Posted: Mon Dec 21, 2009 11:49 am
In jacks design I would worry about how much current the battery is putting out when it is just sitting there. A half an amp (4.6V/45ohm) or so means the battery is probably getting warm and the voltage is drifting a bit. I would think a $0.5 voltage regulator would be a good idea. In addition the ~2.5 watts being dissipated in the resistor means it's resitance is probably drifting as well.
I would cut the current way down to reduce heating affects and include a voltage reg so the only real variable is the cap. Boost the resitor 100X then drop the cap 100X. (or 1000x on each)
Cap leakage woudn't be a huge problem since the voltage is low (less than 5V), besides, it is easy enough to measure the leakage, just charge the cap up with the battery, disconnect battery and monitor the voltage drop with a high impedance voltmeter. (DVM are typically 10 MOhm resitance, a 1000 UFD cap would take a couple hours to discharge through the meter.)
RP181
You might still be able to get the piezo setup to work. (I've fiddle with piezo's a fair amount)
I would add a mass to the other side of the piezo. As it is setup now the piezo is responding to sound and probably not responding all that much to the flexing of the sheet of plastic. In seismometers the sensing element is generaly sandwiched between the ground and a large mass. As the ground moves the mass tends to not move and that generates the pressure that the seismometer records. With one side of the piezo free floating I would thing the signal would be very weak. (Piezos aren't usually used in seismometers since earthquakes are usually very low frequency signals.)
Can you set the sheet of plastic on top of the piezo and then put the piezo on the ground? Or set it so the piezo is resting on a suitably diametered steel rod.
I suspect that the best way to decide if the setup is giving a signal related to KE or momentum is to just calibrate it. I would do it that way anyway, calibration is always better than simulation. Set it up and drop varying mass from varying heights on to it. Easy enough to calculate the energy and momentum of the falling weights. You can then try to generate a calibration curve for your setup. Doesn't even matter if it is linear, it just need to be a reasonably smooth curve and need to get close to the energy range you are interested in.
I think you'll have to do some experiementing on how to process the signal. Should you use the hieght of the first peak or the area under the curve? Should the damped sign wave be mathematically recitified then integrated?
If you want a more accurate height and area under the first peak then fit a parabola to the three data points. (You can also fit a curve if you have more data points.) That usually works surprising well for getting the peak time, peak height and peak area.
I would cut the current way down to reduce heating affects and include a voltage reg so the only real variable is the cap. Boost the resitor 100X then drop the cap 100X. (or 1000x on each)
Cap leakage woudn't be a huge problem since the voltage is low (less than 5V), besides, it is easy enough to measure the leakage, just charge the cap up with the battery, disconnect battery and monitor the voltage drop with a high impedance voltmeter. (DVM are typically 10 MOhm resitance, a 1000 UFD cap would take a couple hours to discharge through the meter.)
RP181
You might still be able to get the piezo setup to work. (I've fiddle with piezo's a fair amount)
I would add a mass to the other side of the piezo. As it is setup now the piezo is responding to sound and probably not responding all that much to the flexing of the sheet of plastic. In seismometers the sensing element is generaly sandwiched between the ground and a large mass. As the ground moves the mass tends to not move and that generates the pressure that the seismometer records. With one side of the piezo free floating I would thing the signal would be very weak. (Piezos aren't usually used in seismometers since earthquakes are usually very low frequency signals.)
Can you set the sheet of plastic on top of the piezo and then put the piezo on the ground? Or set it so the piezo is resting on a suitably diametered steel rod.
I suspect that the best way to decide if the setup is giving a signal related to KE or momentum is to just calibrate it. I would do it that way anyway, calibration is always better than simulation. Set it up and drop varying mass from varying heights on to it. Easy enough to calculate the energy and momentum of the falling weights. You can then try to generate a calibration curve for your setup. Doesn't even matter if it is linear, it just need to be a reasonably smooth curve and need to get close to the energy range you are interested in.
I think you'll have to do some experiementing on how to process the signal. Should you use the hieght of the first peak or the area under the curve? Should the damped sign wave be mathematically recitified then integrated?
If you want a more accurate height and area under the first peak then fit a parabola to the three data points. (You can also fit a curve if you have more data points.) That usually works surprising well for getting the peak time, peak height and peak area.