Overstated Computer Speed and Distance

[keno]

I hate when I start to think about this kind of thing, but it's happened again. It's like an acid flashback.

While I know that the 2096 (for a 700x23 tire) bike computer manuals I'm familiar with give as the magic circumference in mms is right in some sense, it is wrong in practice. I've never bothered to do anything about it in setting up any computer, even if possible. I know that I'm overstating speed and distance. Maybe not enough for most to care about for themselves, but I'm not without my problems.

Clearly, those manuals overstate the rolliing circumference by reason of the fact that the tire on the wheel from which the measurment is determined is depressed a bit by some portion of the weight of the bicycle and myself, and in that way decreasing the effective radius, therefore circumference, of the tire. In my own case, that's about 205 total on two 700x23 tires at 110 psi.

Has anyone felt strongly enough about this to try to determine the truth for yourself? I have to believe that any of you who worry about gms would be inclined to futz with this, too. My sense for myself is about a 2%+ overstatement.

While ordinarily not of much interest, if you are training for a time trial or other endurance race against the clock, it becomes an important thing to know in order to determine how competitive you will be. If 25 mph is your goal and that's what your bike computer is telling you, and what you've actually been doing is about 24.5mph, not good. There will be some 'splainin to do.

(All of those years in engineering school have left an unfortunate scar.)

keno

keno

[thejen12]

While ordinarily not of much interest, if you are training for a time trial or other endurance race against the clock, it becomes an important thing to know in order to determine how competitive you will be.
keno
I think the clock will tell you, if you're training for a race against the clock.

However, every time I set up a new bike/computer combo, I ride a known course to make sure the distance comes out right. In my case, it's around the block for the initial setting, double-checked by a commute to work.

Jenn

[92degrees]

My last 'puter, a VDO, had a procedure where you rolled the wheel for a complete revolution and measured the distance and plugged that in during set-up.

[Ozz]

I posted a similar question a couple years ago cuz I was too lazy to get help rolling my tire to get the circumference....I think the answer came back as: "sit on your bike, (accurately) measure center of hub to the ground...the the pi math thing for finding circumference, and use that number."

[Dave]

A value of 2096 should not be too large. My book from Vetta suggests 2105.
The best way to be sure is to find a flat length of smooth sidewalk or road, long enough for at least three tire revolutions. Carefully mark a start point where the tire meets the road and roll through three revolution, preferably with your weight on the saddle. Repeat the measurement at least three times and take the average. If your measurements are not close, then you're doing something wrong.

In theory you could also sit on the bike and have someone carefully measure from the floor to the center of the axle to get a tire radius, then multiply by 6.28, but it's real easy to have errors. The bike has to be sitting plumb and the measuring tape has to be plumb, or you'll get a larger number than is correct.

[cadence231]

My last 'puter, a VDO, had a procedure where you rolled the wheel for a complete revolution and measured the distance and plugged that in during set-up.


I do that too.
On my SRM, I have the speed/distance sensor mounted for the rear wheel.

1)What I do is inflate that tire(rear) to the psi I usually use.
2) Place a mark on the tire surface that will make contact with the riding surface with some paint or White Out etc.
3)Then I get on the saddle with everything I take with me for a ride(weight).
4)Roll forward in a "straight" line one revolution of the tire. You will know one revolution when you see two marks from the above mentioned paint.
5)Measure that distance between the two marks and enter that into your computer.

[MarleyMon]

I hate when I start to think about this kind of thing, but it's happened again. It's like an acid flashback...
keno
You need to find some better acid, dewd.
Have someone measure the radius while you are on the bike.
Then calculate the value of Pi for the duration of your next century.
Multiply by the wavelength in nanometers of the color "Harlequin".
Divide by the specific gravity of EnSure and you're set! :)

[Blue Jays]

Or multiply results by .95 to lowball the training results and then subsequently pleasantly surprise yourself on raceday! :)

[keno]

Sorry, I know how to do all of those things, but as I said I haven't bothered to do any of them.

What I was asking, poorly it seems, was what number did you get when you did one of those things.

keno

[maunahaole]

Won't the final number depend on what tire you use? A michelin 700x23 is a different size than a Conti 700x23.

[regularguy412]

... the shortest distance between two points is still the straightest line. How much 'side to side' you do during a ride will affect the distance covered compared to the 'actual' distance. Yup. I know. It's stating the obvious. On our group rides, almost everyone comes up with a slightly different number at the end of the ride. This is probably due to computer calibration, but also can be because we didn't all take the same 'line'.

I like to low ball my computer. I use several different sizes of tire/wheel combo from time to time and I just don't feel like recalibrating at every change. If you're within a half mile per hour of the other riders' speeds, I wouldn't worry about it.

Mike in AR

[vaxn8r]

Get a Garmin...

Or just overestimate. It's fun to say you averaged 23mph on your last double century. Impressive!

[Ozz]

Sorry, I know how to do all of those things, but as I said I haven't bothered to do any of them....
sounds like you are getting the same response I did when I asked the same question....everyone has the solution, but no one gives you the answer....I'll look up the number I use, but I think it runs about a 1/3 mph high.

sorry dude.

[JohnS]

Won't the final number depend on what tire you use? A michelin 700x23 is a different size than a Conti 700x23.Here's the answer. You have to find out for yourself. Even different batches of the same tire could have a few mm difference in their diameter. Even tire pressure could make a small difference if the tire stretches.

[dvs cycles]

Won't the final number depend on what tire you use? A michelin 700x23 is a different size than a Conti 700x23.As well as size 25 is bigger than a 23. Weighted rollout is the only way to have accuracy. :)

[cadence231]

Here's the answer. You have to find out for yourself. Even different batches of the same tire could have a few mm difference in their diameter. Even tire pressure could make a small difference if the tire stretches.


True that.

[PaulE]

Sorry, I know how to do all of those things, but as I said I haven't bothered to do any of them.

What I was asking, poorly it seems, was what number did you get when you did one of those things.

keno

Is the number I've gotten on 2 different bikes rolling out the front tire one revolution with my weight on it, marking the driveway with chalk when the valve stem is perpendicular to it, measuring the distance with a tape measure in inches and multiplying by 25.4. Both bikes had new Vittoria Open Pro tires. YMMMV (your milimeters may vary) Now I just use that 2096. :)

[Dave]

I got to thinking about the percentage of error and realized that you have to be off by over 20mm per revolution to have a 1% error. It seems like a three revolution rollout should keep the error to a minimum.

[keno]

I know my answer. I use 700x23 Michelin Pro Race 2. My bike and I weigh about 200#. I pump to 110 psi. It ain't 2096, no how, no where. Show me yours and I'll show you mine.

That's that.

keno

[Dave]

I know my answer. I use 700x23 Michelin Pro Race 2. My bike and I weigh about 200#. I pump to 110 psi. It ain't 2096, no how, no where. Show me yours and I'll show you mine.

That's that.

keno

I've got the same tire. I'll try to do a 3 revolution measurement and post my results, but remember we don't all weigh the same either. I've been using 2110 for years, based on some measurement I took long ago. My bet is the answer comes out within 1% of 2096. In that case you might exaggerate you mileage by 10 miles for each 1000 that you ride. Is that a big deal?

[Dave]

I did a 3 revolution weighted rollout this morning, with a 23mm Pro 2 race and came up with 2095 as a very good number for somone with my 140 lb weight. A hevier rider might want to put in 2090. I sat on the top tube to put more weight on the front and got a reading as low as 2090.

What I did was mark a starting point, and four marks representing values for 2080 through 2110 (all times three of course). That made it easy to eyeball where the third revolution ended, with the valve stem at the bottom. Every one of my tests runs was right between the 2090 and 2100 marks.

I also checked my 700 x 28 Conti 4-seasons winter tire and came up with 2105 as a good number.

It seems that I've been exaggerating my mileage by nearly 1% for awhile.

[RPS]

Clearly, those manuals overstate the rolliing circumference by reason of the fact that the tire on the wheel from which the measurment is determined is depressed a bit by some portion of the weight of the bicycle and myself, and in that way decreasing the effective radius, therefore circumference, of the tire. In my own case, that's about 205 total on two 700x23 tires at 110 psi.I rolled mine at 100 PSI loaded, and then rolled back without my weight, and the difference in one revolution was only about 1 CM. Did you mean 205 MM?

Has anyone felt strongly enough about this to try to determine the truth for yourself? I have to believe that any of you who worry about gms would be inclined to futz with this, too. My sense for myself is about a 2%+ overstatement.I don't worry about grams, but have calibrated tire size for needed accuracy. My 700X23 Michelin Pro2 at 100 PSI came out at 2090 MM.

Before you ask, the tire has some wear and I didn't calibrate my pump gauge. However, the 2096 number for a new tire looks good to me -- well within 1 percent of what I'd get.

[jerk]

ride a measured mile and correct by the percentage it is off.

jerk

[Dave]

I get about the same result as RPS. If I do a 3-revolution rollout with no weight, it increases the tire circumference to from 2095 to 2108. I can't imagine putting enough weight on the front to reduce the average circumference to less than 2080. I sat on the top tube, about as far up as I could get and still got about 2085.

[J.Greene]

To make up for computer error I just skipped the brownie at lunch.

JG

[ClutchCargo]

ride a measured mile and correct by the percentage it is off.

jerk

and how, pray tell, should you
measure the mile?
signed,
wise guy
:banana:

[RPS]

and how, pray tell, should you
measure the mile?
signed,
wise guy
:banana:The only time I thought I needed accuracy it turned out it didn’t help at all – mostly for the reason you state above.

I needed to navigate roads during a series of century rides while following cue sheets and because some of the roads were unmarked it made it easy for me to get turned around.

As it turned out the calibrated computer did not help one bit because some days I was long and others short. Apparently the guys who made the cue sheets used different bikes and cars for the different routes, making my computer more accurate than it needed to be for the data I had to work with. That was the last time I worried about it.

[BumpyintheBurgh]

and how, pray tell, should you
measure the mile?
signed,
wise guy
:banana:

1ft-2ft-3ft-4ft-6ft-oops! 1ft-2ft-3ft...............................5280ft, whew!

[vaxn8r]

and how, pray tell, should you
measure the mile?
signed,
wise guy
:banana:
Easy, with your Garmin. It's accurate within a couple of feet.

[stevep]

i thk its useless exercise.
when yr racing you'll know how you stand.
accurate numbers ain't gonna help you or mph stuff.
i thk its unimportant what that fool thing tells you.
the guage is the race...you best some guys and some guys will beat you.
hopefully the next week you can beat some of the guys who beat you.
go by that.

but yr an engineer. im not.




While ordinarily not of much interest, if you are training for a time trial or other endurance race against the clock, it becomes an important thing to know in order to determine how competitive you will be. If 25 mph is your goal and that's what your bike computer is telling you, and what you've actually been doing is about 24.5mph, not good. There will be some 'splainin to do.

keno

[JohnS]

i thk its useless exercise.

.
Exercise is useless? You mean I've been wasting my time all these years? :confused:

[stevep]

loose interpretation but pretty much
you wont live forever no matter.
just buy a new bike and worry later.


Exercise is useless? You mean I've been wasting my time all these years? :confused:

[Fixed]

bro imho you are going half /fast if you have to look at your computer .they get in the way of the fun ...for me imho :beer: cheers