On my route the other day I used my "cateye" computer for the first time. (not sure which version, but one of the lower end models). It was installed on the bike by the LBS who sold me the bike. On the ride, I also wore my Garmin Forerunner 310xt on my wrist. This GPS watch is more for running / and triathlons. I've been using the Garmin Forerunner watches for running for the last several years.

I only wore the Garmin on this bike ride to test the accuracy of the Cateye.

I noticed after mile Number one that there was only about a .03 mile discrepancy. I thought, "not too bad.....That's fairly close....."

But after I multiply that by the short 11 mile ride, it adds up..... (11 x .03 = .33)

Here's what they each read -

Cateye -
11.35 miles

Garmin (using GPS) -
11.01 miles

If I were to get up to a 50 mile ride, I'd be looking at a mile and half discrepency.

I was telling a friend of mine this and he said he's heard the cateyes are not "exact".

My questions are -
1. Is it probably an issue with the cateye needing to be adjusted more accurately?
2. is this discrepency just a norm for bike computers and I should pay no attention to it? (this was the view of the friend I spoke to).

You can select your wheel diameter in the Cateye setup. I am not sure if the cateye being ahead of the GPS means your wheel is programmed at a too big, or too small, setting. I am guessing too big.

Read your manual and maybe dial down the wheel size on the Cateye's computer just a hair, and see if that makes it closer.

I'd bet the Cateye is more accurate.

considering the cateye uses a 4-bit (2^4 =16) microcontroller I'm surprised it's even that accurate.
So in theory over a mile you can only resolve 330ft increments (5280/16) , which is +/- 6.25% error. This also assumes there is no errors associated with the rims and tires.

The Garmin claims an accuracy of +/- 1%

I found the same issue during a triathlon this past weekend. Top + average speed and distance were higher on the Cateye (and slightly higher than the course distance).
If I had to bet, I'd go with the Garmin as far as accuracy. If the Cateye is just slightly off, each mile will add more to the discrepancy. I don't ever recall GPS telling me I'm 1.5 miles from where I really am. (not sure if it's a fair comaprison...but you know what I mean)

Could possibly also depend on the placement of the sensor magnet on your wheel...

My Cateye and Garmin are within a tenth over ~20 miles. Check the wheel size chart for your Cateye and see if the shop set it right.

Huh? It counts revolutions... Placement has nothing to do with it.

Incorrect. The GPS is very accurate, particularly over some distance. (i.e. immediate speed reading can fluctuate with reception, but over a number of miles its very accurate.)

The Cateye is only as accurate, as the programmed wheel circumphrence. If you really want to dial that in, you need to do a rollout, with the bike loaded with your weight, and the tires at the same pressure you are riding.

Absent such a rollout, just using the numbers off a chart will not be nearly as accurate as GPS.

If you put the magnet closer to the hub, it goes around much faster.

https://sheldonbrown.com/cyclecomputer-magnet.html

I assumed since it was keeping track of "time", too.... That it would matter...?


This is what i was thinking. Exactly. But I may be wrong..

????

I've compared my Cateye distances to Google Maps distances up to around 25 miles and they are almost identical (within a tenth of a mile). I have it set to 700x23c = 210.

haha, same number of times though

OK.... You learned me.

ok...it doesn't matter where the magnet is on the spoke. The whole spoke sweeps through the plane with the same angular velocity (though they have differing linear velocities); each spoke goes around the same number of times per minute. The number of times the spoke goes around is the same as the number of times the wheel goes around. From the wheel size you enter, the distance traveled is calculated.

I know you are joking, but you still have it backwards.

The farther away from the center, the faster the magnet will be going. But it's going a greater distance around a larger circumference.

and yet, I can see this as correct at the same time.

On a grand scale, consider this.

Let's say it takes the same amount of time for Mercury to make one full revolution around the sun as it does Neptune. (I know this is NOT true, but for the example...)

(Mercury is closest to the sun, Neptune is farthest).

They both take X amount time to get around one full revolution.....

Which planet was moving faster? Take a 1/4 of the distance of each planet, and measure that speed. You'd think if Neptune had to travel SOOO much more distance in the same amount of time, wouldnt' it be traveling faster?


But at the same time, they end at the same points, at the same time, yet Neptune travel milliions of miles farther.....

true.

If there was a system where it had 2 magnets next to eachother and it measured the gap between them, then it would matter where you put the magnet on the spoke.

Actually if you put the magnet so close to the hub that the sensor picks it up twice a revolution it will make you really fast.

This can be a problem when using a Powertap with a magnet, as just a cyclecomputer. ( It has to be close to the sharktooth to pickup, but too close and it picks up twice.)

Makes for an impressive file however.

Incorrect error propagation by the way.

Wiki has a decent article but it may be more than you care for: https://en.wikipedia.org/wiki/Propagation_of_uncertainty

This is a completely irrelevant analogy. The distance traveled on a bike is entirely dependent on the distance traveled by a given point which is the distance of the contact point from the center of the wheel.

The magnet is simply a tool to measure revolutions, and then the distance (and speed) are calculated as a function of these revolutions and the radius entered into the computer.

In other words, it does not matter what point you use to measure the revolutions, because we don't care how far that point travels at all. We only care about how quickly that point makes revolutions, because that is the only input that matters for the computer. It should be obvious that the distance of the point from the center has no impact on number of revolutions.

Edit: You'll notice, however, that the radius is a key variable in the calculations that the computer makes. Having fuller or flatter tires WILL have an impact on distance recorded. Accurately measuring from the center to the contact point (while loaded) will yield the most accurate computer measurements.

I think this one finally sunk in with me....

Glad to be of help ^^

Also, I just reread my post and realized it might come across a little aholeish. It wasn't intended, sorry about that!