-
#2820 by oyzw on 07 Dec, 2017 00:44
-
I think it would be more appropriate to use a boat as a test vehicle in a static pool..... test whether vibrating devices can produce false-positive thrust results on a torsional pendulum. ...
Hello, your experiment is very rigorous, but I suggest you use the air suspension platform as an experimental vehicle as soon as possible.
O, no, he shouldn't.
The air suspension platforms are very unreliable instruments to work with. For critiques, see, e.g., Marc Millis, Nonviable mechanical “antigravity devices, in: M.G. Millis and E.W. Davis (eds.), Frontiers of propulsion science, AIAA, 2009, pp. 249–261.
Monomorphic started his experiment with an air suspension rail. There were reasons why it was not good. -
#2821 by otlski on 07 Dec, 2017 00:47
-
.... test whether vibrating devices can produce false-positive thrust results on a torsional pendulum. ...
Hello, your experiment is very rigorous, but I suggest you use the air suspension platform as an experimental vehicle as soon as possible.
O, no, he shouldn't. The air suspension platforms are very unreliable instruments to work with. For critiques, see, e.g., Marc Millis, Nonviable mechanical “antigravity devices, in: M.G. Millis and E.W. Davis (eds.), Frontiers of propulsion science, AIAA, 2009, pp. 249–261.
Be careful not to generalize too much on this subject. I will say that without a doubt a professionally manufactured, well characterized air bearing will out perform all but perhaps a perfectly executed torsion balance. And at that, I would have to think about it some more. The air bearing is much more expensive yes, but inferior, no. I say this having built hundreds of air bearings as part of our mass properties instruments, and having built hundreds of torsion instruments for measuring MOI. The torsion instruments measure to within 0.1% for parts as small as a few grams and as large as 24,000 lbs. The air bearings support similarly sized payloads. I have purchased perhaps fifty air bearings from three different competitors that perform similarly to ours.
I have reliably measured torques down to tens of millions of a lb-inch. Extracting said signal out of motoring torque, viscous drag, turbulent drag, etc. If my company were tasked with construction of an instrument to test EM Drives, our design kick-off meeting would settle on an air bearing as the heart of the instrument in two minutes. The only complicating factor would be the application in a vacuum chamber where vacuum level required and pump capacity, I might go to one of our competitors for solutions.
The one Mono made was a valiant attempt but a linear bearing would be my last choice and the execution was the best he was willing to do at the time. There is (both literally and performance-wise) a sizable gap between that attempt and one professionally made.
All that said, I would love to read the paper you cited if you had a link. Always looking to learn more... -
#2822 by flux_capacitor on 07 Dec, 2017 10:23
-
About problems with air bearings: do you remember Paul March's remarks back to thread 8, after Eagleworks has tested their cavity on a rotary low-friction Cavendish balance supported on a spherical air-bearing?
Emphasis mine:Quote from: Star-Drivethe EW team ran the same Integrated Copper Frustum Test Article (ICFTA) on a battery powered, spherical air-bearing supported, Cavendish-Balance (C-B) last summer, and it self-accelerated in both directions when the ICFTA was reversed on its mount. Past that I can't reveal anymore on the C-B test campaign until Dr. White gets around to publishing those test results after some improvements are made to the spherical air bearing, which had some annoying swirl torques that disturbed the data runs, but did not hide the already noted results.
TheTraveller calculated the required force to accelerate the platform to the observed speed is only 17 µN, so he admitted the results of these tests are very uncertain. -
#2823 by otlski on 07 Dec, 2017 23:53
-
What Paul March is calling swirl torque we call motoring a.k.a turbine effect. Motoring can be virtually nonexistent or problematic of course depending on magnitude. Well made air bearings can motor but the motoring will be repeatable and easily characterized (using two different methods).
in order for TT to calculate the torque required to achieve a particular angular acceleration rate, he would need to know the MOI of the test rig. I did not see that posted anywhere back then. Furthermore he would have to separate the retarding torque caused by viscous air drag and perhaps added mass. The bare minimum required would be one full revolution of the test rig. -
#2824 by xyzzy on 09 Dec, 2017 11:06
-
Monomorphic,
Looking at your vibrational thrust test graphs, I think there is a significant experimental error: You seem to have neither eliminated nor accounted for the changing offset / DC component in the movement of the plunger in your shaking mechanism. The reason is that you are using a unipolar signal source for the excitation and that the source turns off (rather than maintaining the average bias current through the solenoid) between the test pulses. This change in biasing conditions and the corresponding mass shift in the shaker mechanism results in a mass shift of the entire torsional balance, which, together with the long time constant of the balance, makes it look like a thrust signal.
To perform such an experiment "cleanly" it would need 2 signal sources for the shaker:
1. A fixed, DC, current (not voltage) regulated, constant current bias supply that is always on for the duration of the experiment.
2. A AC-only signal source with no DC component in the output (preferably also a current source, but voltage would still be OK in a pinch) for the shaking signal.
Both sources need to drive the shaker in parallel but to prevent signal flows in undesired directions the bias DC source should be coupled through an inductor while the signal AC source is coupled through a capacitor. This makes sure that each signal component (DC and AC) can only take its appropriate path.
It's important that the DC source is current regulated (not voltage) because the solenoid coil will change resistance as it warms up and from a source that stays on a long time it will be warming up. The AC signal can be voltage regulated if need be, but beware that its absolute current magnitude must never exceed the DC bias, otherwise the polarity of the coil will start flipping at the peaks and this will make the average B-field and the average plunger position no longer fixed and well-defined.
Also beware that the entire signal must never drive the mechanical spring-mass-system out of linear operation (like by compressing a spring too much) and this also includes effects of resonance (the control of which is not always trivial).
P.S. It's also important for the bias current source to consistently maintain correct bias even when resonance effects drive the voltage across the solenoid coils to high peak levels (possibly negative and/or even higher than the voltage of the power supply that supplies the bias current source in the first place) - the bias source needs to maintain linear operation all the time and this also may not always be trivial. -
#2825 by Monomorphic on 09 Dec, 2017 12:32
-
Looking at your vibrational thrust test graphs, I think there is a significant experimental error: You seem to have neither eliminated nor accounted for the changing offset / DC component in the movement of the plunger in your shaking mechanism. The reason is that you are using a unipolar signal source for the excitation and that the source turns off (rather than maintaining the average bias current through the solenoid) between the test pulses. This change in biasing conditions and the corresponding mass shift in the shaker mechanism results in a mass shift of the entire torsional balance, which, together with the long time constant of the balance, makes it look like a thrust signal.
I'm using a Voice Coil Actuator (VCA), not a solenoid. The VCA is driven by an analog audio signal. I use mono WAV files created in Audacity to generate the various signal frequencies and waveform shapes to test. Those are stored on microSD, which is accessed by the arduino, which then converts it to analog, sends it to the 3W amplifier, and finally to the VCA.
In the image below, the two bottom waveforms are sinusoidal chirps, in case anyone was wondering what those looked like.the corresponding mass shift in the shaker mechanism results in a mass shift of the entire torsional balance, which, together with the long time constant of the balance, makes it look like a thrust signal.
The shakers were designed to test the following criticism of the mach effect: "Due to the asymmetric motion inside the box, the center of mass of the box and its contents shifts relative to the box. But the center of mass must still remain where it was before (relative to the laboratory). So the box moves aside, while its center of mass stays put. Newton's laws were working properly, as they always do."
The idea being that it is fairly easy to produce a false-positive thrust trace on a torsional pendulum with a vibrating device. I certainly do not believe I'm producing actual thrust with the shaker.
-
#2826 by xyzzy on 09 Dec, 2017 13:02
-
OK, a voice coil is quite another story. Sorry for the confusion.
A question remains however: how does the arduino output the analog signal and how does the power amplification work? From the graphs (especially the lower frequencies) it looks like the signal is not entirely DC-free.
A voice coil does not need a bias source, so forget everything I was telling about a bias supply. But a voice coil needs to be free of any DC bias however.
It looks like the digital-to-analog conversion is somewhere introducing an offset, each time that the shaker is operating. Maybe it's not a real DAC but some sort of PWM output mechanism with a filter - and the PWM is being switched off entirely (output going to zero) between the periods of shaker operation, rather than continuing to drive out a "neutral" 50% duty cycle.
Can you check this: what is the average voltage that comes out of the amplifier? Best to check with an analog DC voltmeter. If the signal is fully symmetric and has the correct "neutral" output between the periods of shaker operation (and if each period is itself DC-free) then the meter needle should not visibly move away from zero (it would vibrate in place, but not move on average).
Regards
P.S. Of course it's clear that the shaker does not produce thrust. I did not imply any such thing or belief
But it's easy to introduce a DC offset in the drive circuitry - either with each pulse or in the time between the pulses. The effect on the measurement of such a mistake would be enormous, orders of magnitude more than the effect you were testing for (mechanical nonlinearities in the materials when subject to vibration), and it would easily dwarf such nonlinearities in comparison if it happened.
P.P.S. If the meter needle makes a (visible) swing in opposite directions at the start and end of each pulse (of each period of operation) then something is clearly wrong with the DC offset difference between operation and non-operation. -
#2827 by Monomorphic on 09 Dec, 2017 13:29
-
Can you check this: what is the average voltage that comes out of the amplifier? Best to check with an analog DC voltmeter. If the signal is fully symmetric and has the correct "neutral" output between the periods of shaker operation (and if each period is itself DC-free) then the meter needle should not visibly move away from zero (it would vibrate in place, but not move on average).
This is something I can easily check. And if present, I can compensate for it by adjusting the bias of the waveform in software rather than requiring physical biasing. That is much exaggerated in the image below to illustrate.
-
#2828 by xyzzy on 09 Dec, 2017 14:18
-
This is something I can easily check. And if present, I can compensate for it by adjusting the bias of the waveform in software rather than requiring physical biasing. That is much exaggerated in the image below to illustrate.
Please do. And if you happen to find anything strange, please also check the code logic for simple configuration mistakes - like forgetting to output a "neutral" 50% duty PWM (or to set a DAC to an appropriate "neutral" value) between output pulses (or generating the output pulses themselves with a wrong midpoint in amplitude). Note that some "hard" cases of such code mistakes cannot be completely compensated in the waveform offset (e.g. when the compensation would have required the output to go below a power supply rail). -
#2829 by WarpTech on 09 Dec, 2017 15:56
-
Can you check this: what is the average voltage that comes out of the amplifier? Best to check with an analog DC voltmeter. If the signal is fully symmetric and has the correct "neutral" output between the periods of shaker operation (and if each period is itself DC-free) then the meter needle should not visibly move away from zero (it would vibrate in place, but not move on average).
This is something I can easily check. And if present, I can compensate for it by adjusting the bias of the waveform in software rather than requiring physical biasing. That is much exaggerated in the image below to illustrate.
If your chirp is shifting the frequency, to avoid a DC offset you must maintain the same Volt-seconds in each half cycle. The period should only change like a staircase, shift some Hz ONLY after 1 complete cycle, not a continuous ramp. Otherwise, the DC will be inherent in the chirp input.
I know this because it's what causes transformer saturation in DC to AC Inverters that use power transformers. DC offset of the AC output when adjusting frequency can be a big problem and has to be accounted for in the design. -
#2830 by ThatOtherGuy on 12 Dec, 2017 08:14
-
Can you check this: what is the average voltage that comes out of the amplifier? Best to check with an analog DC voltmeter. If the signal is fully symmetric and has the correct "neutral" output between the periods of shaker operation (and if each period is itself DC-free) then the meter needle should not visibly move away from zero (it would vibrate in place, but not move on average).
This is something I can easily check. And if present, I can compensate for it by adjusting the bias of the waveform in software rather than requiring physical biasing. That is much exaggerated in the image below to illustrate.
Jamie, did you try using a mix of sin and square (or trapezoidal) waves ? -
#2831 by Monomorphic on 12 Dec, 2017 11:51
-
Jamie, did you try using a mix of sin and square (or trapezoidal) waves ?
I am not sure how to go about creating a waveform shaped like that. My guess is it would have characteristics between the sine and square. Sine waves are the least energetic of the three from what I can tell. It's a gentler shake compared to square and sawtooth waves of equal magnitude.
-
#2832 by PotomacNeuron on 12 Dec, 2017 14:15
-
I am not sure how to go about creating a waveform shaped like that. My guess is it would have characteristics between the sin and square. Sin waves are the least energetic of the three from what I can tell. It's a gentler shake compared to square and sawtooth waves of equal magnitude.
Monomorphic, symmetrical waveforms with no DC does not guarantee symmetrical movements. Your voice coil works with a magnetic field, usually generated by permanent magnets. The field strengths usually are not equal on different side of the coil rest position. -
#2833 by VAXHeadroom on 13 Dec, 2017 16:16
-
Jamie, did you try using a mix of sin and square (or trapezoidal) waves ?
I am not sure how to go about creating a waveform shaped like that. My guess is it would have characteristics between the sine and square. Sine waves are the least energetic of the three from what I can tell. It's a gentler shake compared to square and sawtooth waves of equal magnitude.
I have a full recording studio and can generate any waveform you might need. Let me know frequency, any dF/dt, and waveshape and I'll see if I can help out.
-- Emory
-
#2834 by Stormbringer on 13 Dec, 2017 17:59
-
The mischievous part of me wants to know if the following frequency pattern equals warp drive when used as driver frequencies for an EM drive or MET Drive? Its the voyager pulsar map coordinates for earth from Voyager's gold record with the map pulsar frequencies or harmonics thereof along with three vocal synthesizers.Jamie, did you try using a mix of sin and square (or trapezoidal) waves ?
I am not sure how to go about creating a waveform shaped like that. My guess is it would have characteristics between the sine and square. Sine waves are the least energetic of the three from what I can tell. It's a gentler shake compared to square and sawtooth waves of equal magnitude.
I have a full recording studio and can generate any waveform you might need. Let me know frequency, any dF/dt, and waveshape and I'll see if I can help out.
-- Emory
-
#2835 by Peter Lauwer on 13 Dec, 2017 19:19
-
How about the Polish guys? Are they still having positive results?
-
#2836 by Monomorphic on 13 Dec, 2017 21:13
-
How about the Polish guys? Are they still having positive results?
They are in the process of upgrading their test rig for the next level of the competition. My understanding is they are currently trying to find a second-hand laser displacement sensor (LDS) in Poland with decent stats. They are also considering switching from a teeter-totter to a torsional pendulum, but I do not think they have made the final decision. -
#2837 by Monomorphic on 15 Dec, 2017 21:31
-
The on-board PC has audio functionality, but it is "audio over HDMI." The signal is digital, not analog. This is why I originally went with an arduino as it can output an, albeit low quality, analog signal directly. The arduino is not only tedious, but also redundant since I already have a much more sophisticated on-board computer.
I had to use an HDMI audio extractor and display emulator so that I can use the computer to generate the signals. These signals are much higher quality than what the arduino can handle. I also upgraded the 3W amplifier to 25W
I would be further along but I got a stomach bug last week and then it snowed a bunch. But these upgrades should allow me to finish testing with the asymmetric shaker more quickly as now I can send any signal I wish at will without having to save to microSD.
-
#2838 by zellerium on 17 Dec, 2017 17:31
-
Hi all,
I wanted to publish my collected EM Drive report that I finished back in Dec 2015 on arXiv but realized I need an endorsement from someone who already has publications. Is anyone willing to endorse me?
Attached is the collected report, nothing has been updated since I finished it. Hopefully someone can find something useful in there!
Thanks,
Kurt
-
#2839 by PotomacNeuron on 18 Dec, 2017 00:41
-
Hi all,
I wanted to publish my collected EM Drive report that I finished back in Dec 2015 on arXiv but realized I need an endorsement from someone who already has publications. Is anyone willing to endorse me?
Attached is the collected report, nothing has been updated since I finished it. Hopefully someone can find something useful in there!
Thanks,
Kurt
Ability to endorse expires in 6 months. So I can not endorse you. But you can do as follows: browse current papers of your target field (I used pop-ph), find somebody who has the endorsement ability, directly send an endorsement request to him/her. Follow that request with an email with your paper attached. Often that person will just endorse you. If you do not hear back from him in 1 or 2 days, ask somebody else. Good luck!
The air suspension platforms are very unreliable instruments to work with. For critiques, see, e.g., Marc Millis, Nonviable mechanical “antigravity devices, in: M.G. Millis and E.W. Davis (eds.), Frontiers of propulsion science, AIAA, 2009, pp. 249–261.
