|   Weird question for weird people | Fast Eddy
Nov 27, 2001 2:58 PM | | I'm looking for a graphical representation of what's inside a bike speedo sensor. |
| Well, I'm not that weird. But here's something else. | KriZkit
Nov 27, 2001 4:02 PM | | I've forgotten your friends name. I've got the rest of the pictures finally. They turned out okay but didn't scan too well. |
| What a nut. | Fast Eddy
Nov 27, 2001 4:32 PM | | He's Mikes buddy. I don't remember either. I forget how old I am too. |
| You either want a picture of the internals of a computer | Twilight Error®
Nov 27, 2001 4:04 PM | | a chart of some manner showing the steps a computer takes to calculate speed,distance,time, etc...? |
| Nope | Fast Eddy
Nov 27, 2001 4:37 PM | | I made my own computer, and I had to write all that software. Now I'm writing a report, and I need to explain the sensor itself.
As I said below, it's a physical contact of some kind. |
| re: Weird question for weird people | Kafin8ed
Nov 27, 2001 4:14 PM | | the part on the wheel is usually a magnet, the part on the fork/frame is a small coil of wire. as the magnet passes in close proximity to the coil a current is induced in the coil that travels to the computer. the magnitude or (depending on the type of computer) frequency of the current is used to calculate all the computer data. that's the generalized non-graphical representation of a speedometer sensor. a graphical representation would not be anymore exciting as coils and magnets are rarely very thrilling to look at... *must be read as an impersonation of ben stein* |
| The magnet... | JRinCA
Nov 27, 2001 4:30 PM | | attached to the fork only tells the computer each time the wheel magnet passes. The computer does all the calculating based on the entered wheel diameter. |
| weird magnets... | Kafin8ed
Nov 27, 2001 4:40 PM | | some old-school computers (avocets i think)had magnets that were actually donuts that went all the way around the wheel (down by the hub), therefore the induced current must have been continuous & speed was calculated based on amplitude of the current...using the entered wheel diameter. that's why i stated that it could work either way, depending on what type of computer you had... |
| Actually, I know that's wrong | Fast Eddy
Nov 27, 2001 4:36 PM | | If what you said was true, you'd only need one wire. There are always two.
The sensor has a mechanical contact of some kind. The passing magnet makes the two wires connected for a millisecond or so. I've got one working, I just want to know what kind of contacts are in that little thingy. |
| it might be different but it's not wrong | Kafin8ed
Nov 27, 2001 4:47 PM | | well it is one wire sort of, it goes down the fork, around in a coil, and back up the fork. if it was only one simple wire it would not make a complete circuit, it has to be 2. there are probably 20 different ways to make this work, i outlined the most common way to do it (in my experiences) certainly some company could have come up with a "better" mousetrap, however i gather that you're trying to fix something so the term "better" is sure to be used loosely above =) |
| Sorry, I didn't mean you were wrong | Fast Eddy
Nov 27, 2001 5:15 PM | | I just meant that's not how the one I'm using works. Actually I'm really glad too, because it would require a more complicated circuit to detect the speed.
Right now, I send 5v down, and it comes back, only for an instant, once per rev. I measure the time in between, do some fancy math-e-matics, et voila. |
| Gonna have to agree with Kaf ... | K'Endo
Nov 27, 2001 5:32 PM | | It's a coil. Anything with contacts - internal moving parts - would wear out and fast. A coil has no moving parts and yes, it requires two wires to complete the circuit that would sense the voltage induced by the passing magnet.
Also, your contact based system would likely stop operating when you move fast enough - not enough time for the magnet to pull the contact all the way closed. With a coil, you'll always get some measurable rise and fall of the voltage induced by the magnet.
Hey, designing control systems has been my life for 15 years. I couldn't resist!
Kn. |
| If it was a coil... | Fast Eddy
Nov 27, 2001 5:55 PM | | ...I don't think the system as I've designed it would work. I have two wires coming from the sensor:
1 -----------
(sensor)
2 -----------
My circuit looks like this:
+5 - 1 -----------
(sensor)
int - 2 -----------
where int is the interrupt line going into a PIC microprocessor. I think that if the sensor was a coil, I'd get a high signal on the interrupt line all the time.
Also, if I connect a circuit like this:
+5 - 1 -----------
(sensor)
gnd - 1kohms - led - 2 -----------
I can get the led to stay on by placing the magnet and sensor in close proximity. No motion required |
| I may have some of the connections wrong, but ... | K'Endo
Nov 27, 2001 6:30 PM | | How bout something like this? My transistor theory was never strong, but I think this is the general idea.
When I mention 'moving parts', I mean internal to the contact unit. To open and close on a magnetic impulse, some little chunk of metal must be moving around inside. It'll fail eventually. A coil has no movement, no wear.
Kn. |
| That's the "more complicated" circuit I was talking about. nm | Fast Eddy
Nov 27, 2001 7:27 PM | | |
| No no ... the correct response is ... | K'Endo
Nov 27, 2001 7:45 PM | | Dat's jest crazy enuff to work!!
Well, g'luck anyhoo!
Kn. |
| My sensor makes a clicking sound when the magnet passes. | Zignzag
Nov 27, 2001 6:17 PM | | Avocet 25. That would support your mechanical contact theory. |
| Never noticed before, but mine does too. | Fast Eddy
Nov 27, 2001 11:29 PM | | Just a little click when the magnet is near. Cateye AT-100.
So now I'm worried that my sensor, thus whole circuit is going to be
unreliable. I might remember it not working so well when I was using it. |
| Maybe someone's got one that doesn't work??? | Fast Eddy
Nov 27, 2001 4:38 PM | | Give it a good bashing with a hammer, and take a picture. |
| If I do, can I ask, Why? | KriZkit
Nov 27, 2001 4:51 PM | | I have a broken one. I'll gladly break it open for you if I can find it. A guy at work has a digital camera. Tomorrow though. Someone may beat me to it. |
| I built a computer for a school project... | Fast Eddy
Nov 27, 2001 5:18 PM | | ...now I have to write a report. I want to describe the sensor.
Maybe you better not. I'd hate to have you spend a bunch of time and find out it's the coil kind like kafin8ed was describing. |
| not sure about this but... | kafin8ed
Nov 27, 2001 10:21 PM | | it's certainly possible that they have included other elements in the circuit. perhaps a capacitor so that a signal is only sent when the sensor passes by and induces a "large" current spike, this could also filter out background noise...resistors could also be used for filtration purposes. these electronic trinkets certainly could create multimeter mischeif when checking the leads??? just a thought i'm not mr wizard though... |
| Another homework idea, no bashing required. | Fast Eddy
Nov 27, 2001 5:59 PM | | If you have a multimeter, take off your speedo, and measure the resistance between the contacts.
If there's a coil down there, there will be a little resistance.
If there are contacts, it will be the same as putting one probe in your nose and waving the other one around like a magic wand.
The second part of the experiment is entirely optional. |
| Queer answers for queer people | FagDetective
Nov 27, 2001 7:03 PM | | Speedo's? Little resistance? Magic wands? Entirely optional experiments?
Take this stuff to boy-oh-boy.com! |
| Waving the free probe around certainly IS nessecary! (nm) | Twilight Error®
Nov 27, 2001 7:04 PM | | |
| Here ya go.... | CraigH
Nov 27, 2001 4:43 PM | | Used the words "magnet coil" in google.com image finder. |
| YAAAAY! (nm) | Kafin8ed
Nov 27, 2001 4:53 PM | | ... |
| I've opened up 2 or 3 of them..... | alibi
Nov 28, 2001 9:53 AM | | ...and all of them were just a momentary contact leaf-spring kind of thing. A very thin band of ferrous metal moves into contact with a second tab of metal when drawn by the passing magnet, thus momentarily closing a circuit. It's just an on/off switch; the computer counts the closures per unit time and calcalutes eveything from that. |
| This is correct... | Rick O'Shay
Nov 28, 2001 10:15 AM | | The idea was borrowed from the alarm industry. Magnetic switches are available as normally closed or normally open and are usually used on doors and windows. When the door/window is opened, the magnet moves away from the switch and changes the contact state. It's a pretty low-tech approach and because it is mechanical, will fail after "x" number of contacts. An all-electronic design would be better but is probably more expensive. |
| |
