The Case of the Overly Sensitive Sensor

THE MYSTERY

One of ACES’ clients does metal fabrication and manufacturing for the government, which has exacting standards. In this case the client needed to report and display the rotational speed of a spin lathe shaft. In some applications it might be possible to calculate the shaft speed from the motor speed, but this was a belt-driven system — and belts slip, introducing inaccuracies.

The Control Systems Investigator (CSI) on the job devised a 12” diameter ring studded with a circle of 45, ¼” bolts, to be read by a magnetic proximity sensor. Once everything was installed and the computer programmed, the unit seemed to be working fine…and then it wasn’t.

THE CLUES

Many times the sensor was reading two signals for each bolt head. The CSI reasoned that perhaps the holes in the Allen bolts were causing the double signals. He changed out the Allen bolts for regular bolts, with no improvement. Then the CSI had a hunch that the rounded profile of the bolts was the issue, so he went to a solid, flat head bolt. No luck.

THE PERP

After some research, the CSI discovered that — even though the manufacturer’s manual for the proximity sensor claimed that the component could read bolt heads — further investigation clarified that the sensor has to see a rectangular shape coming off the top — not rounded or hexagonal. Any shape other than a sharp rectangle could result in what’s known as “double tapping”.

THE SOLUTION

The CSI sketched out a custom bolt head using 5/8” square stock, which was then drawn out and custom built. The techs tapped a ¼” hole on one side and filled it with all thread.

The customer installed the new bolts, which came complete with a lock washer to place between the adaptor and the ring so they could tweak each bolt until it was angled precisely to the center of the shaft. In the end they had 45 smooth, ¼” thick, 5/8” wide rectangles of steel for the sensor to scan — providing a sharp on and sharp off.

After reprogramming, the proximity sensor worked perfectly and transmitted the speed of the spin lathe independent of the motor.

Another solution could have been to design a ring with 45 gear-like teeth and have it laser cut. This approach would have eliminated the final step of angling each individual bolt.

CASE CLOSED