Computer math

[email hidden] wrote in message news:<[email hidden]>...

Quoted message said:

On 21 Sep 2004 01:02:18 -0700, [email hidden] (John

McGraw) said:

"tcmedara" <[email hidden]> wrote in message news:<m4L3d.3033$%t3.1720@lakeread01>...

Quoted message said:

John McGraw <[email hidden]> wrote:
> "S o r n i" <[email hidden]> wrote in message
> news:<[email hidden]>...
>> tcmedara wrote:
>>> Took the boy wonder out for an off-road adventure ride today. As I
>>> trailed along somewhat bored watching boy wonder negotiate the trail
>>> ahead with the speed of an overly cautious 10 year old, I started
>>> fiddling with my computer.
>>>
>
> I imagine your Cateye works fine. Read the instructions. They explain
> it very well. If you can't find the instructions, look em up on line
> at "cateye"
>
> Good luck, John

It seems to work well enough for my uses ("skirt on a bull" and all that), I
was merely curious. I read through the instructions and didn't find much by
way of explaination. The instruction sheet does say average speed is ±0.3
mph. Obviously my readings were well within that range, but it begs the
question. How is the figure derived if the only two data points are
distance and time? The average is not a direct measurement so it must be
derived from the data available to the machine and then go through some sort
of calculation. In this case the calculation itself introduced the error.
Bizarre events for my analog brain!

I'll go with Dave Lehnen's response elsewhere in the thread. Essentially:
low budget performance for a low budget computer. That I can grasp.

Tom

Sorry, I must have missed something. I'll reread the thread. It's
always made sense to me, but then maybe I've been blissfully ignorant.
For >15 yrs I've used Avocet's w/ the ring around the hub & have been
astounded by their accuracy. But recently I've had some physical
problems w/ Avocets, & started using Cateyes. Perhaps I should check
the cateye accuracy.
John

Dear John,

Here's the problem from the original post:

Right at the hour mark I managed to flip through the Cateye
Endro 2's functions and noted:
Time: 1:00
Distance 5.00 miles
Ave: 5.2 mph *<-----------???????*

Given 60 minutes and 5.00 miles, the question is why is the
average calculated as 5.2 mph?

If it's any reassurance, the accuracy of individual
measurements is likely to be ridiculously precise. That is,
the wheel probably made enough turns at its calibration of
20xx mm per revolution to reach 5.00 miles instead of 5.01
or 4.99, and the clock probably measured 1:00 hours far more
precisely than plus or minus 60 seconds.

So the problem is likely

a) bad programming (oops!)

b) odd rounding (the average function reads only in
odd-looking steps such as 4.3 4.8, 5.2, 5.7)

c) the use of a different internal timer for averaging (the
1:00 is total time since the ride began, while the hidden
internal timer noted when the wheel paused here or there on
the ride and reads a little over 2 minutes lower)

Whatever the explanation, I for one will not sleep well
until I have more confidence in the instruments that tell me
the rate at which I'm having fun.

Carl Fogel

I thought that the OP was probably aware that, now a days, most
cyclometers stop recording time for average speed when the speed drops
below a set speed. Usually ~2mph. Did the OP stop for a few seconds
during his ride? that would account for it, if the time on the ride
timer doesn't stop, & the average speed does.
Best, John

On 28 Sep 2004 18:42:38 -0700, [email hidden] (John
McGraw) wrote:

[snip]

Quoted message said:

I thought that the OP was probably aware that, now a days, most
cyclometers stop recording time for average speed when the speed drops
below a set speed. Usually ~2mph. Did the OP stop for a few seconds
during his ride? that would account for it, if the time on the ride
timer doesn't stop, & the average speed does.
Best, John

Dear John,

Hmmm . . . this is the first suggestion that makes sense.

So the idea is that there's a hidden timer just for average
speed that's separate from the visible elapsed-time? And
this timer goes to sleep when the speed drops to under 2
mph?

Given the 5 mph average, it seems plausible that the rider
might have dropped under 2 mph now and then. And a lower
recorded time would produce a higher average speed.

Now I have to go out to the garage, hang my bike in the
stand, zero things, and gently twirl the wheel at under 2
mph to see if it works on my cheap Nashbar computer.

Thanks,

Carl Fogel

On Tue, 28 Sep 2004 20:04:06 -0600, [email hidden]

Quoted message said:

On 28 Sep 2004 18:42:38 -0700, [email hidden] (John
McGraw) wrote:

[snip]

Quoted message said:

I thought that the OP was probably aware that, now a days, most
cyclometers stop recording time for average speed when the speed drops
below a set speed. Usually ~2mph. Did the OP stop for a few seconds
during his ride? that would account for it, if the time on the ride
timer doesn't stop, & the average speed does.
Best, John

Dear John,

Hmmm . . . this is the first suggestion that makes sense.

So the idea is that there's a hidden timer just for average
speed that's separate from the visible elapsed-time? And
this timer goes to sleep when the speed drops to under 2
mph?

Given the 5 mph average, it seems plausible that the rider
might have dropped under 2 mph now and then. And a lower
recorded time would produce a higher average speed.

Now I have to go out to the garage, hang my bike in the
stand, zero things, and gently twirl the wheel at under 2
mph to see if it works on my cheap Nashbar computer.

Thanks,

Carl Fogel

Drat! Another nice theory that doesn't seem to work.

With my bike in the stand, I zeroed my Nashbar cyclocomputer
and began spinning the wheel gently, trying to keep it under
2 mph. For about 0.20 miles, I think that I kept it under 2
mph, 0.9 mph and 1.x mph (and sometimes 0.0 mph, which meant
that I had to give the tire a slap to wake things up).
Occasionally it drifted up to the 2.x range.

At 0.20 miles, I spun it up to 8 to 10 mph and watched until
the odometer turned over from 0.29 mph to 0.30 mph and
grabbed the wheel to stop the experiment.

The cyclocomputer showed 10.9 mph max, 0.30 miles, a time of
5:35, and a 3.2 mph average, suggesting that I had indeed
kept it under 2 mph for the first 0.20 miles.

The computer's average of 3.2 mph is pretty close to the
3.224 mph average that a time of 5:35 and 0.30 miles
produces. It's nowhere near the 5.0 and 5.4 mph discrepancy
reported in the Cateye. Since I had a number of failures
(0.0 mph, time to crank the tire a bit faster), there should
have been at least a few seconds lost to any hidden timer
that went dead at ultra-low speeds.

So either there's no hidden timer for averages that goes
dead below 2 mph, or else it's a questionable feature of
more expensive cyclocomputers.

Carl Fogel

Dear Carl
Usually when a cyclometer has the separate timer for the aver. Spd.,
It will say "Auto Start - Stop" in the instructions or on the box or
in ads.
This is really a great feature, if one is trying to get an accurate
idea of energy expenditure.
Let's say hypothetically that a rider is riding on a boulevard 10
miles long & gets stopped every 1/2 mi. by a traffic light, for 3 min
;-)(well that's how long it seems anyway) Now lets say he rides 20mph
btn. lights. W/o any lights it would take him 1/2 hr to ride that
distance. W/ lights he will spend 60 min sitting at 20 lights. Add the
2 together & he has spent 90 minutes to ride 10 miles for a 6.66 mph
average speed. His energy expenditure will be much greater that 6.66
mph would indicate. Discounting the time stopped gives a far more
accurate picture of training effort &/ or calories burned. Granted it
is a little optimistic, but not nearly as much as the 6.66 speed is
low. He can't say "I just rode for 90 min. @ an average speed of
20mph" But he can say I just rode 10 miles @ 20 mph & be pretty close
to the truth.
Compendia? John

[email hidden] wrote in message news:<[email hidden]>...

Quoted message said:

On Tue, 28 Sep 2004 20:04:06 -0600, [email hidden]

Quoted message said:

On 28 Sep 2004 18:42:38 -0700, [email hidden] (John
McGraw) wrote:

[snip]

Quoted message said:

I thought that the OP was probably aware that, now a days, most
cyclometers stop recording time for average speed when the speed drops
below a set speed. Usually ~2mph. Did the OP stop for a few seconds
during his ride? that would account for it, if the time on the ride
timer doesn't stop, & the average speed does.
Best, John

Dear John,

Hmmm . . . this is the first suggestion that makes sense.

So the idea is that there's a hidden timer just for average
speed that's separate from the visible elapsed-time? And
this timer goes to sleep when the speed drops to under 2
mph?

Given the 5 mph average, it seems plausible that the rider
might have dropped under 2 mph now and then. And a lower
recorded time would produce a higher average speed.

Now I have to go out to the garage, hang my bike in the
stand, zero things, and gently twirl the wheel at under 2
mph to see if it works on my cheap Nashbar computer.

Thanks,

Carl Fogel

Drat! Another nice theory that doesn't seem to work.

With my bike in the stand, I zeroed my Nashbar cyclocomputer
and began spinning the wheel gently, trying to keep it under
2 mph. For about 0.20 miles, I think that I kept it under 2
mph, 0.9 mph and 1.x mph (and sometimes 0.0 mph, which meant
that I had to give the tire a slap to wake things up).
Occasionally it drifted up to the 2.x range.

At 0.20 miles, I spun it up to 8 to 10 mph and watched until
the odometer turned over from 0.29 mph to 0.30 mph and
grabbed the wheel to stop the experiment.

The cyclocomputer showed 10.9 mph max, 0.30 miles, a time of
5:35, and a 3.2 mph average, suggesting that I had indeed
kept it under 2 mph for the first 0.20 miles.

The computer's average of 3.2 mph is pretty close to the
3.224 mph average that a time of 5:35 and 0.30 miles
produces. It's nowhere near the 5.0 and 5.4 mph discrepancy
reported in the Cateye. Since I had a number of failures
(0.0 mph, time to crank the tire a bit faster), there should
have been at least a few seconds lost to any hidden timer
that went dead at ultra-low speeds.

So either there's no hidden timer for averages that goes
dead below 2 mph, or else it's a questionable feature of
more expensive cyclocomputers.

Carl Fogel

On 30 Sep 2004 01:54:27 -0700, [email hidden] (John

McGraw) said:

Dear Carl
Usually when a cyclometer has the separate timer for the aver. Spd.,
It will say "Auto Start - Stop" in the instructions or on the box or
in ads.
This is really a great feature, if one is trying to get an accurate
idea of energy expenditure.
Let's say hypothetically that a rider is riding on a boulevard 10
miles long & gets stopped every 1/2 mi. by a traffic light, for 3 min
;-)(well that's how long it seems anyway) Now lets say he rides 20mph
btn. lights. W/o any lights it would take him 1/2 hr to ride that
distance. W/ lights he will spend 60 min sitting at 20 lights. Add the
2 together & he has spent 90 minutes to ride 10 miles for a 6.66 mph
average speed. His energy expenditure will be much greater that 6.66
mph would indicate. Discounting the time stopped gives a far more
accurate picture of training effort &/ or calories burned. Granted it
is a little optimistic, but not nearly as much as the 6.66 speed is
low. He can't say "I just rode for 90 min. @ an average speed of
20mph" But he can say I just rode 10 miles @ 20 mph & be pretty close
to the truth.
Compendia? John

Dear John,

I think that you may be misunderstanding things.

Yes, speedometers have an auto start-stop feature to avoid
adding the time spent at stop-lights to the ride.

This simply stops the displayed timer whenever the sensor
records no magnet passing in about three seconds and then
re-starts on the first magnet pass, whenever it occurs.

It happily times things down to 0.9 mph, far less than any
normal pedalling speed.

And this elapsed time is displayed. That is, there are not
two elapsed times, one for total time since the ride began
regardless of any pauses and another for only the time spent
moving. There's just the timer for the time spent moving.

This is why you can park your bike overnight, forget to
re-zero things, and get an accurate average speed and
elapsed time anyway. The timer and its average function work
only when the magnet appears to be turning. I sometimes
forget to set my speedometer for my daily 15 mile ride and
end up showing 30.40 miles, a time around 1:35, and the
familiar 19.x mph average.

The original poster was interested in why his Cateye was
showing an average of 5.4 mph for 5.0 miles in 60 minutes.
The 60 minutes is the elapsed time that the speedometer
spent noticing movement for 5.0 miles.

So far, no one has come up with a plausible explanation that
flatters Cateye.

Elsewhere in this thread, it's been mentioned that almost
ten years ago people were posting complaints about their
Cateye speedometers doing strange things.

The latinate word that you were probably searching for was
probably "comprende"--the "r" is crucial. ("Compendia" is a
latinate plural that means a collection of summaries, each
being a compendium.)

Carl Fogel

our down to the beach 12 mile run from town center is used for
sumkinda foot race and is mile marked-in case one expires, the
location is knowable.
i tracked my enduro-no slippage-and LO! the enduro was 250-300 feet
off at 12 miles. BUMMER!
then one day i found a second mark for this previously aforementioned
pavement mark and on rechecking found this second mark was right on
the button with the enduro.
reason? the guys doing the first mark screwed it up!
the second mark agreeable with the enduro is the correct mark.