Ok Hi All,
Here is what I want to do and I think I have got my head around the components needed but I really need to know exactly what to order. I do know that buying the kit is probably the easiest part 
So I want to build Wireless Telemetry for Athletes. In principal an IMU encased into Track Spikes that will give me data that can be converted into useful information so I can produce stride frequency, stride length, speed, number of strides, reaction, foot angles, mapping of the feet and legs, and I should also be able to calculate knee lift etc with a few measurements and time data etc.
The other thing that I really really really need to add is a pressure / load sensor that will calculate downward power, duration of foot strike, force etc but that I am guessing will have to be an external sensor to the IMU (physically it will need to be mounted to the footplate of the track spikes)
My vision for this is huge but I need to make one.
I have seen a few products that are good starting blocks but they all look like made from products from Sparkfun anyway using an IMU/Xbee module etc and some sort of plastic casing with a slim like phone battery.
ALL the help will be MASSIVELY MASSIVELY helpful !!!
Thanks ever so much
Warren
You should read this thread …
https://forum.sparkfun.com/viewtopic.ph … mu%2C+foot
IIRC the last poster in that thread did complete his project and I saw more details on the WWW … somewhere. If your GoogleFu is moderate, you should be able to find it.
Yes I have seen that and the kit he uses is from X10 but they are not responding to any messages or emails on their site so not sure if they are still around.
http://www.x-io.co.uk/products/x-bimu/#!prettyPhoto you can see the components and he mentions sparkfun so I am gfuessing that it all comes from there ???
People maybe able to identify it from the pics.
HELP PLEASE
The IMU you linked to looks to be “better” than the 3 similar items SF sell, if only because it allows an XBee to be added more easily than do the 3 others. So as a starting point you might want to look there. See what other inputs it has for the other sensors you want to add. It might make a good development platform. And I would advise you to go that route. Do Not try to make your final product on the first go ! I got the impression from your OP that somehow all the components would be integrated into the shoe sole. IMO that’s going to require sophisticated assembly of components on a flex print type of PCB, not the standard glass boards. Frankly I think that’s out of most hobbyist’s capabilities. Moreover I don’t even know what components (integrated accel/gyro/magnetometer) you need. I don’t know the forces and rates and etc likely to be encountered. I don’t know what resolution(s) you need or what sampling rates. And that’s on purpose of a dev platform, to measure real-world values and experiment with algorithms before you commit to the hard to do and big $$$s of the “final” design. For all I know the integrated ICs like those used on …
https://www.sparkfun.com/products/12636 or …
https://www.sparkfun.com/products/11486
… aren’t the best suitable to do your task. You’ve not set out a set of requirements to choose components by. How much pressure/force are you expecting to see ? Would this be sufficient or not ? I don’t know.
https://www.sparkfun.com/products/9376
Just another example … I assume you’ll want to power this device from some battery. For how long ? The power and run-time requirements of the system will guide the choice of components. Since space will be very limited you’ll probably need to go with ICs that all run off the same voltage. That may exclude certain components that otherwise would be ideal.
Hi There
No it is literally going to be a module like the one shown as my vision is that athletes just “strap” it on and get telemetry rather than making it proprietary in some fashion unless someone like nike pays me millions of pounds pmsl. No this is going to be aimed at athletes from county level to national level. So yes the module shown is ideal its just I cant seem to get hold of them.
big wazza:
No it is literally going to be a module like the one shown as my vision is that athletes just “strap” it on and get telemetry rather …
OK, I got sidetracked reading this part …
In principal an IMU encased into Track Spikes that will give me data
So neither of the 2 IMU’s I’d consider from SF are optimal. The Razor doesn’t have any of it’s spare analog pins routed out to a “connector”. You’d have to tack wires onto the MCU pins itself.
https://www.sparkfun.com/products/10736
The DIYDrones IMU has spare analog lines but runs off 5V and so it won’t be easy to run off a single cell LiPo. You need 2 cells in series or booster converter and a single cell.
https://www.sparkfun.com/products/11055
Of course other people may have other similar products, I can’t say.
And then you could go a separates route, building your own IMU system from a single cell LiPo, MCU and sensors (IMU breakout and pressure sensor) and some RF link (XBee, Nordic or Bluetooth device). It won’t be as small as the system you linked too but it might not be that much worse.
https://www.sparkfun.com/products/11486
https://www.sparkfun.com/products/12636
https://www.sparkfun.com/products/11114
https://www.sparkfun.com/products/12587
https://www.sparkfun.com/products/8483
(not at all sure what the “good” suggestions might be for an RF link. Do you want the data to go to a PC or a smartphone ?)
In any case I would still advise building a development platform from something like the above list for a prototype. Then consider making your own custom PCB to incorporate the ICs needed in a smaller size PCB (or 2).
Thank you so much you are a legend
Hi All,
Ok I am slowly, very slowly moving forward and have a few quesions.
So I was looking for a pressure pad/sensor that would be external to the IMU and fit to the sole of the shoe being worn. This would measure contact time, force, etc but this has proven very difficult because a sprinter generates about 75kg of force from the blocks and about 15kg per stride and to find those ranges in a small lightweight pad is either impossible or impossibly expensive.
I then realised that a high end accelerometer with adjustable ranges can measure in m/sec or G force. So my question is this…can it measure different axis in different values so Y in gforce and X in m/sec ?
But that then gets ultra complex because I am guessing you can only set ranges for the accelerometer as a whole and not per axis or can you ?
If thats possible then using the Y axis data in G would I think enable me to calculate A) the force generated and B) duration of foot strike and C) Number of footstrikes ???
Then the X axis data in m/sec will give me directional speed ???
Or…do I need 2 accelerometers per IMU ???
HELP HELP HELP HELP
Thanks
Warren
First off … accelerometers measure acceleration and the units of acceleration are m/sec/sec or ft/sec/sec or “Gs”. 1 G = 9.806 m/sec/sec or 32.2 ft/sec/sec. Thus you can go from one to another w/simple math. You can’t go from m/sec, which is a measure of speed, to acceleration as easily. They aren’t equivalent things. And I don’t think you want to, at least for the discussion immediately above. I have no doubt that you could count footfalls/footstrikes (and time on the ground) using an accelerometer, after all that’s what happened in the link I posted waaay above.
As for getting force from the accelerometer … ideally that could be done as Mr Newton wasn’t joking when he said F=M*A. If you know the acceleration and mass, you could theoretically compute the force involved. Alas I’m not sure what the mass is and there’s also some "springs involved in the shoe cushioning and leg muscles.
Getting back to the pressure pads … I would urge you to look into how they work. Do they measure force or pressure ? If you put a 10 kg weight on a stick w/an area of 1 cm^2 and put that in the middle of a pad, what would it output ? What would it output if the stick had an area of 10 cm^2 ? Frankly I don’t know. But I do know the entire force of a footstrike isn’t concentrated in just 1 cm, it’s spread across the foot and it changes in it’s intensity and distribution across the foot from the time of first contact w/the ground until the foot is lifted. An accelerometer could never tell you that story.
As for getting speed from an accelerometer … that’s conceptually easy but difficult in real life due to real world “defects” in the accelerometers you can buy. Speed in the X direction is the integration of the X accelerations over time. If you accelerate at 1 m/sec/sec for 5 secs, how fast are you going in the X direction ? Your problem is (at least) twofold. All accelerometers have some amount of offset, that is they read out some non-zero amount of acceleration even when there is no acceleration. That amount keeps getting added/integrated over and over leading to a drift error in the speed estimate. Then, in your case, the X direction of the accelerometer changes as the leg moves. If it corresponds to directly ahead when standing, it won’t when running as the leg’s angle wrt the ground changes. And this leads to gravity causing drift as well as the accelerometer will measure the force due to gravity. Again all this was faced by the writer of the link I posted to waaay above. You might want to look at his algorithms.
Once more you are a legend Thank You
In particular read this article and it’s links.
http://www.x-io.co.uk/gait-tracking-with-x-imu/
A read of this paper, at least the parts re: drift errors, is also recommended.
http://www.x-io.co.uk/res/doc/madgwick_ … report.pdf