Velocomputer

Post Reply
Pete
Posts: 110
Joined: Wed Mar 05, 2014 8:09 am
Location: Melburn, AUS

Velocomputer

Post by Pete »

Had to look twice at the name :)

http://velocomputer.com/views/index

Image
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: Velocomputer

Post by Velocomp »

These guys keep trying to figure out how to get people interested in their product.

They say a standard speed sensor measures distance to a precision of 7 feet. That's true in a slippery marketing sense: at any given point of time, with a conventional bike computer sensor your spoke magnet has to pass by the speed sensor in order for the distance reported to increment by 7 feet. So it is theoretically possible for the total distance reported by your bike computer to be "wrong" by 7 feet.

Assuming that the "error" happens randomly, in the real world the total distance reported on a normal bike computer has an average precision of 3.5 feet.

But the missing link is the distance over which you assess the precision. Here's where marketing obscures reality.

For example, on a 20 mile ride, your standard bike computer distance measurement might be "wrong" by 0.003%. Assuming their claim of 100 x more precision is correct, they will measure distance with an error of 0.00003%.

Of course, NO bike computer of which I'm aware reports distance traveled in feet (except maybe theirs...). Does anyone want to look at their bike computer and, for distance traveled, and at the end of their 20 mile ride see on their screen they rode 105600 feet, (+/- 3.5 ft or +/- .035 ft)?

Yawn.

On a hard, 10 second sprint, where you might be traveling at a rate of 30 mph, during the sprint you'll move down the road about 440 feet. The standard bike computer distance measurement might be "wrong" by 0.7%. Their device: 0.007%.

Yawn.

Oh, by the way: is your 700 x 23c wheel really 2096 mm in circumference? If you REALLY want extreme precision you'll need to account for the actual tire you're using, its inflation pressure, and the wear and tear on the tire as its rubber is consumed.

For example: if your actual wheel circumference is wrong by 1 mm (2097 instead of 2096), on a 20 mile ride your distance error caused by wheel circumference error will be comparable to that caused by the 3.5 foot precision metric.

This is much ado about nothing...
John Hamann
Pete
Posts: 110
Joined: Wed Mar 05, 2014 8:09 am
Location: Melburn, AUS

Re: Velocomputer

Post by Pete »

I was mostly interested in the magnet-less aspect because that seems to be a common issue for some folks here and I've had some initial trouble with that too.
Anyway I see Garmin now have something out and the review by DC Rainmaker mentions others on the way.
http://www.dcrainmaker.com/2014/05/garm ... nsors.html
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: Velocomputer

Post by Velocomp »

More toys for boys :-)
John Hamann
Pete
Posts: 110
Joined: Wed Mar 05, 2014 8:09 am
Location: Melburn, AUS

Re: Velocomputer

Post by Pete »

Was just thinking this morning that we soon have track season coming (in AU) and I'll be planning some drills to work on TT efforts from standing starts.
It will be interesting to track progress with the Newton and other software.

What is the precise trigger for Newton to start calcutating?
Is it when the speed magnet passes the sensor or the cadence or do you need both to have been detected?

I'm interested to know what is the earliest point from a standing start that power might be calculated, assuming the cranks and/or wheel magnets are aligned to optimize it. Specifically I'm interested to be able to pick up the maximal readings you expect to see from a standing start.

It is generally recommended to start left foot forward but right foot forward is viable.

There is of course some delay in what we see on the Newton screen due to filter/averaging,

NB I used this thread because I was thinking some means like this to detect 'activity' sooner might be useful in this instance.
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: Velocomputer

Post by Velocomp »

Pete wrote:Was just thinking this morning that we soon have track season coming (in AU) and I'll be planning some drills to work on TT efforts from standing starts.
It will be interesting to track progress with the Newton and other software.

What is the precise trigger for Newton to start calcutating?
Is it when the speed magnet passes the sensor or the cadence or do you need both to have been detected?

I'm interested to know what is the earliest point from a standing start that power might be calculated, assuming the cranks and/or wheel magnets are aligned to optimize it. Specifically I'm interested to be able to pick up the maximal readings you expect to see from a standing start.

It is generally recommended to start left foot forward but right foot forward is viable.

There is of course some delay in what we see on the Newton screen due to filter/averaging,

NB I used this thread because I was thinking some means like this to detect 'activity' sooner might be useful in this instance.
In physics power is defined as force x speed.

Naturally, to do this calculation you've got to start the calculations at some defined point.

In the Newton (and in all other devices for that matter) the ride file begins when motion or cadence is detected.

So, if you're starting from a dead stop, in a new ride file, recording won't begin until motion/cadence is detected. That could be after you've already taken off...

If you want to beat the system, there is a simple workaround: ride a bit so that ride file recording is started, then stop, then take off within a few seconds. The Newton will capture your first sprint.

We could make a special version of FW that would allow the user to start the recording manually, but so far there has not been sufficient demand.
John Hamann
Pete
Posts: 110
Joined: Wed Mar 05, 2014 8:09 am
Location: Melburn, AUS

Re: Velocomputer

Post by Pete »

Velocomp wrote:In physics power is defined as force x speed.

Naturally, to do this calculation you've got to start the calculations at some defined point.

In the Newton (and in all other devices for that matter) the ride file begins when motion or cadence is detected.

So, if you're starting from a dead stop, in a new ride file, recording won't begin until motion/cadence is detected. That could be after you've already taken off...

If you want to beat the system, there is a simple workaround: ride a bit so that ride file recording is started, then stop, then take off within a few seconds. The Newton will capture your first sprint.

We could make a special version of FW that would allow the user to start the recording manually, but so far there has not been sufficient demand.
Thanks John
We might be OK then if I'm understanding correctly that a motion detector (accelerometer?) has a role.
You'd usually roll down from the fence to a holder then get a "3-2-1 Go".
Just so long as we didn't need to wait for the 2nd pass of the cadence magnet which would take ~6.0m.
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: Velocomputer

Post by Velocomp »

Pete wrote:
Velocomp wrote:In physics power is defined as force x speed.

Naturally, to do this calculation you've got to start the calculations at some defined point.

In the Newton (and in all other devices for that matter) the ride file begins when motion or cadence is detected.

So, if you're starting from a dead stop, in a new ride file, recording won't begin until motion/cadence is detected. That could be after you've already taken off...

If you want to beat the system, there is a simple workaround: ride a bit so that ride file recording is started, then stop, then take off within a few seconds. The Newton will capture your first sprint.

We could make a special version of FW that would allow the user to start the recording manually, but so far there has not been sufficient demand.
Thanks John
We might be OK then if I'm understanding correctly that a motion detector (accelerometer?) has a role.
You'd usually roll down from the fence to a holder then get a "3-2-1 Go".
Just so long as we didn't need to wait for the 2nd pass of the cadence magnet which would take ~6.0m.
Yes, the accelerometer has the dominant role in these sprints. You might not see instantaneous response on your screen, but the ride file will record it all.
John Hamann
Post Reply