-1
\$\begingroup\$

Does this chip exist, there only seem to be hall sensors?

\$\endgroup\$
4
  • 1
    \$\begingroup\$ Mounted on rotor without magnetic field? Please explain how this is? \$\endgroup\$
    – AltAir
    Commented Apr 29, 2017 at 19:31
  • \$\begingroup\$ @AltAir I suppose that the rotor might not have a magnetic field if it was, say, part of a pneumatic motor. \$\endgroup\$ Commented Apr 29, 2017 at 20:21
  • 1
    \$\begingroup\$ Its called a rotary encoder \$\endgroup\$
    – Voltage Spike
    Commented Apr 29, 2017 at 20:24
  • 2
    \$\begingroup\$ Potential XY problem. \$\endgroup\$
    – Andy aka
    Commented Apr 29, 2017 at 20:35

3 Answers 3

2
\$\begingroup\$

You need an electro-mechanical detector for your mechanical system.

A timing disc is often used, consisting of a disc mounted concentrically on the shaft with square holes punctured at regular intervals all around it.

enter image description here

A slotted opto-sensor can then be mounted so that the sensor's beam shines through the hole. As the motor spins, you get a square-wave and its frequency is proportional to the rotary speed of the sensor. The disc shown above also has a hole to indicate a 'home position'. This is at a different radius so that a second slotted opto-sensor can be fitted to detect a home position once per revolution.

A progression from this is the quadrature encoding system which uses two sensors to get speed and direction. The one shown below also has a home position sensor but that is separate to the quadrature encoding principle.

enter image description here

Quadrature encoding here uses two circles of holes, offset from each other radially and on a different radius. When the motor spins, the two opto-sensors have their beam broken and let through but at a different phases of the revolution. This results in a waveform from the sensors like this.

enter image description here

By examining the frequency and relationship of the two waveforms, both speed and direction can be obtained and to a greater precision because there are more edges per revolution in the two waveforms than in just one.

I have used quadrature encoding quite a bit. A simple scheme is to look for a rising edge on sensor A and take direction from the sensor B level at that instant. There are far better methods but it illustrates the idea.

\$\endgroup\$
1
  • \$\begingroup\$ You could also use a single circle of holes or notches, with the two sensors positioned so that they see the holes 90 degrees out-of-phase. \$\endgroup\$ Commented Apr 29, 2017 at 23:54
1
\$\begingroup\$

Certainly. It's called a photo interruptor.enter image description here

It contains an LED on one arm and a photodetector (usually a phototransistor) in the other. A protrusion of some sort on the edge of the rotating object passes between the two arms and breaks the beam once per rotation (of you can provide multiple protrusions for better angle resolution).

Then there are capacitive sensors which do the same thing by measuring capacitance to the object.

And there are optical proximity detectors which look at light being reflected from the object, with either a light or dark spot taking the place of the protrusion.

And there are other approaches as well. If nothing else, you can put a protrusion on a rotating body and use a microswitch to sense its passage. You did not, after all, specify that the sensor must not contact the object.

\$\endgroup\$
0
\$\begingroup\$

A capacitance probe can produce a signal that can be converted to rpm. It needs to work on some eccentricity such as gear teeth or a notch in an axle. It doesn't use a magnetic field and there are no moving parts other than what naturally spins in the motor. It's fairly good at almost standstill speeds too. Good for incidental frequencies up to and beyond 100 kHz.

It can also have a very small intrusive footprint and can be designed to work up to temperatures well in excess of 1000 degC.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.