-
#3440 by meberbs on 02 Jun, 2018 07:11
-
Thanks for doing this. It is TE013 at 2.453GHz as predicted by my spreadsheet as attached. Your resonance is shown at 2.455GHz or 2MHz higher than my value. Seems we agree.
The shape is basically the definition of a TE313 mode. That first "m=3" specifically refers to the circular symmetry, which has 6 nulls (2*m). The lower frequency mode from the simulation has the pattern of a TE214 mode. TE013 is a mode that should be lower frequency than either of these modes.
I find the last one in that sweep interesting. It seems like a TE02X mode, but those don't come up much, so I am not sure. I am not sure what X is, since there seems to be half of a "lobe" next to the big end. It might be one of those hybrid modes that seems to be a mix of TE and TM.Haven't used this spreadsheet for some time now. Will do some work on it as it will be released as part of the KISS thruster design & build package.
Unless you can fix your spreadsheet to accurately predict these different kinds of modes, it is just a waste of time.The wavy side wall eddy currents are caused by the coupler not being exactually at 1/4 guide wave from the big end. Which is why in the KISS thruster build, the coupler will be able to be moved closer or further away from the big end to achieve ideal coupler excitation vs travelling wave phase matching.
No, problems due to poor antenna placement would appear as an asymmetric distortion. Usually with a descent Q, this would not noticeably affect the field pattern, though it is a problem for getting high return loss.BTW in TE313 mode the small end is heavily cutoff with a cutoff freq of 2.672GHz. So any excitation needs to be above the cutoff freq.
Still no reason that cutoff would be a problem, other than "Shawyer said so" and it has been long established that Shawyer has no clue what he is talking about. Also, the depicted mode is TE313, and cutoff clearly is not present.Actually the delay was a good thing as I have a lot more experience to throw into building the thruster, the Rf system, how to tune the coupler,
Yet you keep making incorrect claims about basic RF principles.explaining how it works, etc.
Then maybe you could now actually answer the simple questions I have asked you about what direction something moves when you bounce a ball off of it. Or better yet, you could recognize that Shawyer's claims are self contradictory nonsense, and that he has failed at basic physics. -
#3441 by X_RaY on 02 Jun, 2018 11:40
-
Using the dimensions as suggested by yourself there is no TE013 mode between 2.4 GHz & 2.5 GHz. As shown by Monomorphic in this post there are three other field pattern instead. The one you called TE013 is TE313, clearly indicated by the current vectors.......
It is TE013.
Feko does wavy side wall eddy current patterns when the position of the coupler is not quite correct. Example is the thruster Roger and I worked on for NASA, with Feko help from Jamie.
The patter you show above in your post is of course TE013 but with some deformations. This is most likely due to the visualization frequency is slightly beyond the resonance frequency (typically feko visualisize the calculated frequencies, not each other possible point on the interpolated graph).
https://forum.nasaspaceflight.com/index.php?topic=39214.msg1614781#msg1614781
edit:
mode index corrected
-
#3442 by TheTraveller on 02 Jun, 2018 14:29
-
Using the dimensions as suggested by yourself there is no TE013 mode between 2.4 GHz & 2.5 GHz. As shown by Monomorphic in this post there are three other field pattern instead. The one you called TE013 is TE313, clearly indicated by the current vectors.......
It is TE013.
Feko does wavy side wall eddy current patterns when the position of the coupler is not quite correct. Example is the thruster Roger and I worked on for NASA, with Feko help from Jamie.
The patter you show above in your post is of course TE013 but with some deformations. This is most likely due to the visualization frequency is slightly beyond the resonance frequency (typically feko visualisize the calculated frequencies, not each other possible point on the interpolated graph).
https://forum.nasaspaceflight.com/index.php?topic=39214.msg1614781#msg1614781
edit:
mode index corrected
Would suggest you apply the Bessel J function for TE313 and see that you get.
Then explain the 6 side wall modes observed in this Feko analysis of TE013 due to the coupler being not at the ideal side wall position? Please let me know your opinion why those very visible big end eddy currents in the attached image, Jamie's work not mine, show a clear 6 node wddy current image that you now claim is TE313 mode?
In reality everything you claim as fact will become null and void when you see the KISS thruster going round and round. Are you ready for that?
It Is Time For The EmDrive To Come Out Of The Shadows
-
#3443 by meberbs on 02 Jun, 2018 15:52
-
Would suggest you apply the Bessel J function for TE313 and see that you get.
Using the exact analytic, solution for curved end plates in spherical coordinates, Bessel functions apply to variations in the radial direction (from the small plate to the large plate). The third number (p) by convention of mode naming describes which scaling factor to use for the radial variations, effectively determining the number of lobes between the end plates. The first number on the other hand describes variations in the spherical Phi direction, which are simply a sin(m*phi) or cos(m*phi).Then explain the 6 side wall modes observed in this Feko analysis of TE013 due to the coupler being not at the ideal side wall position? Please let me know your opinion why those very visible big end eddy currents in the attached image, Jamie's work not mine, show a clear 6 node wddy current image that you now claim is TE313 mode?
X_Ray literally just answered this in the post you quoted. It is not due to antenna position, it is due to the tool using a slightly off resonance frequency for the visualization. It is clear in the image that you are posting, that fields are continuously present around the phi direction, just slightly distorted, whereas the TE313 image shows clear and obvious nulls.
besides you are ignoring the TE214 mode also present at a lower frequency in the sweep. If the TE313 mode was in fact a TE013, it would make no sense for any TEx14 mode to exist at a frequency lower than it.In reality everything you claim as fact will become null and void when you see the KISS thruster going round and round. Are you ready for that?
In the unlikely event that happens (unlikely based on the experimental data that no one else who has ever tested one of these devices has come within orders of magnitude of the force required for that), it will say nothing about the existing mode shapes we are discussing. 1+1 will still equal 2. The solutions to Maxwell's equations will not change, nor the fact that Maxwell's equations have predicted resonance correctly in every cavity tested.
The question for you (besides the other ones you have ignored) is if your experiment sits there without moving, beyond some small motion easily attributed to air currents or magnetic fields, are you ready for that? -
#3444 by Monomorphic on 02 Jun, 2018 19:57
-
Please let me know your opinion why those very visible big end eddy currents in the attached image, Jamie's work not mine, show a clear 6 node wddy current image that you now claim is TE313 mode.
Phil, the reason the TE013 looks different in the images for Paul March was because of a mistake in how the file was set up. If I recall correctly, that was when I was using Higher Order Basis Functions (HOBF) with too course of a mesh. The mode was distorted because of an error on my part. It is not a proper TE013 but since then we stopped using HOBF and use a dense mesh, the mode is no longer distorted. -
#3445 by Peter Lauwer on 02 Jun, 2018 21:10
-
In the meantime, I was working on my torsion balance (in the picture below it is shown without cover). Up to now, I have not been able to achieve the stability and low noise I had in the past. Most of the noise is mechanical. It turned out not to be a good idea to place it on the (concrete) floor. Even if I move a little, the balance notices the distortion of the floor.
I am mounting it on a construction attached to the wall now. I will report when this works better.
First going on a short holiday now.
Cheers, keep up the good work, I see a lot of progress with some of you now.
Peter
-
#3446 by X_RaY on 02 Jun, 2018 21:56
-
I had HOBF in mind but was not quite sure that this was the reason, however you are right i did a look to this feko solver funktion again.... in contrast when using a normal dense mesh there isn't this abnormal pattern anymore.Please let me know your opinion why those very visible big end eddy currents in the attached image, Jamie's work not mine, show a clear 6 node wddy current image that you now claim is TE313 mode.
Phil, the reason the TE013 looks different in the images for Paul March was because of a mistake in how the file was set up. If I recall correctly, that was when I was using Higher Order Basis Functions (HOBF) with too course of a mesh. The mode was distorted because of an error on my part. It is not a proper TE013 but since then we stopped using HOBF and use a dense mesh, the mode is no longer distorted.
Thanks for clarify.
Not as clear as in the case of Monomorphic's run but a good visible example of the effect when using HOBF:
-
#3447 by Kenjee on 03 Jun, 2018 00:51
-
Using the dimensions as suggested by yourself there is no TE013 mode between 2.4 GHz & 2.5 GHz. As shown by Monomorphic in this post there are three other field pattern instead. The one you called TE013 is TE313, clearly indicated by the current vectors.......
It is TE013.
Feko does wavy side wall eddy current patterns when the position of the coupler is not quite correct. Example is the thruster Roger and I worked on for NASA, with Feko help from Jamie.
The patter you show above in your post is of course TE013 but with some deformations. This is most likely due to the visualization frequency is slightly beyond the resonance frequency (typically feko visualisize the calculated frequencies, not each other possible point on the interpolated graph).
https://forum.nasaspaceflight.com/index.php?topic=39214.msg1614781#msg1614781
edit:
mode index corrected
There is nothing in that range, only complex modes with real and imaginary parts (i)
(The ratio of the imaginary to the real part gives the relative damping factor)
-
#3448 by Kenjee on 03 Jun, 2018 00:56
-
IMO, only modes worth pursuing with this dimensions are:
-
#3449 by Monomorphic on 03 Jun, 2018 01:39
-
IMO, only modes worth pursuing with this dimensions are:
Thanks Kenjee, I was hoping you would weigh in on this. When you have a chance, could you also do TT’s latest dimensions from the plastic form he found? I would do it myself, but I am staying at a friend’s lake house for a couple of days and do not have access to my workstation. Here are the dimensions:
Bd: 300mm
Sd: 180mm
Len: 280mm -
#3450 by TheTraveller on 03 Jun, 2018 03:18
-
IMO, only modes worth pursuing with this dimensions are:
Thanks Kenjee, I was hoping you would weigh in on this. When you have a chance, could you also do TT’s latest dimensions from the plastic form he found? I would do it myself, but I am staying at a friend’s lake house for a couple of days and do not have access to my workstation. Here are the dimensions:
Bd: 300mm
Sd: 180mm
Len: 280mm
Checked with Roger.
He calcs 2.2608Ghz in TE013.
I get 2.2821GHz in TE013. Clearly a 21.3MHz error is not accceptable.
So it seems something has changed in my spreadsheet. Roger is feeding me more data so I can work out what is not correct.
It Is Time For The EmDrive To Come Out Of The Shadows -
#3451 by TheTraveller on 03 Jun, 2018 03:32
-
IMO, only modes worth pursuing with this dimensions are:
In your 2nd image, the small is clearly in cutoff as the eddy current density on the small end plate is very much below that in the small end side wall ring.
Normally the small end plate eddy current ring shoud be MUCH higher intenssity than the side wall ring at the small end.
This occurs because as the diameter decreases, the ping angle off the side walls the photon take gets steeper and steeper until they never hit the end plate and just ping between the small end side wall lobe before returning to the big end. So YES you can have resonance in a tapered waveguide without the use of a small end plate. See attached. If you do not believe me, please check out the patent in the image.
This cutoff effect is clearly shown by progressively reducing the small end diameter and observing that as the small end starts to get into cutoff, the intensity of the side wall small end ring starts in increase as the small end plate ring intensity starts to decrease. Roger calls that point cutoff.
It Is Time For The EmDrive To Come Out Of The Shadows
-
#3452 by TheTraveller on 03 Jun, 2018 03:40
-
Please let me know your opinion why those very visible big end eddy currents in the attached image, Jamie's work not mine, show a clear 6 node wddy current image that you now claim is TE313 mode.
Phil, the reason the TE013 looks different in the images for Paul March was because of a mistake in how the file was set up. If I recall correctly, that was when I was using Higher Order Basis Functions (HOBF) with too course of a mesh. The mode was distorted because of an error on my part. It is not a proper TE013 but since then we stopped using HOBF and use a dense mesh, the mode is no longer distorted.
Jamie,
I found a simiar wavy side wall eddy current ring effect as the couper position was adjusted up and down the side wall. Can't replicate the effect as no longer have access to Feko.
Did you ever try that?
It Is Time For The EmDrive To Come Out Of The Shadows -
#3453 by TheTraveller on 03 Jun, 2018 03:58
-
Here is the patent, US 3425006 "Cavity Resonator with mode discriminating means", for the cavity resonator that does not use a small end plate.
It is an interesting read.
It Is Time For The EmDrive To Come Out Of The Shadows
-
#3454 by Kenjee on 03 Jun, 2018 14:59
-
IMO, only modes worth pursuing with this dimensions are:
Thanks Kenjee, I was hoping you would weigh in on this. When you have a chance, could you also do TT’s latest dimensions from the plastic form he found? I would do it myself, but I am staying at a friend’s lake house for a couple of days and do not have access to my workstation. Here are the dimensions:
Bd: 300mm
Sd: 180mm
Len: 280mm
-
#3455 by flux_capacitor on 03 Jun, 2018 19:07
-
Bd: 300mm
Sd: 180mm
Len: 280mm
Checked with Roger.
He calcs 2.2608Ghz in TE013.
COMSOL TE013 2.2548 GHz vs Shawyer TE013 2.2608 Ghz
Difference of +6 MHz for COMSOL vs Shawyer's private method! Not too bad
So finally Kenjee found resonance with TE013 mode for the shape TT provided.
Seems Monomorphic and X_RaY didn't find TE013 as their sweeps were restricted to a slightly higher range out of reach (2.4–2.5 GHz).
According to Shawyer's controversial "cutoff rule" the small end diameter is large enough for such frequency, i.e. it would operate above the threshold below which thrust would collapse. Besides, according to Shawyer's guidelines if I made no mistake: Df = 0.64, loaded Q = 60k, predicted thrust @ 100W = 38 mNFreq is outside the ISM S Band, so need to work on the dimension sets that the form can deliver. Can increase freq by reducing length but doing that increases the small end diameter, which causes the Df to drop, which drops the generated force.
This frequency is still microwaves. What is the problem of running outside the ISM S Band? Especially as solid-state RF power generators can produce any frequency? -
#3456 by Monomorphic on 03 Jun, 2018 23:36
-
This frequency is still microwaves. What is the problem of running outside the ISM S Band? Especially as solid-state RF power generators can produce any frequency?
In the US, ~2.25Ghz is used for NASA satellite tracking, telemetry and control (space-to-Earth, space-to-space). 2.29–2.3 GHz is the NASA Deep Space Network. In the UK and Australia, ~2.25Ghz looks unused/reserved, and is close to 3G Cellular Communications - Base Station Downlink (2.11–2.17 GHz).
This is why the 3D printed cavity resonator was designed to work on the 2.4–2.483GHz Industrial, Scientific and Medical (ISM) band, which is widely used for low power unlicensed microwave devices.
Fortunately, a well-built cavity will contain a vast majority of the microwave energy. I rebuilt the small end twice to get my leaking down to ~-40dB at ~0.5 meters from the cavity when off resonance. What's really convenient is that when the cavity goes into resonance, the leaked RF goes down even further, to lower than -60dB, which is the limit of the detection ability.
-
#3457 by oyzw on 04 Jun, 2018 00:08
-
30/5000This frequency is still microwaves. What is the problem of running outside the ISM S Band? Especially as solid-state RF power generators can produce any frequency?
In the US, ~2.25Ghz is used for NASA satellite tracking, telemetry and control (space-to-Earth, space-to-space). 2.29–2.3 GHz is the NASA Deep Space Network. In the UK and Australia, ~2.25Ghz looks unused/reserved, and is close to 3G Cellular Communications - Base Station Downlink (2.11–2.17 GHz).
This is why the 3D printed cavity resonator was designed to work on the 2.4–2.483GHz Industrial, Scientific and Medical (ISM) band, which is widely used for low power unlicensed microwave devices.
Fortunately, a well-built cavity will contain a vast majority of the microwave energy. I rebuilt the small end twice to get my leaking down to ~-40dB at ~0.5 meters from the cavity when off resonance. What's really convenient is that when the cavity goes into resonance, the leaked RF goes down even further, to lower than -60dB, which is the limit of the detection ability.
Hi, Mr. Jamie, is it all right now? How is your experiment progressing? -
#3458 by Monomorphic on 04 Jun, 2018 00:35
-
Hi, Mr. Jamie, is it all right now? How is your experiment progressing?
My experiment is ready to go. I purchased LabView and have used the sample code to communicate with the relay board. I need to work on getting the relays on timers in LabView and then I should be able to run multiple identical experiments and average the results.
I'm not sure how long that will take, so I plan on running several manual tests in the next few days. I will be posting that data here on NSF.
After these first high-power manual tests, I will also begin working on Oyzw's solid copper frustum. I need to 3D print some parts for the impedance matching controls, fashion a new antenna, and increase the size of the cavity some using a gasket ( to get it in the ISM band). I also need to build the aluminum frame that attaches the cavity to the thrust balance. It will be painted so IR cameras can be used to observe surface heating and mode shape. After that, it will be covered with insulation, attached to the thrust balance, and tested.
What most here do not know, but I have told others privately, is that my condo complex had two main water pipes burst within a week from one another, so we and 40 other homes have been without water for almost two weeks. I was one of those selected by the board of directors to help oversee the emergency and it ended up being one of the worst experiences of my life. One leak was under a giant crepe myrtle and the other was 10 ft below someone's brick patio. My spouse and I decided to go to a friend's lake house this weekend to get away from the fiasco and rest. Water has been restored today, finally, and I am happily home now and things are getting back to normal. -
#3459 by TheTraveller on 04 Jun, 2018 01:31
-
Thanks Kangee for doing that,
Like Jamie I no longer have Feko access and thus placing the side wall 1/4 excitation freq wavelength stub antenna coupler was going to be a hit and miss affair. Now I have a much better idea where to place the couper as shown. Plus will be using the attached SMA bulkhead fitting to allow the fitting to be screwed in and out from the fixed bulkhead mounting nuts to obtain both best length and best up and down side wall position.
As I will show, being able to tune the coupler, in length, position and angle is critical to obtaining optimal coupler efficiency and a very low VSWR without needing the use of an external 3 pot tuner.
Plus a side wall position is needed to generate dual travelling waves that, via superposition, form the resonant standing waves. I may be wrong but don't think an end plate coupler will be able to generate dual direction travelling waves if it is not placed 1/4 guide wavelength away from the big end plate.
In Roger's latest patent application he did use a small end plate located coupler which was placed 1/4 guide wave from the small end plate as shown. Which again shows the coupler in an EmDrive neeeds to be placed 1/4 guide wavelength from an end plate so as to obtain phase matching and to generate travelling waves that propograte in both direction, which then generate the resonant standing waves.
It is my opinion that using an external 3 pot tuner may allow the Rf amp to see an ideal load but it will not tune what is happening inside the EmDrive.
It Is Time For The EmDrive To Come Out Of The Shadows
