x x x x
x x x x x x x
x x x x x yyyyyyy
0000000000000000000000000
I have a curve that is shaped like a series of half a sine curve (the
acceleration of a bike for half a pedal stroke)
i need to find x - y without knowing y, or the average of the curve
without the offset. It seams easy enough, if it was a sine curve
=highpass(x)+root(2)*rms(highpass(x))
but the way its shaped has me stumped.
thankyou
almost rms
Started by ●April 12, 2007
Reply by ●April 12, 20072007-04-12
docherty@onlink.net wrote:> x x x x > x x x x x x x > x x x x x yyyyyyy > > 0000000000000000000000000 > > I have a curve that is shaped like a series of half a sine curve (the > acceleration of a bike for half a pedal stroke) > i need to find x - y without knowing y, or the average of the curve > without the offset. It seams easy enough, if it was a sine curve > =highpass(x)+root(2)*rms(highpass(x)) > but the way its shaped has me stumped.I'm stumped too. What do 'x' and 'y' represent to you? Offset of what from what? Proper ankle and calf use make the crank torque much more constant than would be deduced by hanging a weight from the pedal. Jerry P.S. The area under a sine of peak value 1 from zero to pi is 2/pi. -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Reply by ●April 12, 20072007-04-12
docherty@onlink.net wrote:> x x x x > x x x x x x x > x x x x x yyyyyyy > > 0000000000000000000000000 > > I have a curve that is shaped like a series of half a sine curve (the > acceleration of a bike for half a pedal stroke) > i need to find x - y without knowing y, or the average of the curve > without the offset. It seams easy enough, if it was a sine curve > =highpass(x)+root(2)*rms(highpass(x)) > but the way its shaped has me stumped. > thankyou >-- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Reply by ●April 12, 20072007-04-12
On Apr 12, 10:34 pm, Jerry Avins <j...@ieee.org> wrote:> doche...@onlink.net wrote: > > x x x x > > x x x x x x x > > x x x x x yyyyyyy > > > 0000000000000000000000000 > > > I have a curve that is shaped like a series of half a sine curve (the > > acceleration of a bike for half a pedal stroke) > > i need to find x - y without knowing y, or the average of the curve > > without the offset. It seams easy enough, if it was a sine curve > > =3Dhighpass(x)+root(2)*rms(highpass(x)) > > but the way its shaped has me stumped. > > I'm stumped too. What do 'x' and 'y' represent to you? Offset of what > from what? Proper ankle and calf use make the crank torque much more > constant than would be deduced by hanging a weight from the pedal. > > Jerry > > P.S. The area under a sine of peak value 1 from zero to pi is 2/pi. > -- > Engineering is the art of making what you want from things you can get. > =AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF==AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF= =AF=AF=AF=AF=AF=AF=AF=AF=AF=AFthanks jerry, but not the answer im looking f= or.maybe i should clairify (understatment.) in the doodle, y is the acceler= ation from gravity going down a hill x is the aditional acceleration from = pedaling it is shaped like a half a sine curve and 0 at the top of the peda= l stroke i want to remove y. im measuring y+x. all help will be apreciated
Reply by ●April 12, 20072007-04-12
docherty@onlink.net wrote:> On Apr 12, 10:34 pm, Jerry Avins <j...@ieee.org> wrote: >> doche...@onlink.net wrote: >>> x x x x >>> x x x x x x x >>> x x x x x yyyyyyy >>> 0000000000000000000000000 >>> I have a curve that is shaped like a series of half a sine curve (the >>> acceleration of a bike for half a pedal stroke) >>> i need to find x - y without knowing y, or the average of the curve >>> without the offset. It seams easy enough, if it was a sine curve >>> =highpass(x)+root(2)*rms(highpass(x)) >>> but the way its shaped has me stumped. >> I'm stumped too. What do 'x' and 'y' represent to you? Offset of what >> from what? Proper ankle and calf use make the crank torque much more >> constant than would be deduced by hanging a weight from the pedal.First: the acceleration is a half sinusoid only if force on the pedal is constant and doesn't change direction for the entire down stroke. When I ride, my dropped ankle near the top of the stroke coupled with a forward push below the knee starts the motive force before top dead center. Extending the ankle near the bottom of the stroke and pulling the leg backward continues to apply pressure to the pedal after bottom dead center. I suspect that a plot of torque vs. angle is flattened on top, and that I apply useful pressure for about 110 degrees. Moreover, I pull up on the rear foot to keep the downward pressure on the other from lifting me out of the saddle when I'm peddling hard. If a good numerical answer is important to you, you may need to correct your model. Second: if you want to calculate the acceleration due to peddling according to your model, superposition is applicable. If instead you want to calculate the torque from measured (or computed from measured velocity) acceleration, that's just a scale factor and a constant, torque and acceleration being proportional. The constant should be discernible from a few cycles of the action. Third: is this really a signal processing question? Jerry -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Reply by ●April 13, 20072007-04-13
Reply by ●April 13, 20072007-04-13
=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF thanks again but my questin is likely simpler.. on http://analyticcycling.com/PedalAcc_Page.html. I want to get from acceleration vs time(graph 2) to propultion forces vs time(graph4).
Reply by ●April 13, 20072007-04-13
docherty@onlink.net wrote:> ��������������� > thanks again but my questin is likely simpler.. on > http://analyticcycling.com/PedalAcc_Page.html. I want to get from > acceleration vs time(graph 2) to propultion forces vs time(graph4).Note "In this analysis we approximate a rider's pedaling force with a sinusoidal force tangent to the pedals. A comparison with real data suggests that this is a good approximation." I suggest that it is a poor approximation when the rider feels the need for exertion. I could ride up a short 45-degree slope in my low gear of 45" when I was in better shape. It's not possible to do that by standing on the pedals, which is what the approximation implies. When a dog takes after me, I pedal with the foot on the far side only, holding the foot on the dog's side still and raised, so as not to excite the beast. That usually works, but once I had to jam my heel down the dog's throat. The first graph puzzles me in the introduction page. The vertical axis is marked "[Distance - Initial Speed x time] (m)". Now, distance is speed integrated over time, so the dimension is correct, but the graph is puzzling. It is primarily negative and varying, but the initial speed is constant by definition and the cyclist is probably moving forward. What does the graph really represent? The first graph after the model is run is equally puzzling. Its caption is "Abs[Tangent Force/Acceleration] For 1 Pedal Stroke" and it shows four peaks. Tangential force/acceleration should be constant (and never change sign), so why "abs()"? What does the graph really represent? Whatever assumption is made about the force on the pedal, the force of the wheel on the ground is twice the crank length divided by the gear number (with a minor allowance for transmission loss).* On level ground, the acceleration is that force divided by the mass of bicycle, rider, and any gear. On a slope, add or subtract the sine of the angle. Correction for the angular momentum of the wheels will be lost in noise except perhaps on heavy-tired mountain bikes. You didn't answer my last question to you: why is this a topic for comp.dsp? Jerry ___________________________________________ * The scales of the last two graphs show that the diameter of the crank circle is 7/50ths of the effective wheel diameter (gear number). The gear number is .617 * 53/15, or 2.18 meters. Dividing by the crank circle diameter, 2.18/.34 gives 6.4, which is far off the 7.14 deduced from the plots. (Isn't 170 mm short for a crank? Mine are about 200, maybe 215.) -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Reply by ●April 13, 20072007-04-13
Clay wrote:>> P.S. The area under a sine of peak value 1 from zero to pi is 2/pi. > > How about just two!Yes. The average value -- what a DC meter with rectifier would read -- popped out of my head. I realized that right after I posted, but I can't see that post any more. so I couldn't correct myself. Thanks for the word. Jerry -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Reply by ●April 13, 20072007-04-13
On Apr 13, 9:39 am, Jerry Avins <j...@ieee.org> wrote:> On level ground, > the acceleration is that force divided by the mass of bicycle, rider, > and any gear. On a slope, add or subtract the sine of the angle. > Correction for the angular momentum of the wheels will be lost in noise > except perhaps on heavy-tired mountain bikes. > > You didn't answer my last question to you: why is this a topic for comp.dsp? > > JerryI don't have the angle of slope but it doesn't matter at the dead spot of a pedal stroke, i know the only acceleration will be from opposing forces. what i want to do is subtract that acceleration The obvious thing is to subtract the lowest value, but that's to easy. what I'm doing to do is use a dsp(actually a 8 bit pic but it wants to be a dsp when it grows up cause its brother is) to: low pass filter + add a % of the rms . What is that percentage. if it was a sine wave rather than a half sine wave the percent would be sqrt(2) but as it is a half sine wave it has me stumped, and apparently jerry too
