Counting Bike Wheel Revs

How about installing a contact pin in a spoke and a corresponding fixed insulated contact in the structural tubing nearby, and counting the contact hits (with bounce suppressed).

There are two such sensors in each mechanical mouse. Can't be more than $1 each. You just need to make the transfer gears and mounting stuffs. You can always buy up all the used mouses for the sensors.

They are? Hall effect sensors are well under $1 each in the DigiKey catalog--probably less than $0.60 in large quantities. If you pick another solution, you could use up that much in signal conditioning and PC boards pretty quickly.

OTOH, you could fasten a playing card so that it hits the spokes. Add a microphone, amplifier, and DSP chip, and you should be able to determine wheel speed pretty accurately! ;-) The sound effects should be a good marketing point!

Mark Borgerson

My Sigma Sport 600 bike speedometer, odometer and trip calculator wa quite innexpensive as a complete system (approx UKP12.00) and uses a magnet on the spokes and a reed switch in the fork mounted sensor. A magnet and a coil would probably do as well.

Well, you could take a look at how it's done in other absolute cheapest mass-production products.

Doubt it. You'd need some amplifier/filter/interface circuitry to sense the current induced by the magnet -- probably about the same price as a hall effect sensor.

Go the a local bike shop, buy a few computers (cheap ones aren't much more than 25USD), and take apart the sensors.

I think a magnet on a spoke and a reed switch attached to the frame is the standard solution.

That would fail at low speeds, and so probably fail the user 'walk the bike to check' test...

You don't say what other resource you have ? Viz: A smarter processor, can often use a dumber sensor.

REEDs have been suggested, and they have a key advantage of being nominally sealed/waterproof, and have a high KISS rating.

Asian Detect switches are cheaper, but fail the splash proof test.

Do you need to know actual revolutions, or is speed of more interest ? Would a one-per-spoke signal, give better display results, than one-per-rev ?

For one per spoke, your choices are optical beam [needs dual cables, one per fork] and magentic proximity sensing [needs ferrous metal spokes]

-jg

Attach a wider reflective or even non-reflective item to one spoke and have a opto sensor attached to one of the forks that detects a longer pulse (either darker or lighter pulse) than the spokes.

Could even be an infra red led on other side and covers to obscure as much sunlight a possible. Thus also be useable at night.

Could be placed nearer axle to reduce dirt/mud issues.

Yeah more like one per rev is fine.... I think it looks like magnet and reed switch testing right now....

VERY KISS!

Thanks guys! Looks like it will be magnet and reed switch testing right now as this whole thing I want to make needs to come in around $1.50 to be feasible!

Let ya know when its done!

Rick.

Either a magnetic reed switch or count the spokes with an opto-interrupter. I suspect that the cost of the amp for spoke counting would offset the cost gains over a cheap Hall sensor though.

Bob

No, but I've heard them.

Perhaps some of the newer ones are using hall-effect sensors, but I'm pretty sure that a few (e.g. 10) years back they were pretty much all reed switches. If you held the sensor close to your ear and moved the magnet past it, you could year a faint clicking.

In my experience, hall-effect sensors don't produce much signal and are rather fussy to interface to. They require a analog amplifier and either a comparity or A/D converter. The output signals aren't very consistent from one unit to the next.

A reed-switch OTOH, is dead simple to interface to. All you need is an input port pin (or IRQ pin) and a pullup resistor.

They have these things now called "Integrated Circuits"!! ;-) The $1 sensors in the DigiKey catalog include a signal conditioning chip that converts the magnetic signal to a digital output. Three pins: Power, Ground, digital output.

True. But does it have the same sensitivity, a built-in schmidt trigger and other features (about which I know little)?

Reed switches are about $0.39 qty 1000 in the Jameco catalog, but you still need to mount it, make sure it is oriented properly, and protect it from breakage. A 3-Pin SIP Hall sensor might be worth the extra $0.60.

I've used reed switches, but never on the fork of a bicycle, where shock and vibration might be an issue. Are those issues in favor of a Hall sensor?

Mark Borgerson

Mark Borgerson

Darn. What will they think of next! Someday I bet they put a whole computer in one package...

I would think that either would work fine.

I don't remember bicycle computers ever having any problems with them.

Did anyone suggest a conductive metal flag projecting from the side of the wheel (or both sides uf balance is critical). It could close a contact (soft and/or flexible) and make like a momentary switch. If there is a micro involved it could even detect missed hits based on timing.

----------------------------------------------------------------------- (Please pardon any typos. I am typibg one handed for a while due to recent hand surgery)

I've used a magnet on the wheel spokes and a reed switch on the fork, both taped with packing tape. It seemed to be reliable at speeds up to 46 mph.

-----

A reed switch doesn "Schmidtt triggering" naturally, though there may be a slight problem with contact bounce, that a Hall effect sensor won't have.

Don't you also need to mount a Hall effect sensor, make sure it is oriented properly, and protect it from breakage? Okay, it's not glass like a reed switch, but it still needs to be 'adjusted' in relation to the magnet on the wheel, just as does the reed switch.

Reed switches also have the advantage of being totally passive and not needing power. An integrated-circuit Hall sensor probably takes (just a swag, I haven't looked at a data sheet lately) 100 or 1,000 times as much power as the average electronic bicycle speedometer/computer.

-----

Or a missing pulse could be an indication of wheel lockup. It could pulse the brakes when that is detected, thus making a bicycle with an Antilock Brake System. To do this seriously would require about ten equally spaced magnets on the wheel.

-----