RoadBikeReview.com's Forum Archives - Components


Archive Home >> Components(1 2 3 4 5 6 7 8 9 10 )


Stupid Magnet Question(9 posts)

Stupid Magnet QuestionPete
Sep 27, 2001 10:16 AM
K, really stupid question that I'm sure that somebody will know the answer and have a good laugh at the same time. Is there a typical radial placement for wheel magnets? I had to buy a magnet for the comp. on my bike, put it on, but think that I may have it set too far out. Is there a standard distance from the hub that a magnet should be set? I would apreciate any help.

Thanks in advance,

Pete
Need some more info...Cima Coppi
Sep 27, 2001 11:19 AM
Pete,

What brand of computer do you have? Different computers have different placements on the wheel. Also, have you checked out the website for your computers manufacturer?

CC
Need some more info...Pete
Sep 27, 2001 11:31 AM
It's a Specialized, and no I haven't checked thier website yet, but I'll do that right now.
Magnet placement irrelevantBipedZed
Sep 27, 2001 1:27 PM
Bike computers calculate velocity and distance based upon the time it takes for the magnet to make one revolution to the sensor. It then calculates velocity and distance based upon the inputed wheel circumference or diameter. Without getting into geometry/trigonometry, it doesn't matter where you place the magnet on the radius (spoke) of the wheel, the time it takes to make one revolution will be the same regardless of whether it's near the hub or near the rim.
Actually it is relevant...Ken2
Sep 27, 2001 3:21 PM
There is another factor that seems to affect the computer sensor unit (may vary by manufacturer), and that's the actual speed that the magnet moves past it. I've noticed it more at slow mtn. bike speeds, but basically as the magnet moves past slower and slower, there gets to be a point where the readout zeroes out, usually around 3 mph.
It follows that putting the magnet farther out towards the rim will lessen this effect; if it's really close in on the hub be prepared for no reading below a certain speed, and therefore no mileage or ave. speed accruing.
You are correct sir.Atombomber
Sep 27, 2001 8:47 PM
The further the magnet is from the center of rotaion, the faster is passes the contact point. This allows for less of a sweep time in the contact zone. Greater accuracy.
Design Dependantgrzy
Sep 28, 2001 11:28 AM
Being able to pickup the pulse is more a factor of the design of the sensor - good engineering would dictate that it should work in the entire envelope. Avocet claims that there design is better since they use a magnetic ring at the hub with several reversals in magnetism which gives many more pulses per single revolution. This allows the unit to more quickly and accurately detect changes. No valid reason for why a sensor shouldn't work at virtutally any reasonable speed at any location.

I question your asseertation that accuracy is improved. Given that one pulse is going to occur at the same moment the magnet passes the sensor there doesn't seem to be anything that would systematically make this vary. Ultimately the accuracy is tied to getting the wheel size correct. Does slowly sweeping through the sensor zone produce a lead or a lag given that we're operating an LC circuit? Now, this would affect accuracy since the lead/lag would varry with speed. Also, there is typically an accuracy spec with a unit that seems to be a smaller error than intriduced through measuring the circumfrence of the wheel under riding conditions.

Accuracy - the ability to get the right reading
Percision - the ability to get repeatable results, even if they're not quite right
Design Dependence of magnets, pickups, cyclometersbykboi
Oct 1, 2001 4:15 PM
Don't you suspect that the closure of the switch is the event the cyclometers' circuits use to start a calculation? I do. Given that, I also suspect that although Chen's idea carries some weight (pun there), the real deal here is what Bipedzed talked about. That was the likely solution! It prob'ly doesn't matter how far the magnet and pickup are from the hub as long as the cyclometer has more than 1 pulse to compare time for from one revolution to the next in a given time period, right?

We also don't know how long the cyclometer circuit is designed to wait for the next pulse do we? If it's one second, for example, then a 1.01 sec wait would give a zero reading and perhaps log zero distance. If it's 3 or 4 seconds, the rider would have that much time to input the next pulse (by rotating the magnet past the pickup again-by rotating the wheel being measured) thus causing a comparison from the last input pulse for the known wheel circumference in the inputted time passage between pulses, making a calculation possible by the cyclometer's circuits.

Since we don't know this stuff and most cyclometer mfrs don't tell us (I had an old Paramount cyclometer that updated based on everything it got in 4 seconds per the specs), we don't really know the cutoff point for Pete's Specialized cyclometer to read zero speed or zero distance.

All this was kind of impressive but possible bad science. In the end, Bipedzed was likely nearest the solution mark. To paraphrase Bipedzed, "...if the wheel carries the magnet past the pickup (sensor), the thing oughta work no matter where it's radial location."

By the way grzy...Avocet doesn't make all ring magnet models any more. They're mostly magnet and pickups like the ones we've discussed here. I think only 2 are ring magnets. I've used one successfully for 6-7 years now and it'd be longer, but one of the older models I had couldn't be repaired when it gave out after 5 or 6 years (too out of date for repalcement parts), so for $25 they sold me a newer model that still fit the mounts on all 4 of my bikes. A testimonial to good service for good products.

Now let's consider a REAL engineering faux pas...The Shimano chain pin that only works once. They can make 105 almost look, act and last like D-A but they can't figure out how to make a replaceable chain pin. Hmm-m-m. Good marketing, bad engineering, no? 2 or 3 dollars to change a chain pin? "It is the weakest link...good-bye."

Good riding!

bykboi
re: Stupid Magnet QuestionChen2
Sep 27, 2001 3:07 PM
As far as the computer is concerned it does not matter how far out you place the magnet. But if you really care, it will make a difference to the balance of the wheel. Usually the wheel's heavy side is either the side with the valve stem, or the side directly opposite the valve stem where the rim's seam is. You can test for this by spinning the wheel slowly and seeing which side stops at the bottom. Place the magnet opposite the heavy side. You can also use the valve stem lock ring and plastic cap to add or reduce weight on the valve side of the wheel.

-Al