Here's the video of the flight test:
Definitely not Hollywood...take it easy on me 
Congratulations on an excellent test !
The setup seems very stable and if you by varying some parameters later should get a tiny thust I think the quality of your build will make it both easy to see it and also to reproduce it reliably.
Wonderful !
Here's the video of the flight test:
Definitely not Hollywood...take it easy on me
I'm not convinced you have a totally null result.
If you compare the frames at timestamp 2:50 (no weight), 3:35 (calibration weight) and 10:49 (100% power) and just flip between the frames, there is displacement, or at least my feeble eyes delude me so.
Whether that's the device or some kind of systematic drift is a different issue, but the measurement points are not static. The laser shows drift while the rest of the image elements are unchanged in position. Try flipping between those times while paused.
NSF-1701 update - Test complete! Myself and the whole assembly came through without a scratch. I started this project with the commitment to report on exactly what I measured. What I measured was no apparent movement of the fulcrum in response to the magnetron being powered on. However, my ability to measure below about 20 mg of gram-force is just not there.
Therefore, in the configuration of the frustum with power insertion from the small base, 3.5 cm from the side wall; the result is Null to the best of my knowledge.
The video is still uploading and will be for some time. I'll post a link when it becomes available. Those of you who may want to download it and zoom in tightly on the laser target for further analysis are welcomed to do so. Power on is easily noted by transformer hum.
Congratulations on your efforts.
Did you get my PM?
As I said, the thing is that the galinstan is a heavy alloy, 6 times heavier than water, so you need a bit more thrust to overcome the buoyancy force. Those 3 wires are quite thick so to avoid oscillating or the frustum standing still, you need a several dozen milinewtons more. In addition the emdrive is ballooning.
So my recommendation is to place the magnetron close to the big end or at the center of it in order to get more unbalanced force and/or turn the cavity upside down.
Best of luck.
Peter.
....
Ok then. I may be mistaken because I'm representing in my head the loop antenna as a flat solenoid. A solenoid there would trigger an axial magnetic field as you say, but would tighten the magnetic lobes towards the axis of axi-symmetry because of the return path of the magnetic field lines around the coil. An antenna may behave differently.
...
And as X_Ray and cee said, a cavity resonates in certain mode shapes due to its dimensions, which should help stabilize the desired mode.
Your picture shows magnetic lobes that vary in the azimuthal direction.
What magnetic toroidal lobes in TE012 
?:
The magnetic vector component in the azimuthal direction is zero for TE012.
The magnetic vector component in the polar angle theta direction is constant in the azimuthal direction for TE012, it varies in the theta and longitudinal direction. It does NOT vary in the azimuthal direction. No magnetic toroidal variation in the azimuthal direction.
It is my understanding that the blue lines in Eagleworks pictures I've posted are magnetic field lines. Eagleworks even draw vectors on the right of the drawing, showing the magnetic field is swirling from one direction along the axis of axi-symmetry (where the magnetic field is the strongest) with looping return paths near the walls of the cavity in the other direction (where the magnetic field is weaker). Directions are inverted for the lobes near the other end. Are we talking about the same thing?
The picture below is from Eagleworks. I added the captions about magnetic fields.
EDIT: Good Lord, when I wrote "toroidal" I really wanted to write poloidal field lines (around a torus shape) and not toroidal field lines (in the azimuthal direction, there is no magnetic field in that direction of course) 
I edited the drawing in my prior post.
Thank you for your patience. Yes, now I see what you drew, I was thinking of the word "toroidal" so I saw them in the azimuthal direction. Now I understand that what you are drawing are the contours shown in the plots below:
Please take a close look at the contours and vector fields. What we want to excite is the magnetic longitudinal field. They have different nodes and antinodes than the magnetic field in the theta polar direction !
The anti-node (the maximum) of what we want to excite: the magnetic field in the longitudinal direction, corresponds to the node (the zero) of the magnetic field in the polar theta direction.
Please take a look at the red section of the longitudinal vector field: the antenna loop can have a very small perimeter indeed as long as it is located at the center of the axis of axi-symmetry, at the correct location
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?
This?
What about placing a
polarizing metal grid inside the cavity? Would that, depending on the orientation, not enhance a TE mode? You'd probably got some losses in average Q, but a much cleaner resonance mode?
Um, looks like the force (if any) was toward the large base (i.e up). Somebody with some batter software than snipit and GIMP want to do a better screen capture and put some lines on this thing?
Here's the video of the flight test:
Definitely not Hollywood...take it easy on me
First, congrats on the experimentation. But I'm a bit confused:
Everyone, including yourself, have been saying this was a null result. When I heard this, I was a bit disappointed and watched the video anyways. It is clear, when you skip through the test[from start to near-end] that the beam has moved. Whether or not that means there was thrust- who knows. But it is, indeed, moving, so something has to be causing that.
I think the apparatus moved up since the beam went down?
NSF-1701 update - Test complete! Myself and the whole assembly came through without a scratch. I started this project with the commitment to report on exactly what I measured. What I measured was no apparent movement of the fulcrum in response to the magnetron being powered on. However, my ability to measure below about 20 mg of gram-force is just not there.
Therefore, in the configuration of the frustum with power insertion from the small base, 3.5 cm from the side wall; the result is Null to the best of my knowledge.
The video is still uploading and will be for some time. I'll post a link when it becomes available. Those of you who may want to download it and zoom in tightly on the laser target for further analysis are welcomed to do so. Power on is easily noted by transformer hum.
With the small end down wouldn't any thrust be acting
against thermal buoyancy? Would have more confidence in the apparent null result after the same test with small end up. If still null then the effectiveness of the wire mesh against a balloon type effect would be further demonstrated. Greater displacement shown with small end up would be very valuable data. Tajmar registered upward movement in both up and down orientations and had to take the difference, if my memory is correct.
Um, looks like the force (if any) was toward the large base (i.e up). Somebody with some batter software than snipit and GIMP want to do a better screen capture and put some lines on this thing?
Enhance.... enhance more, zoom, increase contrast. It is definitely positional change.
Out to the shop again....
Um, looks like the force (if any) was toward the large base (i.e up). Somebody with some batter software than snipit and GIMP want to do a better screen capture and put some lines on this thing?
For comparison, could someone do the same with the before, during and after positions of the laser light when he added the 200mg (I think) calibration weight to the frustrum before he ran his test?
Um, looks like the force (if any) was toward the large base (i.e up). Somebody with some batter software than snipit and GIMP want to do a better screen capture and put some lines on this thing?
For comparison, could someone do the same with the before, during and after positions of the laser light when he added the 200mg (I think) calibration weight to the frustrum before he ran his test?
Can you give me the time stamps on that?
Um, looks like the force (if any) was toward the large base (i.e up). Somebody with some batter software than snipit and GIMP want to do a better screen capture and put some lines on this thing?
Enhance.... enhance more, zoom, increase contrast. It is definitely positional change.
Out to the shop again....
Those frames are at 2:50, 3:35, and 10:49. Positional changes like that are expected with rfmwguy moving around the lab.
Frame grabs need to be done as the power cycles during a test run. The first run with the 30% power cycle over five minutes would be best, since the rig needs time to settle when rfmwguy starts it and leaves the room. Wait until the first minute into the run and then grab frames. You can hear the power cycle on and off.
Staring at the screen I didn't notice any movement.
Um, looks like the force (if any) was toward the large base (i.e up). Somebody with some batter software than snipit and GIMP want to do a better screen capture and put some lines on this thing?
Enhance.... enhance more, zoom, increase contrast. It is definitely positional change.
Out to the shop again....
Those frames are at 2:50, 3:35, and 10:49. Positional changes like that are expected with rfmwguy moving around the lab.
Frame grabs need to be done as the power cycles during a test run. The first run with the 30% power cycle over five minutes would be best, since the rig needs time to settle when rfmwguy starts it and leaves the room. Wait until the first minute into the run and then grab frames. You can hear the power cycle on and off.
Staring at the screen I didn't notice any movement.
I found before to be anytime before 3:00. During the test I see around 3:24 (or you 3:35 is fine) but I would take the "after" as being around 4:00 - NOT the 10:49, which is after all the tests. Okay, maybe include that, too.
My goal is to see the deflection due strictly to the calibration test - before adding the weight, while the weight is added, and after removing the weight. The final position should be the same as the beginning position of the powered runs.
Um, looks like the force (if any) was toward the large base (i.e up). Somebody with some batter software than snipit and GIMP want to do a better screen capture and put some lines on this thing?
Enhance.... enhance more, zoom, increase contrast. It is definitely positional change.
Out to the shop again....
Those frames are at 2:50, 3:35, and 10:49. Positional changes like that are expected with rfmwguy moving around the lab.
Frame grabs need to be done as the power cycles during a test run. The first run with the 30% power cycle over five minutes would be best, since the rig needs time to settle when rfmwguy starts it and leaves the room. Wait until the first minute into the run and then grab frames. You can hear the power cycle on and off.
Staring at the screen I didn't notice any movement.
I found before to be anytime before 3:00. During the test I see around 3:24 (or you 3:35 is fine) but I would take the "after" as being around 4:00 - NOT the 10:49, which is after all the tests. Okay, maybe include that, too.
My goal is to see the deflection due strictly to the calibration test - before adding the weight, while the weight is added, and after removing the weight. The final position should be the same as the beginning position of the powered runs.
Added a bar using gimps rectangle tool. It looks like the rig returned to roughly its initial state after the calibration test. Can anyone work out the downward deflection? My gut it telling me that it's more than 200mg of force.
When I designed NSF-1701, I built it to be reconfigurable.
So I did just that...Doc, here's the new setup with the magnetron centered on the big base as you recommended. Its all ready for a second flight test!
But, its been a big day and I need a break form the shop. I'll test it again in a couple of days.
A heartfelt thanks for all the support!
...
EDIT: Good Lord, when I wrote "toroidal" I really wanted to write poloidal field lines (around a torus shape) and not toroidal field lines (in the azimuthal direction, there is no magnetic field in that direction of course)
I edited the drawing in my prior post.
I knew what you meant. What @Rodal calls azimuthal, I refer to as "circular", when the flux lines are circling the minor diameter, I call it "toroidal" like you did. The word "poloidal" was not in my vocabulary.
Thanks!
Todd
NSF-1701 update - Test complete! Myself and the whole assembly came through without a scratch. I started this project with the commitment to report on exactly what I measured. What I measured was no apparent movement of the fulcrum in response to the magnetron being powered on. However, my ability to measure below about 20 mg of gram-force is just not there.
Therefore, in the configuration of the frustum with power insertion from the small base, 3.5 cm from the side wall; the result is Null to the best of my knowledge.
The video is still uploading and will be for some time. I'll post a link when it becomes available. Those of you who may want to download it and zoom in tightly on the laser target for further analysis are welcomed to do so. Power on is easily noted by transformer hum.
Congratulations on your efforts.
Did you get my PM?
As I said, the thing is that the galinstan is a heavy alloy, 6 times heavier than water, so you need a bit more thrust to overcome the buoyancy force. Those 3 wires are quite thick so to avoid oscillating or the frustum standing still, you need a several dozen milinewtons more. In addition the emdrive is ballooning.
So my recommendation is to place the magnetron close to the big end or at the center of it in order to get more unbalanced force and/or turn the cavity upside down.
Best of luck.
Peter.
Thanks peter, yes I got the pm but was too busy setting up the flight test. Good comments. Yes galinstan is heavier than I would like an believe it dampens small movements and surface tension is an undesireable effect. Wish I had some other liquid to conduct the supply voltages.
Um, looks like the force (if any) was toward the large base (i.e up). Somebody with some batter software than snipit and GIMP want to do a better screen capture and put some lines on this thing?
Enhance.... enhance more, zoom, increase contrast. It is definitely positional change.
Out to the shop again....
Those frames are at 2:50, 3:35, and 10:49. Positional changes like that are expected with rfmwguy moving around the lab.
Frame grabs need to be done as the power cycles during a test run. The first run with the 30% power cycle over five minutes would be best, since the rig needs time to settle when rfmwguy starts it and leaves the room. Wait until the first minute into the run and then grab frames. You can hear the power cycle on and off.
Staring at the screen I didn't notice any movement.
I grabbed three frames during the five minute run when the power was off (5:25, 6:00, and 6:25) and three frames when the power was on (5:45, 6:10, and 6:35). Then I selected a point in the pattern to measure the pixel location. I looked at each frame at 500% so I could easily see the pixels. Each one is at the same location within +/- 1 pixel. There isn't any appreciable movement at the level of resolution of this experiment.
It is a null result.
The stability of rfmwguy's setup is impressive. This is good clean data that sets an upper limit to any thrust made by NSF-1701. Of course, the antenna location may not be optimal, but Dr. Rodal's suggested follow on test will see if the meep analysis helps.
When I designed NSF-1701, I built it to be reconfigurable.
So I did just that...Doc, here's the new setup with the magnetron centered on the big base as you recommended. Its all ready for a second flight test!
But, its been a big day and I need a break form the shop. I'll test it again in a couple of days.
A heartfelt thanks for all the support!
Now that's what I like! An engineer who's doesn't waste any time! Excellent!
Todd
Thanks peter, yes I got the pm but was too busy setting up the flight test. Good comments. Yes galinstan is heavier than I would like an believe it dampens small movements and surface tension is an undesireable effect. Wish I had some other liquid to conduct the supply voltages.
Of course, it's current-carrying capability that's at stake here. I don't think these are too severe - maybe 10 Amps for the heater?
You might want to experiment with common low-density electrolytes - like salt water.
?:
