-
#1060
by
SeeShells
on 25 Aug, 2015 13:50
-
...
Please don't be offended but one of those EM Drives designed with help from computer models is scheduled to be tested "Today".
...
As previously posted, the EM Drive to be tested today has the antenna located at the small end of the cavity, while the latest computer analysis indicates that the antenna should be located at the completely opposite location: near the opposite end, near the big base.
Thus today's test will be instructive to see what happens when the antenna is located in the opposite direction than the one recommended by the latest computer analysis and the opposite direction than the one used by NASA Eagleworks in their positive tests.
We have to consider there are many ways to excite modes within the cavity and the general meep would need to be modified (which it can be) to work with them all. Dr. Rodel in taking the CSV files and post processing with mathematica is proving a very good workaround and has given us some very insightful data.
Cavity size rules them all.
Snub (s)
Loop antenna (s) TE and TM
waveguide coupling (s) TE and TM
direct magnetron (s) TE and TM
dipole (s) TM
helical TM?
modified helicals TM?
1/4 wave shifted dipoles TM
For me it has narrowed down to "build it and they will come", in other words select the best option at build and test, move, file, cut to fine tune it all. At least I have the VNA and SA to do it with.
Shell
cleaned up
-
#1061
by
SeeShells
on 25 Aug, 2015 14:26
-
The best thing is to use a loop antenna for TE modes
Ok then but how do you reconcile the need to put the loop antenna into the nodes of the mode (which are several centimeters apart) and the need for the loop antenna to have a perimeter equal to 1/4 wavelength (which gives a loop diameter of 3.5 to 4.5 cm)?
Don't put an antenna at a node. A node is location of zero amplitude. Instead locate the antenna at an anti-node: a location of maximum amplitude.
The orientation of the loop antenna for a TE mode should be such that the magnetic field goes through the center of the loop. The magnetic field in TE01 modes is oriented along the longitudinal axis of axi-symmetry. Thus the perimeter of the antenna is perpendicular to the longitudinal axis and hence perpendicular to the longitudinal propagation direction of the wave.
Fair enough, wrong terminology from my side. But my question is still valid for antinodes.
I wanted to talk about the centers of the blue or red "blobs" in Meep simulations.
I think this is right because of the asymmetry of the cavity the anti-nodes are not at 1/4 wavelengths from the side walls. Putting the antenna 1/4 wave will cause phase distortions and cancellations.
-
#1062
by
aero
on 25 Aug, 2015 15:47
-
The best thing is to use a loop antenna for TE modes
Ok then but how do you reconcile the need to put the loop antenna into the nodes of the mode (which are several centimeters apart) and the need for the loop antenna to have a perimeter equal to 1/4 wavelength (which gives a loop diameter of 3.5 to 4.5 cm)?
Don't put an antenna at a node. A node is location of zero amplitude. Instead locate the antenna at an anti-node: a location of maximum amplitude.
The orientation of the loop antenna for a TE mode should be such that the magnetic field goes through the center of the loop. The magnetic field in TE01 modes is oriented along the longitudinal axis of axi-symmetry. Thus the perimeter of the antenna is perpendicular to the longitudinal axis and hence perpendicular to the longitudinal propagation direction of the wave.
Fair enough, wrong terminology from my side. But my question is still valid for antinodes.
I wanted to talk about the centers of the blue or red "blobs" in Meep simulations.
Don't read to much into those blue or red "blobs" on the latest meep images. They are only there to show the antenna location. I cut the simulation off after only 6/10
th of a cycle so we can see the antenna but nothing has yet reached the side or end walls. So the blobs are only the antenna. They are located where Harminv calculates the best resonance from the antenna/cavity combination, but that is a different run, different image.
-
#1063
by
BL
on 25 Aug, 2015 15:50
-
Re post #1025 by rq3
Be wary of spark plug wire. The insulation rating will be more than adequate, but they are not designed to carry appreciable current. Most spark plug wire has a resistance of several thousand ohms/ft, by design, to suppress RFI. (One web post said that Ford specs around 7 k ohms/ft.), but plug wire from auto parts stores could vary from 5-10 k ohms/ft. Some plug wires however have low resistance cores, but they may be hard to find. If you decide to go with plug wire, just be sure that you use the ones with a low resistance wire core rather than the normal resistance cores.
-
#1064
by
TheTraveller
on 25 Aug, 2015 16:40
-
-
#1065
by
rfmwguy
on 25 Aug, 2015 17:25
-
1:15...everything ready to go. Never could convert ip cam stream to rtmp used for live streaming, so I will record with fixed camera starting in 45 minutes.
I've read a lot about modes and antenna placements without a direct link to the theory as to why this may move. So... I picked a design to get the data points.
If they are null with my current design, others have guidance. If it produces motion...well back to the drawing board for modeling em fields.
It has been a fun project, regardless. Next post will be a link to unedited video of NSF-1701's flight test.
-
#1066
by
deltaMass
on 25 Aug, 2015 17:45
-
A Time Zone would be helpful
Fingers crossed.
-
#1067
by
Eer
on 25 Aug, 2015 17:59
-
A Time Zone would be helpful
Fingers crossed.
Uhhhh ... 45 min from 1:15, or about now, whatever timezone you are in ... he seems to be in EDT
-
#1068
by
flux_capacitor
on 25 Aug, 2015 18:12
-
The best thing is to use a loop antenna for TE modes
Ok then but how do you reconcile the need to put the loop antenna into the nodes of the mode (which are several centimeters apart) and the need for the loop antenna to have a perimeter equal to 1/4 wavelength (which gives a loop diameter of 3.5 to 4.5 cm)?
Don't put an antenna at a node. A node is location of zero amplitude. Instead locate the antenna at an anti-node: a location of maximum amplitude.
The orientation of the loop antenna for a TE mode should be such that the magnetic field goes through the center of the loop. The magnetic field in TE01 modes is oriented along the longitudinal axis of axi-symmetry. Thus the perimeter of the antenna is perpendicular to the longitudinal axis and hence perpendicular to the longitudinal propagation direction of the wave.
Fair enough, wrong terminology from my side. But my question is still valid for antinodes.
I wanted to talk about the centers of the blue or red "blobs" in Meep simulations.
Don't read to much into those blue or red "blobs" on the latest meep images. They are only there to show the antenna location. I cut the simulation off after only 6/10th of a cycle so we can see the antenna but nothing has yet reached the side or end walls. So the blobs are only the antenna. They are located where Harminv calculates the best resonance from the antenna/cavity combination, but that is a different run, different image.
@aero, Shell and Rodal: To clarify my mind, here is attached what I mean by putting a loop antenna near the big end "in the middle of the
node lobe". My concern is
how could we reconcile such a large loop with the fact its perimeter should be equal to λ/4 to trigger resonance in the cavity (with lambda being between 12 and 15 cm)?
@X_Ray, this is a
simplified sketch of a loop antenna, for didactic popularization purpose
The important thing to consider is the location of the loop WRT the cavity walls, end plates and mode shape. For example the two horizontal rods could be build in such a way they would not electromagnetically interfere with the waves, unlike in the sketch.
-
#1069
by
Chris Bergin
on 25 Aug, 2015 18:13
-
Next post will be a link to unedited video of NSF-1701's flight test.
God speed NSF-1701!
-
#1070
by
graybeardsyseng
on 25 Aug, 2015 18:18
-
A Time Zone would be helpful
Fingers crossed.
Uhhhh ... 45 min from 1:15, or about now, whatever timezone you are in ... he seems to be in EDT
I believe in an earlier post he had indicated that T0 was 1400 EDT.
-
#1071
by
TheTraveller
on 25 Aug, 2015 18:25
-
My concern is how could we reconcile such a large loop with the fact its perimeter should be equal to λ/4 to trigger resonance in the cavity (with lambda being between 12 and 15 cm)?
X_Ray, this is a simplified sketch of a loop antenna, for didactic popularization purpose The important thing to consider is the location of the loop WRT the cavity walls, end plates and mode shape. For example the two horizontal rods could be build in such a way they would not electromagnetically interfere with the waves, unlike in the sketch.
Excite in TE013 mode and put the loop in the middle lobe as I indicated earlier.
-
#1072
by
flux_capacitor
on 25 Aug, 2015 18:37
-
My concern is how could we reconcile such a large loop with the fact its perimeter should be equal to λ/4 to trigger resonance in the cavity (with lambda being between 12 and 15 cm)?
X_Ray, this is a simplified sketch of a loop antenna, for didactic popularization purpose The important thing to consider is the location of the loop WRT the cavity walls, end plates and mode shape. For example the two horizontal rods could be build in such a way they would not electromagnetically interfere with the waves, unlike in the sketch.
Excite in TE013 mode and put the loop in the middle lobe as I indicated earlier.
Climbing to the second floor won't really change the problem.
(the "middle lobe" you want to use in TE013 is barely smaller than the lower lobe, hence a loop perimeter barely smaller than what I draw, and still longer than λ/4)
-
#1073
by
SteveD
on 25 Aug, 2015 18:48
-
If we don't hear from rmfguy by 3:00 EDT, is there someone here who can call him and make sure he's ok?
-
#1074
by
cee
on 25 Aug, 2015 18:49
-
Could someone please recap what resonance mode is expected for NSF-1701, looks like he is launching a TM mode to me.
-
#1075
by
TheTraveller
on 25 Aug, 2015 18:55
-
My concern is how could we reconcile such a large loop with the fact its perimeter should be equal to λ/4 to trigger resonance in the cavity (with lambda being between 12 and 15 cm)?
X_Ray, this is a simplified sketch of a loop antenna, for didactic popularization purpose The important thing to consider is the location of the loop WRT the cavity walls, end plates and mode shape. For example the two horizontal rods could be build in such a way they would not electromagnetically interfere with the waves, unlike in the sketch.
Excite in TE013 mode and put the loop in the middle lobe as I indicated earlier.
Climbing to the second floor won't really change the problem.
(the "middle lobe" you want to use in TE013 is barely smaller than the lower lobe, hence a loop perimeter barely smaller than what I draw, and still longer than λ/4)
So don't use an antenna inside the frustum.
Follow Shawyer & Prof Yang and inject the Rf into the middle lobe of the TE013 excited frustum via a tuned and impedance matched waveguide section as in her latest paper.
-
#1076
by
Rodal
on 25 Aug, 2015 18:59
-
My concern is how could we reconcile such a large loop with the fact its perimeter should be equal to λ/4 to trigger resonance in the cavity (with lambda being between 12 and 15 cm)?
X_Ray, this is a simplified sketch of a loop antenna, for didactic popularization purpose The important thing to consider is the location of the loop WRT the cavity walls, end plates and mode shape. For example the two horizontal rods could be build in such a way they would not electromagnetically interfere with the waves, unlike in the sketch.
Excite in TE013 mode and put the loop in the middle lobe as I indicated earlier.
Climbing to the second floor won't really change the problem.
(the "middle lobe" you want to use in TE013 is barely smaller than the lower lobe, hence a loop perimeter barely smaller than what I draw, and still longer than λ/4)
To excite TE01 you need to excite a magnetic field in the longitudinal direction of the axis of axi-symmetry. If you want to keep the loop perimeter λ/4, what prevents one from making a loop λ/4 perimeter and placing it so that its center goes through the axis of axi-symmetry?
-
#1077
by
Ryan900
on 25 Aug, 2015 19:03
-
Is there a list of materials that the existing builds have been made of?
-
#1078
by
SeeShells
on 25 Aug, 2015 19:11
-
My concern is how could we reconcile such a large loop with the fact its perimeter should be equal to λ/4 to trigger resonance in the cavity (with lambda being between 12 and 15 cm)?
X_Ray, this is a simplified sketch of a loop antenna, for didactic popularization purpose The important thing to consider is the location of the loop WRT the cavity walls, end plates and mode shape. For example the two horizontal rods could be build in such a way they would not electromagnetically interfere with the waves, unlike in the sketch.
Excite in TE013 mode and put the loop in the middle lobe as I indicated earlier.
Climbing to the second floor won't really change the problem.
(the "middle lobe" you want to use in TE013 is barely smaller than the lower lobe, hence a loop perimeter barely smaller than what I draw, and still longer than λ/4)
Only in for a bit for a bite.
Remember wrt the larger the loop exciting a TE mode the more it will seriously start to effect the Q. aero I believe is seeing the same situation using the 1/4 dipoles trying to maintain the higher Qs.
Also remember a loop doesn't need to be a real loop or circle either I've got to get out again but I would like for you to consider this design. Not a loop but a loop nonetheless. Spaced around the cavity in the top or the bottom. Not to scale, just for ref.
Shell
-
#1079
by
X_RaY
on 25 Aug, 2015 19:23
-
The best thing is to use a loop antenna for TE modes
Ok then but how do you reconcile the need to put the loop antenna into the nodes of the mode (which are several centimeters apart) and the need for the loop antenna to have a perimeter equal to 1/4 wavelength (which gives a loop diameter of 3.5 to 4.5 cm)?
Don't put an antenna at a node. A node is location of zero amplitude. Instead locate the antenna at an anti-node: a location of maximum amplitude.
The orientation of the loop antenna for a TE mode should be such that the magnetic field goes through the center of the loop. The magnetic field in TE01 modes is oriented along the longitudinal axis of axi-symmetry. Thus the perimeter of the antenna is perpendicular to the longitudinal axis and hence perpendicular to the longitudinal propagation direction of the wave.
Fair enough, wrong terminology from my side. But my question is still valid for antinodes.
I wanted to talk about the centers of the blue or red "blobs" in Meep simulations.
Don't read to much into those blue or red "blobs" on the latest meep images. They are only there to show the antenna location. I cut the simulation off after only 6/10th of a cycle so we can see the antenna but nothing has yet reached the side or end walls. So the blobs are only the antenna. They are located where Harminv calculates the best resonance from the antenna/cavity combination, but that is a different run, different image.
@aero, Shell and Rodal: To clarify my mind, here is attached what I mean by putting a loop antenna near the big end "in the middle of the node lobe". My concern is how could we reconcile such a large loop with the fact its perimeter should be equal to λ/4 to trigger resonance in the cavity (with lambda being between 12 and 15 cm)?
@X_Ray, this is a simplified sketch of a loop antenna, for didactic popularization purpose The important thing to consider is the location of the loop WRT the cavity walls, end plates and mode shape. For example the two horizontal rods could be build in such a way they would not electromagnetically interfere with the waves, unlike in the sketch.
You are right with the length of the wire, this could trigger other modes(field vectors along the wire). But:
1) If the dimensions of the cavity is designed for a given mode shape, that will be the dominant. The whole thing is a resonant system (RCL) with a well defined resonant frequency.
2) With the cavity length and the angle it is possible to manipulate the frequency position in the way to be sure that the other modes are as far away as possible.
3)Each conventional MW source generate the drive frequency and a noise bandwidth around, but most a small one. The used HF components are able to modify that bandwidth(cavity, antenna, antenna feed, filter, ...)
For a "clean" resonance curve other modes have to be as far as possible.
You are also right with the straight stubs to connect the loop (loop feed) they could trigger other mode shapes, if there are possible for the cavity dimensions (see 1) 2). And the stubs will lower the effectivity of the coupling.
I don't know if i would do it this way(big loop).
That's why we need the simulations.On the other hand it may be the right way to design the cone for overlapping mode shapes to generate thrust. Don't know
The important thing to consider is the location of the loop WRT the cavity walls, end plates and mode shape. For example the two horizontal rods could be build in such a way they would not electromagnetically interfere with the waves, unlike in the sketch.
