Please log in.
Before you can vote, you need to register. Please log in or create an account.
Vehicle: Car: Navigation: Non-GPS
Mechanical Car Compass   (+7, -2)  [vote for, against]
Keep Track of Wheel Rotations

This idea is relatively simple -- suppose that we add an digital encoder to each of an automobile's two rear wheels (or rather, to the half-axles, where they attach to the differential), and send the data from these two encoders to a CPU.

By counting how many times each wheel has turned, subtracting one from the other, and performing a modulo operation, it's possible to measure the vehicle's direction relative to it's starting orientation. That is, we have the rotational analog of an odometer.

This can rotational information be used to assist navigation, so that if the vehicle has a magnetic compass, but travels somewhere that has little magnetic field (inside a steel tunnel, for example), the "mechanical compass"'s direction can be used instead.

Because it accumulates errors (just like an odometer) it can't be used by itself, and must be regularly corrected... however, under many circumstances, it will be at least as accurate as a magnetic compass.
-- goldbb, Jun 15 2010

Directional Gyro http://en.wikipedia...ki/Directional_gyro
Does the same job - whether or not it's simpler is debatable. [DIYMatt, Jun 16 2010]

South Pointing Chariot http://en.wikipedia...th_Pointing_Chariot
Similar [goldbb, Jun 16 2010]

Wouldn't even the most minor difference in air pressure throw the thing off completely after just a short distance?
-- ldischler, Jun 15 2010


That would make it even more fun. Imagine the cross country competitions allowing only navigation by this device.
-- normzone, Jun 15 2010


I had assumed that that's how my car did things when out of range of the magic space clock radios - such as when in a tunnel. Though it might just look at the speed and assume things from there...
-- saedi, Jun 15 2010


If turns were banked, it'd underestimate how far the car had turned, no? (Banking ruins everything, doesn't it?)
-- mouseposture, Jun 16 2010


This is sort of related, aircraft have a directional gyro. The pilot sets the DG to the heading of the aircraft while on the ramp, and it will stay in its original position as long as the vacuum system is working. It is used in conjunction with the magnetic compass.
-- DIYMatt, Jun 16 2010


Better to use an electric or pneumatic gyro.
-- phoenix, Jun 16 2010


Traveller the SF role-playing game from the 70s, which was played with two dice, pen, paper and books, had inertial compasses. This is presumably what they were doing as your character drove around under-explored planets in their ATV.
-- Aristotle, Jun 16 2010


This might work until you spun your wheels, skidded to a stop, picked up your car with a trailering dolley...
-- RayfordSteele, Jun 16 2010


[RayfordSteele] Isn't it true that modern transmissions keep track of which wheels are skidding, so they can deliver power to the wheels that aren't? So, couldn't this device "believe" only the odometry from wheels currently enjoying good traction?
-- mouseposture, Jun 16 2010


ldischler, a minor difference in air pressure in the tires will throw it off, but I'm fairly sure that it will take many miles to cause significant error.

saedi, AFAIK, your navigation system (assuming it's built into the car, and not an add-on) will assume distance traveled based on the car's speed or odometer, but generally (I think) assumes that the vehicle stays pointed in the same direction it was when it last knew for sure.

DIYMatt, a directional gyro will suffer similar errors to an odometer-based direction-ometer. From the DG article on Wikipedia that you cited:

//As such any configuration of the aircraft horizontal which does not match the local earth horizontal results in gimbal error (essentially leading to a variation in the predictable 'apparent' wander known in this instance as drift.)// ... //Because the earth rotates (ω, 15° per hour), and because of small accumulated errors caused by friction and imperfect balancing of the gyro, the HI will drift over time, and must be reset from the compass periodically.//

Thus, banked turns will produce errors in a DG, just as they do for an odometric compass, though for different reasons. Furthermore, a DG will drift due to the earth's rotation, even if the vehicle it's in stays still, while an odometer-compass won't.

RayfordSteele, if your car has a GPS, but no magnetic compass, and you drive along the street, stop, and then use a trailering dolly to turn the vehicle around in-place, what direction will the nav system think that the car is facing?

In other words, no matter what kind of system you use, it will inevitably need to be calibrated on a regular basis.

Besides, the most common situations where the device will be activated will be parking garages and tunnels -- how often do you spin your wills, skid to a stop, move the car onto a trailering dolly, or drive on a banked turn whilst underground?
-- goldbb, Jun 16 2010


//if your car has a GPS, but no magnetic compass, and you drive along the street, stop, and then use a trailering dolly to turn the vehicle around in-place, what direction will the nav system think that the car is facing?// Depends. Am I allowed to have two GPS receivers at opposite ends of the car? And over what time interval am I allowed to average the difference in their estimated locations?
-- mouseposture, Jun 17 2010


All vehicle tyres operate at a slip rate all the time. Even when you're travelling dead straight at constant speed on the flat. The slip rate depends on a whole bunch of factors; vehicle geometry, surface macrotexture (bumpiness) and microtexture (roughness) etc. After just a few rotations, you will see significant errors. The errors suffered by a directional gyro will be insignifcant by comparison. The accuracy required to define to location of an aircraft at the end of a 3000 mile flight is entirely different to that requied to define whether your car is in your street or the next.

Learn to read a map and compass.
-- Twizz, Jun 17 2010


it would be great set into the steering wheel.
-- po, Jun 17 2010


When driving on dirt roads your wheels skid a little quite often.

The transmission doesn't keep track. The ABS sensors and pcm do that, with some input from the chassis control mod.

The simple crowning of a road is going to throw it off pretty fast.
-- RayfordSteele, Jun 17 2010


even on a perfect surface differential action isn't perfect. it wouldn't work.
-- WcW, Jun 17 2010


Deciding whether to performing jumps, handbrake-turns, and 4-wheel powerslides would have to be weighed up against the possibility of knocking the device out of callibration - mind you, if it did go off, you might decide to perform another manouvre to get it back in line.
-- zen_tom, Jun 17 2010


a) the chinese did it, as you pointed out yourself.

b) //we add an digital encoder // and // Mechanical Car Compass // don't really seat well together.

c) mechanical range calculation schemes on youtube (8 part series I think)
-- 4whom, Jun 17 2010


Accelerometers and a CPU would make for a more accurate Initial Navigation System. Probably cheaper, too.
-- whlanteigne, Jan 11 2013


//a minor difference in air pressure in the tires will throw it off, but I'm fairly sure that it will take many miles to cause significant error.//

Really? A little elementary maths perhaps?

Assume that one wheel has a circumference of 180cm, and the other has a circumference of 180.1cm (a tiny difference - 1 millimetre).

After driving ~3km (~1500 wheel revolutions), the smaller wheel will have rotated once more than the larger wheel. If the wheels are 2m apart, this will give an error of around 45 degrees.
-- MaxwellBuchanan, Jan 11 2013


Realistically, that difference would easily be measured, and compensated for, minute-by-minute when the regular compass or GPS was available.

Otherwise, I thinks its a lousy idea, though.
-- Kansan101, Jan 11 2013


Since the rotation is only in one plane, a basic fibre optic gyroscope that reorientates whenever an external r3ference is available would be the best solution.
-- 8th of 7, Jan 11 2013


//r3ference//

you're not lapsing into leet, are you?
-- MaxwellBuchanan, Jan 11 2013


Who has been bunning this?

S4V4NT
-- WcW, Jan 11 2013


Seem to be a good idea for a magnetic compass complement, but I can't see when is really needed. You can't loose driving in a tunnel; I can't see a situation where a magnetic compass it's not enough. Other examples could help to understand the need.
-- piluso, Jan 12 2013



random, halfbakery