How do speedometers work? Can it work if attached to a revolving, moving part?

Hi guys,

I have a question that's been on my mind and I was wondering if anyone knew the answer.

I was wondering if there is any speedometer capable of detecting speed of movement (or even just movement), if attached to that moving part?

It seems that it a way that a GPS speedometer can sort of do this, but is it based on GPS, thus it requires linear movement in a certain direction?

What about something like a moving gear? Would there be a way for any current speedomer tech to measure the speed of the movement of the gear?

Basically, can you measure the basic speed of the moving gear itself? (Not RPM or anything, but MPH/KPH).

I know this seems like a crazy question, but I do have a purpose behind it.

Thanks!

If we're talking about the "speed" of a moving gear, I assume we are talking about the velocity of a point on the outside edge of the gear as it spins in place?

If you are looking for the distance travelled by the outside edge of a gear, you need a tachometer and math.  I'm sure you could calibrate a mechanical speedometer to do that math for you, since it is driven by a spinning cable.  An electronic one could read impulses from magnets placed around the circumference of the pulley.  Again, some math to determine pulses per mile would be required.

What you're asking about is inertial-based speed measurement.  Basically you have a set of accelerometers inside a box, they measure the acceleration of the box in 3 dimensions, subtract out the force of gravity, and then integrate the remaining values over time to compute the speed (and then position) of the device.  A track datalogger contains this kind of tech and combines it with the GPS signal to produce a more accurate location of the car on track.

This would not be a practical way to measure the speed of a rotating gear, however.  These devices are not cheap, require power, and errors in them stack up rapidly the longer they go between being synced to a known location.  Rotational speed sensors like streetwiseguy is talking about are cheap, easy, and readily available (that's what a crank angle sensor or an ABS wheel speed sensor is).

Also look at encoders, they measure rotation of shafts, or if you don't need quite as much precision, a proximity switch and something protruding from the shaft to trigger it every rotation (or fractions of a rotation). I gather this is how some aftermarket speedos are done, with something clamped to the driveshaft to trigger a prox. In industrial settings, these are often paired.

The way the tach and speedo in my 1976 honda bike is designed is as follows (and maybe this is standard for mechanical gauges? I've only taken one set of mechanical gauges apart in my life...):

- The needle is subjected to a counter-clockwise momentum by a spiral spring. So it takes a linear increasing momentum to turn the needle clockwise.

- The base of the needle is attached to a metal disc, not connected to anything else then the needle.

- The cable is turning a disc shaped magnet, placed close to the metal disc, clockwise. The faster the cable turns the magnet, the greater the clockwise momentum on the disc. In other words, the gauge is measuring the rpm of the spinning magnet (which could be your gear?). 

I think you're making this a little over complicated, folks.

As long as it is calibrated correctly for the spinny part of the speedo sensor, it won't matter what size anything is.

Let's say you want to measure the speed of how fast the outside of a bike tire is spinning.  Put the wheel of the speedo on the outside of the spinning tire.  It says 20 mph.  Now switch to a smaller tire and spin it to 20 mph.  The smaller tire is moving at a faster RPM, but the outside of the tire is still moving at 20 mph.

Traditional speedos use a proper calibration of gears based on rear gearing, tire size, etc to get the right signal.  If the reluctor/spinny part is in direct contact with the moving surface itself, it will work on anything; flat linear motion, small tire, big tire, anything... because you are always taking your sample at the exact moment of inertia or in the same velocity vector as the actual motion.

Think about it... if you are driving at 65 mph, the surface of your tires are rotating at 65 mph relative to the hub.  It doesn't matter if you have 37" swampers or 22" donuts.

The usual way to measure the speed of a gear would be to count teeth using a gear tooth sensor (no joke - some types of Hall effect and other proximity sensors are actually sold using that designation) and determine the gear's RPM by dividing the number of teeth counted by the time of the counting period.