About 35 years ago I worked for a company that designed and manufactured altimeters for light aircraft. These altimeters had to be hand-calibrated by connecting them to a vacuum source which simulated altitude. There were several key altitude points that the technicians would bring the altimeters to in order to verify the reading against a calibrated standard called a quartz manometer. To bring them back down to sea level, or even below sea level, pressure was applied by dry nitrogen which was supplied by nitrogen cylinders that were delivered from an industrial gas vendor.
A quartz manometer is a device that contains a chamber to which the unit under test was connected via a hose and finger-knob valves that controlled vacuum or pressure. The inside of the chamber contained a quartz tube wound in a spiral from which hung a mirror. A light beam was shone on the mirror and the beam reflected on a series of photosensors that determined if the chamber needed to be rotated to make sure the light beam remained on the center.
As the chamber rotated, it drove mechanical dials indicating the amount of movement, and thus the simulated altitude. The inside of the quartz tube contained a reference vacuum, so that as pressure varied outside, in the chamber, the whole spiral tube and mirror deflected an amount proportional to the pressure difference.
One day, all the manometers started going crazy. With no altimeter connected, they kept counting up as if they were rapidly going up in altitude. The cause became clear when I noticed one of the technicians joking around by inhaling the gas from the pressure line (supposed to be dry nitrogen), and talking funny. As it turns out, the industrial gas vendor had accidentally delivered a helium cylinder for one of the nitrogen cylinders. When it was hooked up and it got into the manometer chamber, it actually diffused through the quartz tube (yes, it can do that; helium is weird stuff) and destroyed the reference vacuum.
It caused several days of lost time and repair as all the manometers had to have the quartz reference vacuum pumped out and re-calibrated and all the pressure lines flushed of helium.
Tell us your experience in solving a knotty engineering problem. Send stories to Chris Wiltz for Sherlock Ohms, and read more cases here.
Russ Poffenberge is a software architect at Xcerra Corp.