what is the proper battery orientation to minimize the noise?
Orientation of the battery is critical with respect to what direction?
misaligned with respect to what ?
When AA batteries are discharging, they produce a weak magnetic field. If these fields do not align with Earth's magnetic field, then they are constantly torquing the torsional pendulum - which produces back and forth movement we see as noise. Proper orientation of AA batteries seems to be north to ground and south to positive.
The 12V battery is internally composed of three 4V cells. So its orientation is different than the AA batteries.
I do plan on more tap tests now that the mass has been reduced.
Hello, I suggest using magnetic shielding material to cover the battery and wire.
The frequency shift as the photons move vertically is equivalent to what they would experience if the device were strapped to a rocket in free space accelerating at 1 g, which is way more than any horizontal experiment with the emDrive has experienced. The effect would still be almost negligible however (If you actually read X_Ray's attached picture, you would have the numerical answer to your question), and much smaller than and irrelevant to the claimed "guide wavelength" change in Shawyer's theory.
meberbs,
With due respect, I believe this is a mistaken application of the equivalence principle. Initially monochromatic energy in a multi-mode waveguide will be Doppler-spread as the cavity is accelerated, far more than a cavity in a gravitational field with the same acceleration. The gravitational potential gradient is over a significant dist
ance. The reflection off a metal termination is very abrupt in contrast.
No, it is exactly the equivalence principle. As you said:
You inject energy (say, in the middle) of a frustrum that is accelerating. It takes time for that energy to propagate from the injection point, to both the top and bottom. By the time the energy propagates to the top and bottom, they are moving at a new velocity, since I did say, the frustrum is accelerating. Consequently, the energy is Doppler-spread in frequency. The dispersion of the conical frustrum amplifies the effect.
This exact same effect happens to the exact same magnitude when the frustum is pointed upwards in 1 g, or strapped to a rocket in deep space accelerating at 1 g. The size of this effect even at 1 g is miniscule at the frequencies and lengths we are discussing. Saying the dispersion of the frustum amplifies the effect gives an inaccurate perception, since the scales of the 2 effects aren't even remotely similar.
I ought to have my computer do the math. Because it seems obvious to me that an accelerating can of light, being Doppler-scattered by the top and bottom lids, is not the same as a can of light permeated by gravity. Gravity affects the vacuum itself, and light and matter fields/waves warped by the gravity. Accelerating a can of light in deep space doesn't accelerate the vacuum inside the can at all, nor change the light/matter fields/waves propagating in it, until they scatter off the accelerating can boundaries. The can is a shell being accelerated through the vacuum.
If you shake a can, you don't shake the vacuum inside the can, as if a gravitational wave was propagating through it! You shake the can through the vacuum, and consequently with a relative velocity to energy propagating in it.
I ought to have my computer do the math. Because it seems obvious to me that an accelerating can of light, being Doppler-scattered by the top and bottom lids, is not the same as a can of light permeated by gravity.
These two situations being indistinguishable is literally the basis for general relativity. In a closed box there is no way to tell whether you are being accelerated by a rocket in free space or sitting on the ground in a uniform gravitational field. That includes if you reflect light off the top and bottom of the box and measure the Doppler shift. Intuition generally does not match with general relativity.
Hey monomorphic, was wondering if you have a planned test campaign from here in?
If results are good, then I will work on slowly ramping up the RF power to ~20W.
monomorphic: after the various tests and checks (orientation and all that) and before trying to increase the power, have you considered "pulsing" the power ? I mean, instead of constantly powering the drive for a given time, using an on/off approach somewhat "modulating" the power; not sure but I wonder if such a thing may start some kind of resonance effect
Either that, or you could "invert the flow while re-routing to the auxiliary backup" - it always seems to work on TV, and that's where field propulsion has been most successful so far.
monomorphic: after the various tests and checks (orientation and all that) and before trying to increase the power, have you considered "pulsing" the power ? I mean, instead of constantly powering the drive for a given time, using an on/off approach somewhat "modulating" the power; not sure but I wonder if such a thing may start some kind of resonance effect
Pulsing the Rf input can be interesting. Talking about VERY SHORT pulses here. With a measured cavity loaded Q of 8,000 the TC is 1,034sec, so 20% is then around 200nsec.
Suggest to start around 20% of 1 cavity fill TC and to have the reflected power output of the circulator open so there is min energy loss exiting the input radiant antenna, which is now acting like a receiving antenna.
The circulator's reflected power output does need to have a 50 ohm dummy load attached when measuring and tuning for min reflected power but when doing pulsed input, it needs to be open circuit or the dummy load will quickly drain the cavity of stored energy and thermalise it.
Think of what happens when you charge a coil and then open the circuit with or without a snubber circuit across the coil. Shell found out when she fried her dual stub antenna.
what is the proper battery orientation to minimize the noise?
Orientation of the battery is critical with respect to what direction?
misaligned with respect to what ?
When AA batteries are discharging, they produce a weak magnetic field. If these fields do not align with Earth's magnetic field, then they are constantly torquing the torsional pendulum - which produces back and forth movement we see as noise. Proper orientation of AA batteries seems to be north to ground and south to positive.
The 12V battery is internally composed of three 4V cells. So its orientation is different than the AA batteries.
I do plan on more tap tests now that the mass has been reduced.
The earlier posted suggestion to shield your entire balance with mu-metal may be a way too expensive, but putting the batteries in a m-metal box is something I might do. (in the mean time a bit jealous that I could not spend enough time on the project yet to keep up with you).
...
The circulator's reflected power output does need to have a 50 ohm dummy load attached when measuring and tuning for min reflected power but when doing pulsed input, it needs to be open circuit or the dummy load will quickly drain the cavity of stored energy and thermalise it.
...
OK, I did not know that. Thanks! Important.
Hey monomorphic, was wondering if you have a planned test campaign from here in?
Once the battery noise issues are resolved, which I think I am very close to doing, then I will proceed with more powered tests at the ~2W level. This cavity is designed to resonate mode TE013 at 2.45Ghz, so most of the testing will be at that frequency. However, there is another unidentified mode at a higher frequency that I believe is within the range of my RF equipment. I would also like to perform a few tests at that frequency.
Besides the standard forward orientation, I will rotate the frustum 180 degrees, while keeping all the RF equipment unchanged to see if there is a thrust reversal. I will also try a null position which is 90 degrees away from the current orientation. I will also test using a dummy load. I may also test a cylindrical cavity as a control, but that has not been designed or fabricated.
If results are good, then I will work on slowly ramping up the RF power to ~20W.
As for lab assistants, I would be open to someone assisting me, but it would have to be on a volunteer basis.
Thanks for the info. Sounds good.
When I listed frequency as a variable, I was thinking:
1. Using an off resonance frequency as a dummy/null load. It should have the same thermal response (?), but maybe not Lorentz (
) as a frustrum in resonance (as it's the same amount of watts going into the system, whether it's in resonance or not). Useful piece of data showing that there has to be resonance for thrust, and maybe discounting thermal as the (main) source of thrust.
And
2. Exploring the thrust per Q relationships for frequencies near but not at peak resonance. If your frustrum is 8000 Q, can you detune the freq to 4000 Q or 2000 Q and measure thrust, with everything else constant? Or is it all or nothing as far as Q goes? Would be massively interesting to show the relationship.
Also - once you start producing sets of null and positive results, it would be nice to reduce the graphs to some numerical figures that one can do stats on, to compare and average runs and correlate variables. Could be as simple as average (absolute?) displacement before power on, during power on, and after power on? Thoughts?
monomorphic: after the various tests and checks (orientation and all that) and before trying to increase the power, have you considered "pulsing" the power ? I mean, instead of constantly powering the drive for a given time, using an on/off approach somewhat "modulating" the power; not sure but I wonder if such a thing may start some kind of resonance effect
Pulsing the Rf input can be interesting. Talking about VERY SHORT pulses here. With a measured cavity loaded Q of 8,000 the TC is 1,034sec, so 20% is then around 200nsec.
Suggest to start around 20% of 1 cavity fill TC and to have the reflected power output of the circulator open so there is min energy loss exiting the input radiant antenna, which is now acting like a receiving antenna.
The circulator's reflected power output does need to have a 50 ohm dummy load attached when measuring and tuning for min reflected power but when doing pulsed input, it needs to be open circuit or the dummy load will quickly drain the cavity of stored energy and thermalise it.
Think of what happens when you charge a coil and then open the circuit with or without a snubber circuit across the coil. Shell found out when she fried her dual stub antenna.
That's not quite what happened.
Shell
...
The circulator's reflected power output does need to have a 50 ohm dummy load attached when measuring and tuning for min reflected power but when doing pulsed input, it needs to be open circuit or the dummy load will quickly drain the cavity of stored energy and thermalise it.
...
OK, I did not know that. Thanks! Important.
Be careful. I would be afraid of this frying my equipment without a dummy load. Phil, didn't you say you fried your main amp doing these kinds of pulsed tests?
I would need to purchase a different signal generator to effectively do pulsed tests. I guess I could click very quickly!
1. Using an off resonance frequency as a dummy/null load. It should have the same thermal response (?),
Off resonance frequencies are almost fully reflected back into the antenna and circulator, and then absorbed by the attenuator and spectrum analyser. So the thermal response will be different than using a resonance frequency as the frustum has a different thermal response compared to the attenuator due to the large surface area of the frustum.
...
The circulator's reflected power output does need to have a 50 ohm dummy load attached when measuring and tuning for min reflected power but when doing pulsed input, it needs to be open circuit or the dummy load will quickly drain the cavity of stored energy and thermalise it.
...
OK, I did not know that. Thanks! Important.
Be careful. I would be afraid of this frying my equipment without a dummy load. Phil, didn't you say you fried your main amp doing these kinds of pulsed tests?
Jamie,
Amp stopped producing an output because the input variable attenuator failed to produce an output. Output stage was fine. My pulse tests were never at 100W, plus back then I didn't have a circulator between the amp & frustum, so any back EMF from the cavity had a 50 ohm load to deal with. Amp failed before then. I did hear Paul had to repair a few amp output stages. He also didn't use an isolator / circulator with dummy load for reflected power.
https://www.microwaves101.com/encyclopedias/isolatorsThat said, Roger has warned of the dangers that high Q, high stored energy cavities can generate when improperly dealt with. He recommends using isolators and staying some distance away from the thruster when powered.
Please be careful if your circulator's reflected power output port is open circuit, especially if the Rf off time is very quick, small fraction of the cavity 1xTC. Once the Rf is switched off, the now depowered antenna is in a high stored E field and there will be electrical energy generated by the antenna and fed out of the cavity to the external equipment.
monomorphic: after the various tests and checks (orientation and all that) and before trying to increase the power, have you considered "pulsing" the power ? I mean, instead of constantly powering the drive for a given time, using an on/off approach somewhat "modulating" the power; not sure but I wonder if such a thing may start some kind of resonance effect
Pulsing the Rf input can be interesting. Talking about VERY SHORT pulses here. With a measured cavity loaded Q of 8,000 the TC is 1,034sec, so 20% is then around 200nsec.
Suggest to start around 20% of 1 cavity fill TC and to have the reflected power output of the circulator open so there is min energy loss exiting the input radiant antenna, which is now acting like a receiving antenna.
The circulator's reflected power output does need to have a 50 ohm dummy load attached when measuring and tuning for min reflected power but when doing pulsed input, it needs to be open circuit or the dummy load will quickly drain the cavity of stored energy and thermalise it.
Think of what happens when you charge a coil and then open the circuit with or without a snubber circuit across the coil. Shell found out when she fried her dual stub antenna.
That's not quite what happened.
Shell
Shell,
Though you had a theory the dual stub antenna turned into "match sticks" as the cavity discharged?
I ought to have my computer do the math. Because it seems obvious to me that an accelerating can of light, being Doppler-scattered by the top and bottom lids, is not the same as a can of light permeated by gravity.
These two situations being indistinguishable is literally the basis for general relativity. In a closed box there is no way to tell whether you are being accelerated by a rocket in free space or sitting on the ground in a uniform gravitational field. That includes if you reflect light off the top and bottom of the box and measure the Doppler shift. Intuition generally does not match with general relativity.
Yes, there is no need for any "computer do the math" to know the problem with General Relativity, because if mwvp's computer would give an answer that makes a difference between the two proposed situations, then it would mean that mwvp's computer model would not be satisfying one of the basic tenets of General Relativity, which has been experimentally confirmed by numerous experiments (STEP, Galileo-Galilei satellite, etc.)
...
I respectfully disagree with this opinion. While Monomorphic's does have some of the features of the step response of a second order system it lacks some of the key identifying characteristics. A 2nd order step response has a steep slope immediately after the driving force is applied. This can be seen in the capacitive calibration force the EW lab used in their experiments (3rd picture). When the voltage is switched off the inverse of that response is seen. We don't see that in Monomorphic's waveform. Instead it wobbles around a bit and then drops while the RF is still on. ...
I get the impression that you have not been closely following the thread for the last dozens of pages
, and are commenting without the background of the main point that the posts are addressing , as there is no disagreement to be had to what you point out and those points (concerning switch on and off delays) have been addressed in other posts. Some of that has to do with you taking too literally by what has been marked as "RF ON", and "RF OFF" as being identical to "cavity being at peak resonance" at those instances . Also, you do not seem to be taking into account the noise (at least ~+/2 microNewtons) in your remarks. Rather, my post was in response to TheTraveller saying that there were features of what Shawyer calls the "motor" and "generator" modes in the EM Drive experimental response
near its maximum and my post said that nothing strange like that could be seen.
Or are you, zen-in now in agreement with TheTraveller as seeing "motor" and "generator" features in Monomorphic's results at the top (namely the overshoot, the decaying oscillations around the step response)?
Also as has been discussed, what remains to be established is whether this is a step response due to an experimental electromagnetic artifact (like the purported interaction of the battery with the Earth's magnetic field) or a real EM Drive electromagnetic propulsion effect that can be used in Space.
I have been playing hooky for several months. My only interest is whether or not the pendulum response has the characteristics of a 2nd order step response. I don't subscribe to the motor/generator theory of operation. That may be another violation of accepted physical laws and has never been demonstrated to my satisfaction. My interest in system response analysis of the EM-Drive on this forum goes back quite awhile. I agree that spring battery contacts can act like a magnet and could generate a torque by reacting with the geomagnetic field. Maybe that torque can be measured separately.
I would need to purchase a different signal generator to effectively do pulsed tests. I guess I could click very quickly!
Can't you control power using a bit of computer code ?
...
The circulator's reflected power output does need to have a 50 ohm dummy load attached when measuring and tuning for min reflected power but when doing pulsed input, it needs to be open circuit or the dummy load will quickly drain the cavity of stored energy and thermalise it.
...
OK, I did not know that. Thanks! Important.
Be careful. I would be afraid of this frying my equipment without a dummy load. Phil, didn't you say you fried your main amp doing these kinds of pulsed tests?
I would need to purchase a different signal generator to effectively do pulsed tests. I guess I could click very quickly!
Could you use a PIN or FET SPST switch to alternate between a dummy load and the antenna?
monomorphic: after the various tests and checks (orientation and all that) and before trying to increase the power, have you considered "pulsing" the power ? I mean, instead of constantly powering the drive for a given time, using an on/off approach somewhat "modulating" the power; not sure but I wonder if such a thing may start some kind of resonance effect
Pulsing the Rf input can be interesting. Talking about VERY SHORT pulses here. With a measured cavity loaded Q of 8,000 the TC is 1,034sec, so 20% is then around 200nsec.
Suggest to start around 20% of 1 cavity fill TC and to have the reflected power output of the circulator open so there is min energy loss exiting the input radiant antenna, which is now acting like a receiving antenna.
The circulator's reflected power output does need to have a 50 ohm dummy load attached when measuring and tuning for min reflected power but when doing pulsed input, it needs to be open circuit or the dummy load will quickly drain the cavity of stored energy and thermalise it.
Think of what happens when you charge a coil and then open the circuit with or without a snubber circuit across the coil. Shell found out when she fried her dual stub antenna.
That's not quite what happened.
Shell
Shell,
Though you had a theory the dual stub antenna turned into "match sticks" as the cavity discharged?
They turned into matchsticks running at high Q, ionizied the air, arced and suffered a breakdown, lost resonance and then fed back frying the coax and magnetron. It would have worked well as a magnetron can withstand a high VSWR being a tube unlike a SS device but there is a limit in everything. I didn't take into account that the very air inside of the frustum would be the root cause.
Pulsing the RF will open up a different set of issues and a dump load
is needed unless you can take the antenna out of the equation and keep the decaying energy inside of the frustum which then can be manipulated. I believe I've done that, although I'm waiting to finish off a provisional patent before I disclose how.
My Very Best,
Shell
) as a frustrum in resonance (as it's the same amount of watts going into the system, whether it's in resonance or not). Useful piece of data showing that there has to be resonance for thrust, and maybe discounting thermal as the (main) source of thrust.
, and are commenting without the background of the main point that the posts are addressing , as there is no disagreement to be had to what you point out and those points (concerning switch on and off delays) have been addressed in other posts. Some of that has to do with you taking too literally by what has been marked as "RF ON", and "RF OFF" as being identical to "cavity being at peak resonance" at those instances . Also, you do not seem to be taking into account the noise (at least ~+/2 microNewtons) in your remarks. Rather, my post was in response to TheTraveller saying that there were features of what Shawyer calls the "motor" and "generator" modes in the EM Drive experimental response 