| |
Main
Date: 24 Sep 2006 19:46:58
From: Gary Smiley
Subject: FlightDeck Calibration problems
|
I've used a FlightDeck on my road bike for 7 years (Ultegra triple, 700 x 23
tires), and it worked fine until about a month ago. Now, it registers
mileage about 10-12 % too high. I went on a 40 mile ride and it registered
44 miles. So I recalibrated it. Even measured the wheel by rolling the bike
one revolution with me seated on it. But for some reason, it measures way
too high - but - when I am riding alongside a fellow rider and compare
speeds, they are the same. Has this happened to anyone else in this group?
Is there anything I can do about it?
- Gary
|
|
| |
Date: 25 Sep 2006 02:31:45
From: Mike Jacoubowsky
Subject: Re: FlightDeck Calibration problems
|
> I've used a FlightDeck on my road bike for 7 years (Ultegra triple, 700 x
> 23 tires), and it worked fine until about a month ago. Now, it registers
> mileage about 10-12 % too high. I went on a 40 mile ride and it registered
> 44 miles. So I recalibrated it. Even measured the wheel by rolling the
> bike one revolution with me seated on it. But for some reason, it measures
> way too high - but - when I am riding alongside a fellow rider and compare
> speeds, they are the same. Has this happened to anyone else in this group?
> Is there anything I can do about it?
> - Gary
You're using a calibration number in the range of 2080 or thereabouts?
Bike computers basically cannot determine distance incorrectly without also
messing up with the speed. It's a hardwired computation that cannot be
altered or become wrong with time. They can often have problems with maximum
speed without an effect on mileage, because the spikes are momentary and it
doesn't make much difference in the end if, for a couple seconds, it thought
you were traveling at 60mph instead of 20. It might add a hundred feet to
your ride, which is way under the .1 mile display update.
So looks like I only have more questions, not answers for you.
--Mike-- Chain Reaction Bicycles
www.ChainReactionBicycles.com
|
| | |
Date: 26 Sep 2006 15:09:38
From: Buck
Subject: Re: FlightDeck Calibration problems
|
Gary Smiley wrote:
> This is why I posted this thread - it is a mystery. The speed reads the
> same as that of my fellow riders. Every now and then I'll ask my
> friend: "What's your speed now?". And it's always the same as my speed
> - but my distance/mileage is way too high. And no, it's not set to
> kilometers. It's as if the mathematical algorithm for calculating
> distance has failed. Hmm... could it be that the battery is starting to
> go?
That might be the problem. I have found that Cateye computers produce
strange readings when the batteries get low or the computer / mount
connections need to be cleaned.
-Buck
|
| | |
Date: 26 Sep 2006 12:44:51
From: Gary Smiley
Subject: Re: FlightDeck Calibration problems
|
This is why I posted this thread - it is a mystery. The speed reads the
same as that of my fellow riders. Every now and then I'll ask my
friend: "What's your speed now?". And it's always the same as my speed
- but my distance/mileage is way too high. And no, it's not set to
kilometers. It's as if the mathematical algorithm for calculating
distance has failed. Hmm... could it be that the battery is starting to
go?
Chris Y.F.N.W. wrote:
> Bike computers work by counting wheel turns. The number of turns over
> time determines speed, number of turns by itself measures distance, all
> are based upon the wheel size you entered into it. Some computers use
> one measurement along with a little arithmetic to calculate other
> measurements. However all are based upon exactly how big the computer
> _thinks_ your wheel is.
>
> Some units will lose this information when power is lost, such as when
> the battery dies or when changing it. you might want to go into your
> settings and see if the number you originally set is still there.
>
> BTW, simply measuring a wheel turn and punching in a number is SETTING
> the computer, not "calibrating' it. It's not 100% accurate, or even 99%.
> as it doesn't take into account factors that are only encountered while
> actually riding the bike.
>
> To calibrate (test against a measurement of known accuracy) your
> computer, you must ride the bike for an accuratly measured distance (at
> least a few miles) then compare your comp's reading against the known
> distance and make any necessary corrections.
>
> http://www.geocities.com/czcorner/tech17.html
>
> - -
> Comments and opinions compliments of,
> "Your Friendly Neighborhood Wheelman"
>
> Messages sent to the above "E" addy will be lost forever!
>
> to E-mail me:
> ChrisZCorner "at" webtv "dot" net
>
> My website:
> http://geocities.com/czcorner
|
| | | |
Date: 27 Sep 2006 19:41:50
From: Chris Y.F.N.W.
Subject: Re: FlightDeck Calibration problems
|
Group: rec.bicycles.misc
Date: Tue, Sep 26, 2006, 12:44pm (EDT-3)
From: gs@garysmiley.com (Gary=A0Smiley)
>This is why I posted this thread - it is a
>mystery. The speed reads the same as
>that of my fellow riders. Every now and
>then I'll ask my friend: "What's your
>speed now?". And it's always the same
>as my speed
>- but my distance/mileage is way too
>high. And no, it's not set to kilometers.
>It's as if the mathematical algorithm for
>calculating distance has failed. Hmm...
>could it be that the battery is starting to
>go?
Hmm. One of the "factors" affecting comp. readings is how straight a
line you rode whithin the measured distance. Since not everybody follows
the exact same path, there will always be some discrepancy in overall
difference between rider's units, even if current speed reads the same.
How off is "Way off"? if it's a lot (for example 3/4 of a mile after
riding only 25 miles), then try changing the battery. if that doesn't
fix it, then there's probably something more amiss. and I would contact
Shimano. Bike computers are not "user serviceable".
- -
Comments and opinions compliments of,
"Your Friendly Neighborhood Wheelman"
Messages sent to the above "E" addy will be lost forever!
to E-mail me:
ChrisZCorner "at" webtv "dot" net
My website:
http://geocities.com/czcorner
|
| | |
Date: 26 Sep 2006 12:39:41
From: Chris Y.F.N.W.
Subject: Re: FlightDeck Calibration problems
|
Bike computers work by counting wheel turns. The number of turns over
time determines speed, number of turns by itself measures distance, all
are based upon the wheel size you entered into it. Some computers use
one measurement along with a little arithmetic to calculate other
measurements. However all are based upon exactly how big the computer
_thinks_ your wheel is.
Some units will lose this information when power is lost, such as when
the battery dies or when changing it. you might want to go into your
settings and see if the number you originally set is still there.
BTW, simply measuring a wheel turn and punching in a number is SETTING
the computer, not "calibrating' it. It's not 100% accurate, or even 99%.
as it doesn't take into account factors that are only encountered while
actually riding the bike.
To calibrate (test against a measurement of known accuracy) your
computer, you must ride the bike for an accuratly measured distance (at
least a few miles) then compare your comp's reading against the known
distance and make any necessary corrections.
http://www.geocities.com/czcorner/tech17.html
- -
Comments and opinions compliments of,
"Your Friendly Neighborhood Wheelman"
Messages sent to the above "E" addy will be lost forever!
to E-mail me:
ChrisZCorner "at" webtv "dot" net
My website:
http://geocities.com/czcorner
|
| | | |
Date: 27 Sep 2006 00:55:06
From: Mike Jacoubowsky
Subject: Re: FlightDeck Calibration problems
|
> BTW, simply measuring a wheel turn and punching in a number is SETTING
> the computer, not "calibrating' it. It's not 100% accurate, or even 99%.
> as it doesn't take into account factors that are only encountered while
> actually riding the bike.
I think you'd have a difficult time showing that a bike computer is less
than 99% accurate when properly calibrated. The "factors that are only
encountered while actually riding the bike" account for a very tiny
variation in total mileage. To be less than 99% accurate is to suggest that
an error of 1 mile in a century should be considered normal, while I would
argue that an error of .25 miles (representing over 1000 ft of distance
either traveled or not traveled) would be at the upper end of normal.
You can try and come up with all manner of extreme examples (somebody
endlessly weaving back and forth across a wide, long, steep grade?) but in
the real world, cycling computer accuracy is very, very good. And why
shouldn't it be? What could be simpler than measuring the number of turns a
wheel takes, when you know what the diameter of that wheel "actually" is (by
doing a rollout and confirming it based upon known mile kers if you
really want to get picky about it).
--Mike Jacoubowsky
Chain Reaction Bicycles
www.ChainReaction.com
Redwood City & Los Altos, CA USA
"Chris Y.F.N.W." <dedendaddy4spammers@webtv.net > wrote in message
news:15776-451957CD-2734@storefull-3235.bay.webtv.net...
> Bike computers work by counting wheel turns. The number of turns over
> time determines speed, number of turns by itself measures distance, all
> are based upon the wheel size you entered into it. Some computers use
> one measurement along with a little arithmetic to calculate other
> measurements. However all are based upon exactly how big the computer
> _thinks_ your wheel is.
>
> Some units will lose this information when power is lost, such as when
> the battery dies or when changing it. you might want to go into your
> settings and see if the number you originally set is still there.
>
> BTW, simply measuring a wheel turn and punching in a number is SETTING
> the computer, not "calibrating' it. It's not 100% accurate, or even 99%.
> as it doesn't take into account factors that are only encountered while
> actually riding the bike.
>
> To calibrate (test against a measurement of known accuracy) your
> computer, you must ride the bike for an accuratly measured distance (at
> least a few miles) then compare your comp's reading against the known
> distance and make any necessary corrections.
>
> http://www.geocities.com/czcorner/tech17.html
>
> - -
> Comments and opinions compliments of,
> "Your Friendly Neighborhood Wheelman"
>
> Messages sent to the above "E" addy will be lost forever!
>
> to E-mail me:
> ChrisZCorner "at" webtv "dot" net
>
> My website:
> http://geocities.com/czcorner
>
|
| | | |
Date: 26 Sep 2006 20:42:26
From: Paul Hobson
Subject: Re: FlightDeck Calibration problems
|
Chris Y.F.N.W. wrote:
> BTW, simply measuring a wheel turn and punching in a number is SETTING
> the computer, not "calibrating' it. It's not 100% accurate, or even 99%.
> as it doesn't take into account factors that are only encountered while
> actually riding the bike.
>
> To calibrate (test against a measurement of known accuracy) your
> computer, you must ride the bike for an accuratly measured distance (at
> least a few miles) then compare your comp's reading against the known
> distance and make any necessary corrections.
Seems like this would be a useless exercise due to the infinitesimal
probability that you're going to ride perfectly along the same line* as
the road (or trail) was surveyed. Why not just just measure the
circumference of the wheel (in any number of ways) several times and
take the average? (that's what you do with proctor molds in a soils lab)
\\paul
*most streets are surveyed along the center, IIRC
|
| | | | |
Date: 27 Sep 2006 20:22:30
From: Chris Y.F.N.W.
Subject: Re: FlightDeck Calibration problems
|
Group: rec.bicycles.misc
Date: Tue, Sep 26, 2006, 8:42pm
From: gtg611a@mail.gatech.edu (Paul=A0Hobson)
>Seems like this would be a useless
>exercise due to the infinitesimal
>probability that you're going to ride
>perfectly along the same line* as the
>road (or trail) was surveyed. Why not
>just just measure the circumference of
>the wheel (in any number of ways)
>several times and take the average?
My point exactly. Since there's no way you are going to ride "perfectly
straight" that is no wavering in your course, then your wheel is going
to travel a longer path than the road does. Simple geometry.
However, by riding a measured distance of, say 10 miles, then your
"weaving" will be taken into account (along with other factors). And
after riding the "surveyed" ten miles, the distance your computer reads
will also be ten miles. IOW, this will be that actual distance as if you
HAD ridden a perfect, unwavering line. The link I gave in my first post
will take you to a page that explains this better.
Simply measuring the wheel distance, even over a distance of four turns,
with my weight on the bike, only gave me an accuracy of 97%, or about as
accurate as a car odometer. OK, so I'm a perfectionist. This is probably
accurate enough for most people, but I like the 99.99% I get now a lot
better :-3)
- -
Comments and opinions compliments of,
"Your Friendly Neighborhood Wheelman"
Messages sent to the above "E" addy will be lost forever!
to E-mail me:
ChrisZCorner "at" webtv "dot" net
My website:
http://geocities.com/czcorner
|
| | | | | |
Date: 28 Sep 2006 01:26:41
From: Paul Hobson
Subject: Re: FlightDeck Calibration problems
|
Chris Y.F.N.W. wrote:
> Group: rec.bicycles.misc
> Date: Tue, Sep 26, 2006, 8:42pm
> From: gtg611a@mail.gatech.edu (Paul Hobson)
>
>> Seems like this would be a useless
>> exercise due to the infinitesimal
>> probability that you're going to ride
>> perfectly along the same line* as the
>> road (or trail) was surveyed. Why not
>> just just measure the circumference of
>> the wheel (in any number of ways)
>> several times and take the average?
>
> My point exactly. Since there's no way you are going to ride "perfectly
> straight" that is no wavering in your course, then your wheel is going
> to travel a longer path than the road does. Simple geometry.
>
> However, by riding a measured distance of, say 10 miles, then your
> "weaving" will be taken into account (along with other factors). And
> after riding the "surveyed" ten miles, the distance your computer reads
> will also be ten miles. IOW, this will be that actual distance as if you
> HAD ridden a perfect, unwavering line. The link I gave in my first post
> will take you to a page that explains this better.
Ahhh...gotcha. I must clarify: I'm a first semester grad student taking
two classes with very intense labs *and* teaching an undergrad lab
section *and* taking a data analysis/modeling class centered around
what's measured vs. what do dynamics predict vs. what's "correct*". As
such, I've lost touch with what is real and what is useful.
\\paul
*I don't even know what that words means now
|
| | | | | | |
Date: 28 Sep 2006 08:25:18
From: bill
Subject: Re: FlightDeck Calibration problems
|
Paul Hobson wrote:
> Chris Y.F.N.W. wrote:
>> Group: rec.bicycles.misc Date: Tue, Sep 26, 2006, 8:42pm From:
>> gtg611a@mail.gatech.edu (Paul Hobson)
>>> Seems like this would be a useless
>>> exercise due to the infinitesimal
>>> probability that you're going to ride
>>> perfectly along the same line* as the
>>> road (or trail) was surveyed. Why not
>>> just just measure the circumference of
>>> the wheel (in any number of ways)
>>> several times and take the average?
>>
>> My point exactly. Since there's no way you are going to ride "perfectly
>> straight" that is no wavering in your course, then your wheel is going
>> to travel a longer path than the road does. Simple geometry.
>> However, by riding a measured distance of, say 10 miles, then your
>> "weaving" will be taken into account (along with other factors). And
>> after riding the "surveyed" ten miles, the distance your computer reads
>> will also be ten miles. IOW, this will be that actual distance as if you
>> HAD ridden a perfect, unwavering line. The link I gave in my first post
>> will take you to a page that explains this better.
>
> Ahhh...gotcha. I must clarify: I'm a first semester grad student taking
> two classes with very intense labs *and* teaching an undergrad lab
> section *and* taking a data analysis/modeling class centered around
> what's measured vs. what do dynamics predict vs. what's "correct*". As
> such, I've lost touch with what is real and what is useful.
> \\paul
> *I don't even know what that words means now
Too much college and not enough real world. Been there, done that, seems
like a lifetime ago. Things still don't always go as predicted, either
by education or experience. Exceptions seem to be the rule.
Bill Baka
|
| | | | | | | |
Date: 28 Sep 2006 08:55:45
From: R Brickston
Subject: Re: FlightDeck Calibration problems
|
On Thu, 28 Sep 2006 08:25:18 GMT, bill <bbaka@syix.com > wrote:
>Paul Hobson wrote:
>> Chris Y.F.N.W. wrote:
>>> Group: rec.bicycles.misc Date: Tue, Sep 26, 2006, 8:42pm From:
>>> gtg611a@mail.gatech.edu (Paul Hobson)
>>>> Seems like this would be a useless
>>>> exercise due to the infinitesimal
>>>> probability that you're going to ride
>>>> perfectly along the same line* as the
>>>> road (or trail) was surveyed. Why not
>>>> just just measure the circumference of
>>>> the wheel (in any number of ways)
>>>> several times and take the average?
>>>
>>> My point exactly. Since there's no way you are going to ride "perfectly
>>> straight" that is no wavering in your course, then your wheel is going
>>> to travel a longer path than the road does. Simple geometry.
>>> However, by riding a measured distance of, say 10 miles, then your
>>> "weaving" will be taken into account (along with other factors). And
>>> after riding the "surveyed" ten miles, the distance your computer reads
>>> will also be ten miles. IOW, this will be that actual distance as if you
>>> HAD ridden a perfect, unwavering line. The link I gave in my first post
>>> will take you to a page that explains this better.
>>
>> Ahhh...gotcha. I must clarify: I'm a first semester grad student taking
>> two classes with very intense labs *and* teaching an undergrad lab
>> section *and* taking a data analysis/modeling class centered around
>> what's measured vs. what do dynamics predict vs. what's "correct*". As
>> such, I've lost touch with what is real and what is useful.
>> \\paul
>> *I don't even know what that words means now
>
>Too much college and not enough real world. Been there, done that, seems
>like a lifetime ago. Things still don't always go as predicted, either
>by education or experience. Exceptions seem to be the rule.
>Bill Baka
Where have you been? You got that two weeks in jail, Baka? You should
stay on your own planet.
|
| | | | | | | | |
Date: 29 Sep 2006 19:28:05
From: bill
Subject: Re: FlightDeck Calibration problems
|
R Brickston wrote:
> On Thu, 28 Sep 2006 08:25:18 GMT, bill <bbaka@syix.com> wrote:
>
>> Paul Hobson wrote:
>>> Chris Y.F.N.W. wrote:
>>>> Group: rec.bicycles.misc Date: Tue, Sep 26, 2006, 8:42pm From:
>>>> gtg611a@mail.gatech.edu (Paul Hobson)
>>>>> Seems like this would be a useless
>>>>> exercise due to the infinitesimal
>>>>> probability that you're going to ride
>>>>> perfectly along the same line* as the
>>>>> road (or trail) was surveyed. Why not
>>>>> just just measure the circumference of
>>>>> the wheel (in any number of ways)
>>>>> several times and take the average?
>>>> My point exactly. Since there's no way you are going to ride "perfectly
>>>> straight" that is no wavering in your course, then your wheel is going
>>>> to travel a longer path than the road does. Simple geometry.
>>>> However, by riding a measured distance of, say 10 miles, then your
>>>> "weaving" will be taken into account (along with other factors). And
>>>> after riding the "surveyed" ten miles, the distance your computer reads
>>>> will also be ten miles. IOW, this will be that actual distance as if you
>>>> HAD ridden a perfect, unwavering line. The link I gave in my first post
>>>> will take you to a page that explains this better.
>>> Ahhh...gotcha. I must clarify: I'm a first semester grad student taking
>>> two classes with very intense labs *and* teaching an undergrad lab
>>> section *and* taking a data analysis/modeling class centered around
>>> what's measured vs. what do dynamics predict vs. what's "correct*". As
>>> such, I've lost touch with what is real and what is useful.
>>> \\paul
>>> *I don't even know what that words means now
>> Too much college and not enough real world. Been there, done that, seems
>> like a lifetime ago. Things still don't always go as predicted, either
>> by education or experience. Exceptions seem to be the rule.
>> Bill Baka
>
> Where have you been? You got that two weeks in jail, Baka? You should
> stay on your own planet.
Brickston,
You need your own group. Alt.misc.idiots.
I have had to tutor fresh Ph.D's on life in the real world or
electronics. Some were so flustered at the results of their breadboards
that didn't works and they could not figure out why because the
simulations and equations all made sense. Like the OP was saying, there
is data analysis and modeling and then there is the real deal, both of
which do not usually act the same. Add some layers of complexity and
things get even stranger. I have had some 'students' with higher level
degrees than me that just couldn't wrap their heads around the real
world outcomes of some things. Since it is electronics I will just say
that a schematic simulation does not take into account things like wire
resistance, loose connections, and all the things that happen in the
real world.
That is as far as I am going on this one.
Bill Baka
|
|
