“Tiny Orbo replication” 170% efficient
[2010/2/11 update: See "Tiny orbo replication 2" major update]
Today I took more detailed measurements to obtain the efficiency of my “tiny Orbo replication.” Once again it was well over 100% efficient. My “tiny Orbo replication” has changed since the last measurements. The toroid was rotated 180 degrees around so that there are no wires directly facing the magnets. Also, the voltage was low, just 0.36 volts. Furthermore, I moved the toroid closer the magnets, so the inductance is different. The following math and results excludes joule heating from electrical wire resistance, as the goal is to see if there is any excess energy:
General info
Revolutions per second: 26.5 rps (1590 rpm)
Pulses per second: 53.0 pulses/s
Peak voltage pulse across coil: 0.36 V
Peak current pulse through coil: 1.26 A
Inductance of coil during pulse: 60 uH
Power input into inductance: 0.5 * 60 uH * 1.262 A * 53.0 pulses/s = 2.52 mW. Note, inductance was linear in this range. [2010/2/11 update: Nearly all of this inductance energy can be captured back. See "Tiny orbo replication 2" major update]
Deceleration control test:
Time duration: 0.33 s
Starting rps: 27.6 rps
Ending rps: 24.7 rps
Deceleration rate: (27.6 rps – 24.7 rps) / 0.33 s = 8.79 rps/s
Calculated rps from 26.5 rps in 0.1 s = (26.5 rps – 8.79 rps/s * 0.1 s) = 25.6 rps
NdFeB magnets
Quantity: 2
Density: 7400 kg/m3
Length: 0.0033 m
Radius: 0.0028 m
Total volume: [(pi * 0.00282 m) * 0.0033 m] * 2 = 1.63e-7 m3
Total mass: 7400 kg/m3 * [(pi * 0.00282 m) * 0.0033 m] * 2 = 1.20e-3 kg
Average distance from center of rotation: 0.016 m
Average velocity @ 26.5 rps: (2 * pi * 0.016 m) * 26.5 Hz = 2.66 m/s
Kinetic Energy @ 26.5 Hz: 0.5 * 1.20e-3 kg * 2.662 m/s = 4.25 mJ
Average velocity @ 25.6 Hz: (2 * pi * 0.016 m) * 25.6 Hz = 2.57 m/s
Kinetic Energy @ 25.6 Hz: 0.5 * 1.2e-3 kg * 2.572 m/s = 3.96 mJ
Deceleration energy loss in 0.1 seconds: 4.25 mJ – 3.96 mJ = 0.290 mJ
Power output from friction @ 26.5 rps: 0.290 mJ / 0.100 s = 2.90 mW
Black plastic outer rim
Quantity: 1
Density: 1000 kg/m3
Height: 0.00635 m
Thickness: 0.000889 m
Distance from center of rotation: 0.0146 m
Volume: (2 * pi * 0.0146 m) * 0.00635 m * 0.000889 m = 5.18e-7 m
Mass: 1000 kg/m3 * [(2 * pi * 0.0146 m) * 0.00635 m * 0.000889 m] = 0.000518 kg
Velocity @ 26.5 Hz: (2 * pi * 0.0146 m) * 26.5 Hz = 2.43 m/s
Kinetic energy at 26.5 Hz: 0.5 * 0.000518 kg * 2.432 m/s = 1.53 mJ
Velocity@ 25.6 Hz: (2 * pi * 0.0146 m) * 26.5 Hz = 2.35 m/s
Kinetic energy at 25.6 Hz: 0.5 * 0.000518 kg * 2.352 m/s = 1.43 mJ
Deceleration energy loss in 0.1 seconds: 1.53 mJ – 1.43 mJ = 0.100 mJ
Power output from friction @ 26.5 rps: 0.100 mJ / 0.100 s = 1.00 mW
Black plastic flat top disc
Quantity: 1
Density: 1000 kg/m3
Radius: 0.0146 m
Thickness: 0.000686 m
[extra info: Volume: 9.19e-7 m^3]
[extra info: Mass: 0.000919 kg]
Integrated Kinetic energy at 26.5 Hz: 0.679 mJ
Integrated Kinetic energy at 25.6 Hz: 0.633 mJ
Deceleration energy loss in 0.1 seconds: 0.679 mJ – 0.633 mJ = 0.0460 mJ
Power output from friction @ 26.5 rps: 0.0460 mJ / 0.100 s = 0.460 mW
Total friction heating from both NdFeB & black plastic outer rim: 2.90 mW + 1.00 mW + 0.460 mW = 4.36 mW
Efficiency: 4.36 mW / 2.52 mW = 173%
Friction heating was calculated by doing a control experiment. During the control experiment, the “tiny orbo replication” was rotating over 28 rps. The deceleration rate was then recorded after turning the “tiny orbo replication” off. This showed how fast the machine decelerated between 27.6 rps and 24.7 rps. The deceleration rate along with mass and velocity of the pieces provides the loss in kinetic energy over time, which provides the friction heating output of the device at the operating speed.
This the circuit I’m using.
I am not affiliated with Steorn. Orbo is trademarked by Steorn
2010, January 19 at 3:28 pm
Loading ...
2010, February 12 at 3:24 pmInterested EE
link
I have been following Steorn for a while now and stumbled across your circuit diagram. I’m an experienced EE and if you are interested, I could help to optimize your circuit design.
I believe that if you are able to reliable build an over-unity demonstrator then there will be a tremendous interest from hobbyists looking to replicate your work.
Please let me know if you would like some help.
–EE
2010, February 13 at 5:22 amEnergyMover
link | my site | author
Thanks for the offer. I sure could have used you a week ago! The circuit for “Tiny Orbo Replication 2″ is complete enough to do the job. It’s not the greatest design by any means. Instead of using a high-side MOSFET driver, I’m using an ultra high permeability magnetic core for a transformer to get an isolated ground – source to drive the dual back to back MOSFETs. The reason being is that I don’t have on hand a low power high-side MOSFET driver.
Anyhow, the reason for the new circuit is to see if a faster core charging pulse will increase the efficiency. The original design was ~ 60us to 100us, while the new design is 25us, but the circuit can go below 10us. If the faster charging pulse does not help, then the old design is better since it’s simpler. What is meant by “charging pulse” is the time it takes for the pulse current to reach near maximum.
2010, February 28 at 2:36 pmGary Clarke
link
Such lovely work
Are they neoperm cores ?
2010, February 28 at 2:45 pmEnergyMover
link | my site | author
Hi. They’re Metglas MAGAMP cores. Longitudinally annealed with a permeability of 1 million.