-
Categories
-
Platforms
-
Content
Measuring the speed of a Frisbee & Boomerang
- Thread starter LazyBoy
- Start date
- Joined
- Jan 21, 2010
- Messages
- 25,510
imagine an accelerometer placed tangentially right on the edge of a car tyre.
it no longer traces out a circle, but a catenary. It's velocity drops to zero when it is on the roads surface and is the double of the car's velocity at the top of the wheel.
as long as its not at the centre of the wheel it will trace a curve that results in a similar effect which does not vary in frequency.
if a frizzbee is flying in a line, the same effect will be seen.
however I also like the idea of detecting any wobble in the disc. The issue with this is that a well thrown frizzbee will have little wobble. The other issue is that the wobble does not necessarily occur at the rotational frequency (consider our own planet as an example).
it no longer traces out a circle, but a catenary. It's velocity drops to zero when it is on the roads surface and is the double of the car's velocity at the top of the wheel.
as long as its not at the centre of the wheel it will trace a curve that results in a similar effect which does not vary in frequency.
if a frizzbee is flying in a line, the same effect will be seen.
however I also like the idea of detecting any wobble in the disc. The issue with this is that a well thrown frizzbee will have little wobble. The other issue is that the wobble does not necessarily occur at the rotational frequency (consider our own planet as an example).
The velocity is changing as you say, but I believe the acceleration is still toward the center of the of wheel. Think of the rate of change in the tangential velocity. It is zero when at the top or bottom of your catenary curve and highest when at the midpoints between them. This is consistent with the acceleration always being toward the center of the wheel.
edit: Relativity says that you cannot tell any difference between a frame that is at rest and on that is moving at a constant velocity. So the linear motion of the car or the Frisbee cannot possibly be distinguished from a car wheel spinning when the car is up on a jack stand or a Frisbee that is rotating in place without moving linearly, and your description seems to require and detect that difference.
Bob
edit: Relativity says that you cannot tell any difference between a frame that is at rest and on that is moving at a constant velocity. So the linear motion of the car or the Frisbee cannot possibly be distinguished from a car wheel spinning when the car is up on a jack stand or a Frisbee that is rotating in place without moving linearly, and your description seems to require and detect that difference.
Bob
OK, LazyBoy, use the method that you originally intended. Have the axis of the accelerometer that is being measured at a right angle to the center. Use the equation :
f = sqrt(a) / (sqrt(r) * 2 * pi)
a is the acceleration in metres per second per second, eg, this is like the force of gravity, earth gravity exerts 9.8 metres per second per second.
r is the distance from your accelerometer to the center of the Frisbee.
pi is the constant pi
f is your rotations per second and is the value we are finding.
The equation assumes that there is no other acceleration than the rotation, this is not entirely true as the Frisbee is accelerated when thrown and glides, slowly decelerating. For the purpose of this test, ignore this, if you want a more accurate reading of rotational speed, take the mean of the acceleration, this will however increase the time between readings.
If there is anything that needs clearing up, just ask. If you need any help with C or C ++ if you want to make the calculations autonomous, just ask and I can do it.
f = sqrt(a) / (sqrt(r) * 2 * pi)
a is the acceleration in metres per second per second, eg, this is like the force of gravity, earth gravity exerts 9.8 metres per second per second.
r is the distance from your accelerometer to the center of the Frisbee.
pi is the constant pi
f is your rotations per second and is the value we are finding.
The equation assumes that there is no other acceleration than the rotation, this is not entirely true as the Frisbee is accelerated when thrown and glides, slowly decelerating. For the purpose of this test, ignore this, if you want a more accurate reading of rotational speed, take the mean of the acceleration, this will however increase the time between readings.
If there is anything that needs clearing up, just ask. If you need any help with C or C ++ if you want to make the calculations autonomous, just ask and I can do it.
Thank you Guy's, you have been so helpful.
I am going to get a price for a prototype made in China, as I know an electronics factory that helped me develop a couple of other fun little products one being some L.E.D Children's Swimming Goggles around 10 years ago.
Does anyone know the basic components that I will be needing on the PCB
1 x Round PCB
1 x Button cell battery cradle
1 x 3v Button cell battery
1 x 4.0 LE Bluetooth module
1 x Accelerometer
?????????????????????????
I also know a guy who can build me a mobile device app that will display RPM and maybe even total revolutions...oh...or how about a scoring system, so players can compete worldwide for best speed etc...
I am going to get a price for a prototype made in China, as I know an electronics factory that helped me develop a couple of other fun little products one being some L.E.D Children's Swimming Goggles around 10 years ago.
Does anyone know the basic components that I will be needing on the PCB
1 x Round PCB
1 x Button cell battery cradle
1 x 3v Button cell battery
1 x 4.0 LE Bluetooth module
1 x Accelerometer
?????????????????????????
I also know a guy who can build me a mobile device app that will display RPM and maybe even total revolutions...oh...or how about a scoring system, so players can compete worldwide for best speed etc...
Attachments
You will need to pass the acceleration through the formula, you have three options.
1. The formula is substituted and calculated receiver side, this is if you prefer Java or whatever *TSK* high level language the receiver application *shudders* is "programmed" in.
2. The formula is substituted and calculated transmitter side, this is if you prefer C/++ or assembly. *Pick me!* But ideally, use C if going for this option, assembly may take everybody a while...
3. You receive the raw data, take this and plug it into the formula on paper.
Option 2 is my favourite, and you are going to need to do ADC conversion to get transmittable binary data. Are you familiar with atmel? This is the family of voice for such a task. Have you done ISP and C (not arduino) programming on an atmel chip before? If so, write your C and ISP it on there. If not, use an arduino chip.
Make a decision about this, I will probably be on tommorow.
1. The formula is substituted and calculated receiver side, this is if you prefer Java or whatever *TSK* high level language the receiver application *shudders* is "programmed" in.
2. The formula is substituted and calculated transmitter side, this is if you prefer C/++ or assembly. *Pick me!* But ideally, use C if going for this option, assembly may take everybody a while...
3. You receive the raw data, take this and plug it into the formula on paper.
Option 2 is my favourite, and you are going to need to do ADC conversion to get transmittable binary data. Are you familiar with atmel? This is the family of voice for such a task. Have you done ISP and C (not arduino) programming on an atmel chip before? If so, write your C and ISP it on there. If not, use an arduino chip.
Make a decision about this, I will probably be on tommorow.
Hi Guys,
Up-date
I have been in talks with the electronics factory in China.
They want to charge me USD$3,500 (approx GBP£2,450) for the pcb design, software programming and 3 samples
This is the spec:
25mm diameter pcb
1 x 3v button cell (user replaceable)
1 x 4.0 LE Bluetooth module
1 x Accelerometer
1 x Gyroscope
They commented that the battery life is expected to last no more than 10 hours.
Have you guys got any advice for me?
Up-date
I have been in talks with the electronics factory in China.
They want to charge me USD$3,500 (approx GBP£2,450) for the pcb design, software programming and 3 samples
This is the spec:
25mm diameter pcb
1 x 3v button cell (user replaceable)
1 x 4.0 LE Bluetooth module
1 x Accelerometer
1 x Gyroscope
They commented that the battery life is expected to last no more than 10 hours.
Have you guys got any advice for me?
Hi Bob,
Yes I agree however, I think it should be a fair trade off for the feedback...
I have seen Frisbee's with whistles and l.e.d lights attached etc...so should be workable.
I have asked them to position the battery exactly central on the PCB to help with balance.
- Joined
- Jan 21, 2010
- Messages
- 25,510
I have no idea unfortunately. What you could do is look up any local hackerspaces or makerspaces and see if any of them can give you any suggestions. It's even possible that they may be capable of producing a prototype for you.
A lower risk (and thus lower cost) method may be to get someone to build a prototype from available modules. This would allow software development and finally you could take the next step of getting someone to reproduce it on a single board with the other features (size, weight and balance) that you require.
Another advantage of a cheap prototype is that you could place it on a rotating wheel and check that it can respond fast enough to work at the range of speeds you anticipate.
Something like this: https://www.kickstarter.com/projects/guardyen/tiny-bluetooth-sensors-so-you-can-make-and-connect
$55 + P&P
$55 + P&P
- Joined
- Jan 21, 2010
- Messages
- 25,510
I was thinking even cheaper. Something like these and an arduino pro mini could form the heart of a proof of design and software development platform.
Look at Ragworm, they do extremely reasonable prices for small batches. Try them for the prototype. If you are going to be selling these, let me know if and where I can find them!
Similar threads
