Cayeye tire size chart for cycle computers

Re: Cayeye tire size chart for cycle computers

I for one, never bother with manufacturers charts. Too much variation among various tire makers actual circumference. Mark, roll, and measure is the only reliable method IMHO!
Dan

Re: Cayeye tire size chart for cycle computers

Rik O'Shea wrote:
> I just inspected the Cateye tire size chart (well it was a wet night on
> Monday) - this is the chart that relates tyre size to tire
> circumference. I noticed what I think is a discrepency.
>
> Tire size => Circumerference (mm)
> 700 x 28C => 2136
> 700 x 30C => 2170
> 700 x 32C => 2155
>
> How is it that when you go to the wider 32C tire that the circumference
> reduces ? I think the "error" maybe in the 30C value as there is quite
> a big jump from 28C to 30C (34 mm) which is not consistent with the
> other increments in the chart.
>

The only "700 x 30C" tire that I'm familiar with is the IRC Tandem EX,
which has a tall, ellipitical cross-section. This could give it a
larger circumfrence than a rounder 700 x 32C tire. It may be that
they're reporting "real" measurements.

Or it could be a typo.

Jeff

Re: Cayeye tire size chart for cycle computers

anonymous writes:

> I just inspected the Cateye tire size chart (well it was a wet night
> on Monday) - this is the chart that relates tyre size to tire
> circumference. I noticed what I think is a discrepancy.

> Tire size => Circumference (mm)
> 700 x 28C => 2136
> 700 x 30C => 2170
> 700 x 32C => 2155

> How is it that when you go to the wider 32C tire that the
> circumference reduces ? I think the "error" maybe in the 30C value
> as there is quite a big jump from 28C to 30C (34 mm) which is not
> consistent with the other increments in the chart.

> Here is the section for 27" tires:

> Tire size => Circumference (mm)
> 27 x 1 => 2145
> 27 x 1-1/8 => 2155
> 27 x 1-1/4 => 2161
> 27 x 1-3/8 => 2169

> Notice that the mm increments are smaller and Cateye seem to
> directly equate a 700 x 32C tire with a 27 x 1-1/8.

> I noticed that on Cateye's on-line manuals they now express
> circumference in cm (rounding to 3 digits) rather than mm but this
> discrepancy is still there.

> PS before people go off on tangents about "actual" circumference is
> based on rider weight, inflation pressure, blah, blah - yes correct
> but the concern is the "reference" circumference in the the chart.

So if you know all this, why do you care what these charts say when in
reality, for people who want to have accurate readings measure the
roll-out distance and enter it in the instrument?

If you missed it, it's done as follows. With the tire inflated to its
preferred pressure and the rider on the bicycle in the preferred
position, place the front wheel with the valve stem directly over a
line on the floor (floor tile or the like) and roll forward until the
stem is again in its vertical position. Measure the distance in the
units used by the Cyclometer and enter that number.

Jobst Brandt

Re: Cayeye tire size chart for cycle computers

jobst.brandt@stanfordalumni.org wrote:

>anonymous writes:
>
>
>
>>I just inspected the Cateye tire size chart (well it was a wet night
>>on Monday) - this is the chart that relates tyre size to tire
>>circumference. I noticed what I think is a discrepancy.
>>
>>
>
>
>
>>Tire size => Circumference (mm)
>>700 x 28C => 2136
>>700 x 30C => 2170
>>700 x 32C => 2155
>>
>>
>
>
>
>>How is it that when you go to the wider 32C tire that the
>>circumference reduces ? I think the "error" maybe in the 30C value
>>as there is quite a big jump from 28C to 30C (34 mm) which is not
>>consistent with the other increments in the chart.
>>
>>
>
>
>
>>Here is the section for 27" tires:
>>
>>
>
>
>
>>Tire size => Circumference (mm)
>>27 x 1 => 2145
>>27 x 1-1/8 => 2155
>>27 x 1-1/4 => 2161
>>27 x 1-3/8 => 2169
>>
>>
>
>
>
>>Notice that the mm increments are smaller and Cateye seem to
>>directly equate a 700 x 32C tire with a 27 x 1-1/8.
>>
>>
>
>
>
>>I noticed that on Cateye's on-line manuals they now express
>>circumference in cm (rounding to 3 digits) rather than mm but this
>>discrepancy is still there.
>>
>>
>
>
>
>>PS before people go off on tangents about "actual" circumference is
>>based on rider weight, inflation pressure, blah, blah - yes correct
>>but the concern is the "reference" circumference in the the chart.
>>
>>
>
>So if you know all this, why do you care what these charts say when in
>reality, for people who want to have accurate readings measure the
>roll-out distance and enter it in the instrument?
>
>If you missed it, it's done as follows. With the tire inflated to its
>preferred pressure and the rider on the bicycle in the preferred
>position, place the front wheel with the valve stem directly over a
>line on the floor (floor tile or the like) and roll forward until the
>stem is again in its vertical position. Measure the distance in the
>units used by the Cyclometer and enter that number.
>
>Jobst Brandt
>
>
I find it difficult to get an accurate measurement using a rollout. The
minimum difference on my computer is 0.5cm. I know from riding surveyed
routes and measured miles that the correct setting is 211.0 (catalogue
data says 210.5) but I've never been able to actually produce that value
from a rollout. I've used your valve stem method and I've wetted my
tire and measured the space between the wet spots. A 1cm measurement
error is about 1/2 % output error. The perfectionist in me wants to
make the rollout method work and understand what I'm doing wrong, but
the pragmatist already knows the right answer.

I suggest riding a measured course for best accuracy - or you can
measure your own course with GPS or even a map making program. Use the
catalog value, ride the course, make a correction. This has the added
advantage of accounting for real world conditions of not riding
perfectly straight lines; after all, we're not really after perfect
accuracy so much as we're after perfect agreement with correct signage
and route maps, right?

Re: Cayeye tire size chart for cycle computers

Diablo Scott <N0SPAMdiabloscott@terra.es> writes:

>>> PS before people go off on tangents about "actual" circumference
>>> is based on rider weight, inflation pressure, blah, blah - yes
>>> correct but the concern is the "reference" circumference in the
>>> the chart.

>> So if you know all this, why do you care what these charts say when
>> in reality, for people who want to have accurate readings measure
>> the roll-out distance and enter it in the instrument?

>> If you missed it, it's done as follows. With the tire inflated to
>> its preferred pressure and the rider on the bicycle in the
>> preferred position, place the front wheel with the valve stem
>> directly over a line on the floor (floor tile or the like) and roll
>> forward until the stem is again in its vertical position. Measure
>> the distance in the units used by the Cyclometer and enter that
>> number.

> I find it difficult to get an accurate measurement using a rollout.
> The minimum difference on my computer is 0.5cm. I know from riding
> surveyed routes and measured miles that the correct setting is 211.0
> (catalogue data says 210.5) but I've never been able to actually
> produce that value from a rollout. I've used your valve stem method
> and I've wetted my tire and measured the space between the wet
> spots. A 1cm measurement error is about 1/2 % output error. The
> perfectionist in me wants to make the rollout method work and
> understand what I'm doing wrong, but the pragmatist already knows
> the right answer.

I can measure that distance to +-1.5mm which is better than an
estimate from a table that knows neither the inflation pressure nor
the load. Where do you have difficulty? My Cyclometer accepts mm
and my wheel is 2096mm.

> I suggest riding a measured course for best accuracy - or you can
> measure your own course with GPS or even a map making program. Use
> the catalog value, ride the course, make a correction. This has the
> added advantage of accounting for real world conditions of not
> riding perfectly straight lines; after all, we're not really after
> perfect accuracy so much as we're after perfect agreement with
> correct signage and route maps, right?

That is definitely untrue because that depends on how you take
corners, inside or out. The only one that might work is a straight
line course. For that we have California HWY1 with aircraft speed
check markers that are placed at one mile intervals to great accuracy
from Pigeon Point to Santa Cruz. I haven't checked that but it's
an idea I should follow up.

Jobst Brandt

Re: Cayeye tire size chart for cycle computers

jobst.brandt@stanfordalumni.org wrote:

>Diablo Scott <N0SPAMdiabloscott@terra.es> writes:
>
>
>
>>I find it difficult to get an accurate measurement using a rollout.
>>The minimum difference on my computer is 0.5cm. I know from riding
>>surveyed routes and measured miles that the correct setting is 211.0
>>(catalogue data says 210.5) but I've never been able to actually
>>produce that value from a rollout. I've used your valve stem method
>>and I've wetted my tire and measured the space between the wet
>>spots. A 1cm measurement error is about 1/2 % output error. The
>>perfectionist in me wants to make the rollout method work and
>>understand what I'm doing wrong, but the pragmatist already knows
>>the right answer.
>>
>>
>
>I can measure that distance to +-1.5mm which is better than an
>estimate from a table that knows neither the inflation pressure nor
>the load. Where do you have difficulty? My Cyclometer accepts mm
>and my wheel is 2096mm.
>
>
>
Sitting on the bike in riding position and rolling out a straight line
seems to be the difficult part. A 1.5mm difference between your wheel's
true path and the point-to-point distance is pretty impressive, if you
really did it. The fact that you can measure between the two points to
such an accuracy is not the same thing of course.

>>I suggest riding a measured course for best accuracy - or you can
>>measure your own course with GPS or even a map making program. Use
>>the catalog value, ride the course, make a correction. This has the
>>added advantage of accounting for real world conditions of not
>>riding perfectly straight lines; after all, we're not really after
>>perfect accuracy so much as we're after perfect agreement with
>>correct signage and route maps, right?
>>
>>
> <>
> That is definitely untrue because that depends on how you take
> corners, inside or out. The only one that might work is a straightline
> course.

Yes but these are real world conditions and we generally have the same
number of left and right turns. Maybe I've introduced an unintentional
fudge factor to my calibration that accounts for the occaisional sprint
or stretch or pothole dodging maneuver, but I don't care if my distance
function perfectly tracks my rolling contact patch; I want it to
perfectly match the route sheet for the century I'm riding or the map I
downloaded.

> <>For that we have California HWY1 with aircraft speed
> check markers that are placed at one mile intervals to great accuracy
> from Pigeon Point to Santa Cruz. I haven't checked that but it's
> an idea I should follow up.
>
> Jobst Brandt

Yes, you might be surprised. And a straightline course would be a
better control than one with lots of turns. Ride normally, not like
you're trying to stay on a wire.

Re: Cayeye tire size chart for cycle computers

On Thu, 12 Jan 2006 16:11:13 -0800, Diablo Scott
<N0SPAMdiabloscott@terra.es> wrote:

>jobst.brandt@stanfordalumni.org wrote:
>
>>Diablo Scott <N0SPAMdiabloscott@terra.es> writes:
>>
>>
>>
>>>I find it difficult to get an accurate measurement using a rollout.
>>>The minimum difference on my computer is 0.5cm. I know from riding
>>>surveyed routes and measured miles that the correct setting is 211.0
>>>(catalogue data says 210.5) but I've never been able to actually
>>>produce that value from a rollout. I've used your valve stem method
>>>and I've wetted my tire and measured the space between the wet
>>>spots. A 1cm measurement error is about 1/2 % output error. The
>>>perfectionist in me wants to make the rollout method work and
>>>understand what I'm doing wrong, but the pragmatist already knows
>>>the right answer.

>>
>Sitting on the bike in riding position and rolling out a straight line
>seems to be the difficult part. A 1.5mm difference between your wheel's
>true path and the point-to-point distance is pretty impressive, if you
>really did it. The fact that you can measure between the two points to
>such an accuracy is not the same thing of course.
>
>>>I suggest riding a measured course for best accuracy - or you can
>>>measure your own course with GPS or even a map making program. Use
>>>the catalog value, ride the course, make a correction. This has the
>>>added advantage of accounting for real world conditions of not
>>>riding perfectly straight lines; after all, we're not really after
>>>perfect accuracy so much as we're after perfect agreement with
>>>correct signage and route maps, right?
>>>
>>>
>> <>
>
>Yes, you might be surprised. And a straightline course would be a
>better control than one with lots of turns. Ride normally, not like
>you're trying to stay on a wire.


If you would like to improve the accuracy of a "Roll Out", do a
measured 100 feet or so..
I have a long side walk in front of my house and I have measured out
100' with a tape and marked it on the pavement. I use it all the time
for speedo calibration..
Seems very accurate, and yes, the cal factors vary quite a bit from
those in the "Emanuel". <G>
Just check your tire pressure, ride slowly and estimate the change at
the end. It is very repeatable.

Bob F.

Re: Cayeye tire size chart for cycle computers

Rik O'Shea writes:

> Yes, I've always used the tire rollout method when calibrating my
> cycle computer but today out of curiosity I took a straw poll among
> some of the commuter friends at work who use a cycle computer and
> guess what, 4/4 used the chart that came with their cycle computer!

Yes? How large is your sample, and what conclusion do you expect
others to draw from that?

Jobst Brandt

Re: Cayeye tire size chart for cycle computers

Diablo Scott writes:

>>> I find it difficult to get an accurate measurement using a
>>> rollout. The minimum difference on my computer is 0.5cm. I know
>>> from riding surveyed routes and measured miles that the correct
>>> setting is 211.0 (catalogue data says 210.5) but I've never been
>>> able to actually produce that value from a rollout. I've used
>>> your valve stem method and I've wetted my tire and measured the
>>> space between the wet spots. A 1cm measurement error is about 1/2
>>> % output error. The perfectionist in me wants to make the rollout
>>> method work and understand what I'm doing wrong, but the
>>> pragmatist already knows the right answer.

>> I can measure that distance to +-1.5mm which is better than an
>> estimate from a table that knows neither the inflation pressure nor
>> the load. Where do you have difficulty? My Cyclometer accepts mm
>> and my wheel is 2096mm.

> Sitting on the bike in riding position and rolling out a straight
> line seems to be the difficult part. A 1.5mm difference between
> your wheel's true path and the point-to-point distance is pretty
> impressive, if you really did it. The fact that you can measure
> between the two points to such an accuracy is not the same thing of
> course.

I think you can summon the skill to roll straight ahead for one wheel
revolution, however, indoors you can use a hand on the wall if you
like. This is not a good excuse for deferring to a probably
inappropriate chart.

>>> I suggest riding a measured course for best accuracy - or you can
>>> measure your own course with GPS or even a map making program.
>>> Use the catalog value, ride the course, make a correction. This
>>> has the added advantage of accounting for real world conditions of
>>> not riding perfectly straight lines; after all, we're not really
>>> after perfect accuracy so much as we're after perfect agreement
>>> with correct signage and route maps, right?

>> That is definitely untrue because that depends on how you take
>> corners, inside or out. The only one that might work is a
>> straight line course.

> Yes but these are real world conditions and we generally have the
> same number of left and right turns.

Are you suggesting that in the "real world" distances traveled are
different from precise measurement? I don't understand your point.

> Maybe I've introduced an unintentional fudge factor to my
> calibration that accounts for the occaisional sprint or stretch or
> pothole dodging maneuver, but I don't care if my distance function
> perfectly tracks my rolling contact patch; I want it to perfectly
> match the route sheet for the century I'm riding or the map I
> downloaded.

I see. you don't care if the rider who measured the course was off by
an amount, you want some arbitrary correlation counter. Then you had
best get your calibration number from the course measurer rather than
entering an accurate one.

>> For that we have California HWY1 with aircraft speed check markers
>> that are placed at one mile intervals to great accuracy from
>> Pigeon Point to Santa Cruz. I haven't checked that but it's an
>> idea I should follow up.

> Yes, you might be surprised. And a straight line course would be a
> better control than one with lots of turns. Ride normally, not like
> you're trying to stay on a wire.

Are you suggesting that in a straight ahead level mile I cannot
control the line enough to get an accurate reading? I suspect you
haven't tried riding on such a route, especially with a good tailwind
that often prevails on the Coast Highway in that direction.

Jobst Brandt

Re: Cayeye tire size chart for cycle computers

jobst.brandt@stanfordalumni.org wrote:

>
>
> Are you suggesting that in a straight ahead level mile I cannot
> control the line enough to get an accurate reading? I suspect you
> haven't tried riding on such a route, especially with a good tailwind
> that often prevails on the Coast Highway in that direction.
>
> Jobst Brandt

I'm suggesting that in real riding situations, the total distance
travelled as measured along the actual path of the tire's contact patch
(which is what you'd have with a "perfectly" calibrated computer),
doesn't necessarily correspond to the actual sum of straight line
distances between boundry points (which I contend is what you {I} really
want to know). No comment at all on your ability to ride a line over a
flat straight mile with a tail wind, other than perhaps it's better than
my own.

We're arguing small errors here, but I did several rollouts and entered
the average (small deltas) and I was surprised at how far off I was when
I rode a surveyed and stationed road course (neither straight nor
tailwindy). Maybe you will be surprised too. Looks like crappy weather
on Saturday - do it on Sunday.

--
Check out my bike blog!
http://diabloscott.blogspot.com

Re: Cayeye tire size chart for cycle computers

On 12 Jan 2006 15:47:02 -0800, "Rik O'Shea" <rikoshea@yahoo.com>
wrote:

>Yes, I've always used the tire rollout method when calibrating my cycle
>computer but today out of curiosity I took a straw poll among some of
>the commuter freds in work who use a cycle computer and guess what, 4/4
>used the chart that came with their cycle computer !

When measuring rollout, I usually measure the distance of 2
revolutions w/o riding the bike. For a 700x23 tire, I usually select
2090 mm and then ride to a spot that is 11.88 mi from my house. That
spot was verified with a GPS several times. I then readjust the
'puter.

If you are super compulsive, put 2 magnets on when doing the test ride
and divide by 2.

I really have no circumstances when I need such accuracy and
frequently ride w/o a computer.

Re: Cayeye tire size chart for cycle computers

Paul Kopit wrote:

>
>When measuring rollout, I usually measure the distance of 2
>revolutions w/o riding the bike. For a 700x23 tire, I usually select
>2090 mm and then ride to a spot that is 11.88 mi from my house. That
>spot was verified with a GPS several times. I then readjust the
>'puter.
>
>If you are super compulsive, put 2 magnets on when doing the test ride
>and divide by 2.
>
>I really have no circumstances when I need such accuracy and
>frequently ride w/o a computer.
>
>
You seem to contradict yourself several times in that post. But your
key point is that you readjust your computer based on the GPS measured
route; signifying that you want your odometer to agree with a mapped
distance rather than having it be perfectly calibrated to your wheel
circumference. That's exactly my point as well and I've got a similar
measured course.

What percent error do you find between the rollout calibration and the
corrected number from GPS?

There are three sources for error. 1 - Mechanical error caused by
problems with the sensor/magnet or the computer itself. 2 - Calibration
error from entering an incorrect wheel circ number. and 3 - Deviation
error from not riding in a straight line.

The third is the most interesting to me. In Jobst's proposed test he'll
ride a measured mile on his bike with the computer that has perfect
calibration. I asked him not to concentrate too much on maintaining a
perfectly straight line but to ride "normally". In normal riding we
swerve around potholes and debris, we take different lines around
corners, we vary our position relative to the edge of the road, we get
out of the saddle and stretch or sprint - there are a whole host of
things we do that make the path of our front wheels different (longer)
than the straight line distance or whatever path GPS uses in its
measurement. Since Jobst is certain of his calibration data, at the end
of his test he'll know exactly how much he's deviated from a straight
line by how much his computer differs from 1.00 miles.

The question is which value we really want our computers to read. By
calibrating to a measured course of 5 miles or so, I have incorporated a
fudge factor that accounts for my riding style and conditions over that
course. By calibrating based on his roll-out, Jobst is recording the
true distance his front wheel has traveled rather than the true distance
between those two points. I'm not sure why he wants to know the former
rather than the latter except maybe he just doesn't like fudge.

Re: Cayeye tire size chart for cycle computers

Diablo Scott writes:

> The third is the most interesting to me. In Jobst's proposed test
> he'll ride a measured mile on his bike with the computer that has
> perfect calibration. I asked him not to concentrate too much on
> maintaining a perfectly straight line but to ride "normally". In
> normal riding we swerve around potholes and debris, we take
> different lines around corners, we vary our position relative to the
> edge of the road, we get out of the saddle and stretch or sprint -
> there are a whole host of things we do that make the path of our
> front wheels different (longer) than the straight line distance or
> whatever path GPS uses in its measurement. Since Jobst is certain
> of his calibration data, at the end of his test he'll know exactly
> how much he's deviated from a straight line by how much his computer
> differs from 1.00 miles.

I don't understand how you think this is preferable. If your
Cyclometer is based on vagaries of riding ability, what then is a mile
or kilometer? How is it measured and why do you want to redefine
distance?

If you want to record bicycling effort, you should include the gear
ratio so that the distance the feet travel is recorded. That would be
like zig-zagging mileage on a straight course.

> The question is which value we really want our computers to read.
> By calibrating to a measured course of 5 miles or so, I have
> incorporated a fudge factor that accounts for my riding style and
> conditions over that course. By calibrating based on his roll-out,
> Jobst is recording the true distance his front wheel has traveled
> rather than the true distance between those two points. I'm not
> sure why he wants to know the former rather than the latter except
> maybe he just doesn't like fudge.

Tis is similar to the altimeters that have no hysteresis threshold
that prevents surface irregularities from entering cumulative
altitude. They add every bump in the road that falls within the
resolution of the instrument, as well as atmospheric fluctuations.
These devices record climbs on essentially level courses between to
towns at the same elevation with no point being more than 20 feet
higher or lower than the endpoints... including railway underpasses.

This increases trip totals and your suggested calibration decreases
distance traveled. It certainly is not accuracy.

Jobst Brandt