There Is No Such Thing As An Invalid Unit

The Mars Climate Orbiter was a spacecraft launched in the closing years of the 1990s, whose job was to have been to study the Martian atmosphere and serve as a communications relay point for a series of other surface missions. It is famous not for its mission achieving these goals, but for the manner of its premature destruction as its orbital insertion brought it too close to the planet’s atmosphere and destroyed it.

The ill-fated Mars Climate Orbiter craft. NASA [Public domain].
The ill-fated Mars Climate Orbiter craft. NASA [Public domain].
The cause of the spacecraft entering the atmosphere rather than orbiting the planet was found in a subsequent investigation to be a very simple one. Simplifying matters to an extent, a private contractor supplied a subsystem which delivered a reading whose units were in the imperial system, to another subsystem expecting units in the SI, or metric system. The resulting huge discrepancy caused the craft to steer towards the surface of the planet rather than the intended orbit, and caused the mission to come to a premature end. Billions of dollars lost, substantially red faces among the engineers responsible.

This unit cock-up gave metric-using engineers the world over a brief chance to feel smug, as well as if they were being honest a chance to reflect on their good fortune at it not having happened on their watch. We will all at some time or another have made an error with respect to our unit calculations, even though in most cases it’s more likely to have involved a simple loss of a factor of ten, and not with respect to a billion dollar piece of space hardware.

But it also touches on one of those fundamental divides in the world between the metric and imperial systems. It’s a divide that brings together threads of age politics, geography, nationalism, and personal choice, and though it may be somewhere angels fear to tread (we’ve seen it get quite heated before to the tune of 885+ comments), it provides a fascinating subject for anyone with an interest in engineering culture.

Continue reading “There Is No Such Thing As An Invalid Unit”

Standard Resistor Teardowns

What do you do, when you want an ohm? What is an ohm, for that matter? Take a wander over to the textbook definitions, and you’re soon deep in a world of coulombs and parallel infinite planes one meter apart in a vacuum that you probably only half remember from your high school physics class. It’s hard work, this metrology lark.

Of course, you can just order a resistor. A few cents each when you’re buying small quantities or much less when you’re buying a reel of five thousand, and there you have it. An ohm. Only it’s not really an ohm, more like nearly an ohm. Within 1% of an ohm is pretty good, but Vishay or Bourns or whoever don’t have the margins to get philosophical about those infinite planes when you’re only giving them a few cents.

When you REALLY want an ohm, you buy a standard resistor, and you pay a more significant sum. You’re never going to wire one of these up to bias a transistor or drive an LED, instead it’s about as close as it’s possible to get on your bench to the value it says on the box and you can use it for calibration purposes. PPM figures well in excess of the resolution of even superior DMMs sound pretty good to us!

[Zlymex] was curious about standard resistors, so performed a teardown of a few to see what they contain. And after a pithy explanation of the terms involved he’s managed to look inside quite a few of them.

Inside he finds hermetically sealed wire-wound resistors, some oil-filled wire-wound resistors, and the occasional hefty piece of manganin. He also tears down some of the hermetically sealed resistors themselves, finding both wire-wound and foil resistance elements within.

It is a curious obsession that permeates hacker culture, that of standard measurements, and it’s one we’ve covered quite a few times here. Time enthusiasts with atomic clocks like this rather beautiful discrete logic build, or voltage enthusiasts with their temperature compensated references or programmable standards. Surprisingly though, this appears to be the first time we’ve looked at standard resistors.

Thanks [David Gustafik] for the tip.