uninterrupted sunshine. This oversight in design got me thinking about electronic design in general.

How often do you design for the best-case scenario? I have to admit that I normally design for the typical scenario with an eye to the worst case. For example, in RF links, I don’t assume the transmitter and receiver are in immediate proximity but may be at or just past maximum range. When it comes to switching power supplies, I’m constantly trying to squeeze the last bit of energy out of a battery before the supply shuts down.

Design for the best-case scenario is fundamentally different from over-engineering in which the goal is to produce a more robust circuit or system. Examples of over-engineering include using a transistor heatsink significantly larger than necessary to keep the junction temperature below maximum rated temperature, and building a power supply with diodes rated at four or five times the PIV expected. Over-engineering requires knowledge of what’s expected at the extremes of circuit operation, as well as deep pockets. The return is often increased reliability and enhanced performance.

Designing for the best-case scenario seems fundamentally different from and at odds with over-engineering. Take a wind generator. Designing for the worst-case scenario (as I see it) means designing the wind generator so that it requires virtually no wind. In contrast, a wind generator design that considers the best-case scenario would have provisions for operating in excessively high winds.

Do you intentionally over-design your circuits, or do you go for the cheapest solution that should work? Cost is an important, unavoidable consideration in today’s economy, especially if you’re building a circuit for educational purposes. There usually isn’t much at stake if a resistor or transistor in one of your projects overheats and has to be replaced. That’s how you develop an intuitive feel for what you can get away with in terms of power and voltage ratings.

The take-home message is to consider the best-case scenario when you’re designing your next circuit. You don’t want your circuit to fail because of too much of a good thing. NV