SparkFun Forums

[kenny]

Hi,

I am a wheelchair racer doing 100m, 200m and 400m. One thing is that I am looking for a device that can measure/log my speed/acceleration over short durations very accurately to analyse my pushes.

As I said I want to log my speed/acceleration, so that I can see my speed over time at least accurate on the 0,01s. This way I could make a graph and look at my speed against an other trainings exercise. My starting speed is of course 0km/h and accelerate in 50m to about 30km/h. My most important question is, is it possible to get the system that accurate it could calculate the time it took over 100m in lets say about 14 seconds, accurate to 0,01s? This way I could analyse my strokes on the pushrim by force measurement.

Is a system like this possible? Or are there perhaps any complete designs available with this capability already?

Even more exiting would be if there would be a system to record this information “liveâ€

[daemondust]

Assuming your wheels don't slip, I woud do it by putting a rotary encoder (could be as simple as a LED and photo transistor blocked by the spokes of the wheel) to get your speed. Since acceleration is the derivative of speed, you can use calculus to get that.

With the photo transistor method you would take the circumference of your wheel, divide by the number of spokes, and measure the time difference between interruptions.

[kenny]

I have given that a thought as well. Is there perhaps a complete system capable of doing something like this? Though I could get this principle working, but my wheels sometimes do come from the ground, so I'm not sure how much that would affect the results. Maybe a combination of the photo sensor and an accelerometer would be precise?

Is it difficult to create a system with a 1 axis or 3-axis accelerometer sending output to a bluetooth module (for example the one in the bluetooth primer BR-SC11A) and recieved by a pda via a serial link and displays/logs the information to the screen and memory.

[upand_at_them]

What about the Garmin Forerunner?

Mike

[kenny]

What about the Garmin Forerunner?

Mike

That device uses a gps sensor which isn't accurate enough.

I am more and more looking at a device that measures the speed of the wheel. This could be a photosensor scanning markings on the axle hub. What kind of photosensor could I be using? A marker would cross the sensor about 1000 times a second, so I think that would make a sensor capable of at least 2000 checks a second.

[daemondust]

2kHz should be no problem. Just about any reflective sensor would work. I've used Fairchild's QRD1114in the past.

Most microcontrollers can have interrupts triggered when a pin changes state, so you should be able to use that and a timer to get pretty good measurements.

[weingart]

[quote="kenny"]Hi,

I am a wheelchair racer doing 100m, 200m and 400m. One thing is that I am looking for a device that can measure/log my speed/acceleration over short durations very accurately to analyse my pushes.

As I said I want to log my speed/acceleration, so that I can see my speed over time at least accurate on the 0,01s. This way I could make a graph and look at my speed against an other trainings exercise. My starting speed is of course 0km/h and accelerate in 50m to about 30km/h. My most important question is, is it possible to get the system that accurate it could calculate the time it took over 100m in lets say about 14 seconds, accurate to 0,01s? This way I could analyse my strokes on the pushrim by force measurement.

Is a system like this possible? Or are there perhaps any complete designs available with this capability already?

Even more exiting would be if there would be a system to record this information “liveâ€

[kenny]

I have thought of an accelerometer like I said in a previous post. I got a lot of feedback that an accelerometer is too noisy and as soon as the testing time exceeds 2 seconds it drifts to much. Though I hope that a local company specialized in high performance accelerometers will let me test their device so I could see if an accelerometer would suffice, because that would give additional information.

To give you an idea of my racing chair l added a picture: http://www.kennyvanweeghel.nl/images/wheeler.jpg

I think for now the easiest and accurate option would be a rotary encoder / photo transistor. I thought of using a mechanical mouse for measuring the speed of the wheel. Now I got some additional questions:

1. Do you think this is an option?
2. How about timing?
3. How should I make the connection to the racing chair? The racing wheel makes about 6 revolutions per second and if the wheel of the mouse is directly connected it would make about 1100 revs/sec. I counted the openings in the rotary wheel and it had 40 holes, so that means I think it should be able to scan 43200 holes a second. Anybody got an idea? Probably I need to use some gears, but how could I implement that in a small package?

I also found some device that might be able to help with logging, but I'm not completely sure if and how I could use it. It has some remarks that I think might be interesting to use for my purpose. It is the MELD Data Acquisition System. Any experience with that device? I still need to ask the creator about the info, but any information is welcome.

Lots of questions, but any info could be helpfull.

[Caffeine]

Another solution would be an inductive sensor mounted close to the spokes of the wheel. These sensors cost about $80, and output a high level every time a metal object comes into range (usually about 1-2cm)

With what looks like 16 to 18 spokes a wheel, the measurement shouldn't be too hard for a simple microcontroller. The resolution would be +/- 1/16 the wheel circumference, which should be acceptable.

A lot of bike computers have the functionality you're asking for.

[Fuss]

Hi,

I am a wheelchair racer doing 100m, 200m and 400m. One thing is that I am looking for a device that can measure/log my speed/acceleration over short durations very accurately to analyse my pushes.


TIA,
Kenny
[/img]

First Kenny - way to go, I think wheelchair racers are inspirational... I'm a middle distance runner (800/1500m) & always admire the guys (and gals) out there in wheelchairs.

Anyway, regarding your question, I'm an electronics guy & a drag racer.... I think this product family is exactly what you want:
http://www.gtechpro.com/ss.html
http://www.gtechpro.com/rr.html

but I'm not sure if you want plans for something you can build, or a product that provides the solution. These are commercial products, not kits.

I've used the Gtechs on the track and they're very accurate. The RR logs acceleration up to 20 Hz and honestly that is more than enough sampling even for a high-powered drag car...

[Furryice]

Hi Kenny,

Take a look at www.blacklinegps.com

Their GPS system is accurate to 0.2 KPH or 0.12 Mph, and 0.01 second time accuracy. They utilize Bluetooth so you could log your files to a PDA and then observe the respective data after the session.

The application might be perfect for you. You could also log your training session around the track and later compare all the data to see how your training has evolved.

Might be a interesting application!

Nixon.