-
#3060
by
keithpickering
on 07 Jun, 2016 21:06
-
Optimum frustum geometry from MiHsCUsing Mike McCullough's equations, here are optimum geometries for the big diameter and length of frusta, along with predicted thrust for a device with PQ=5000000. Note that: (1) thrust is strongly dependent on the diameter of the big end, with larger big ends preferred; (2) computations here assume a constant small end diameter of .15 meters in every case, consistent with Roger's .82 rule for a magnetron of 2.45 GHz; (3) the "optimum" length for a given pair of big/small diameters has a flat peak, which means you're going to get about the same results for a lot of lengths near the given length. That allows a designer some latitude, if resonance can be made easier with moderate length changes.
PQ = 5000000
Small diameter: .15 m
| Big diameter (m) | Opt. L (m) | T (mN) |
| .151 | .255 | .17 |
| .2 | .294 | 7.2 |
| .25 | .329 | 12.7 |
| .3 | .36 | 17.2 |
| .35 | .389 | 20.9 |
| .4 | .416 | 24.1 |
| .45 | .441 | 26.8 |
| .5 | .465 | 29.2 |
| .55 | .488 | 31.4 |
I'm also assuming that nobody here would want to build anything much larger than half a meter.
EDIT: Perhaps I should also mention that I've made some slight enhancements to Mike's published equations. Instead of assuming a single photon moving along the axis, I've integrated photons across the entire areas of the big and small ends. It doesn't change the results much, but if what you're getting doesn't match, that's the reason.
-
#3061
by
zen-in
on 07 Jun, 2016 21:30
-
.../...
If you consider the supply is delivering .25 A @ 4kV, a load resistance of 16 kOhms would substitute for the magnetron. If you used 100 X 180 Ohm 2 - 10 Watt metal film resistors all wired in series that should work for short duration tests. Mount them on an FR4 card on standoffs inside a sealed box. A metal box is ok if everything is rigidly mounted and well insulated.
The heating of the lead wires are due to current through them, not tension. So if you want to get the component of measurement (assuming linear additivity of contributing components) due to this thermal effect alone it's best to avoid heating anything else by dumping hundreds of W to some dummy load, because at same dissipated power the dummy load will have a different impact on measurements : we won't know if (how much of) the measure is different because of a thermal effect on the wire or different thermal effect of the charge at the end of the wires.
I think it would be better to mimic the same current pattern (if it is known, even only roughly) in the wires only, by short circuiting the 3 wires (where they normally feed the magnetron) and feeding them by a current generator circuit (in place of feeding them with the normal output of the HV power supply). There is a little complication because there is 3 wires so 3 currents to set but I'm sure this is doable with low power resistors and not involving high tensions.
Magnetron power supply off (not wired), magnetron off (not wired), same wattage dissipated in wires only, check with thermal measurements that wires heating is comparable, read measurements induced by this heating alone. Only constraint, devise and build a specific 3 wires current generator. How much of the kW or so that goes through the wires actually is dissipated in the wires ? A few watts at most : this is the required power for this current generator. If someone can tell approximately the 3 intensities (in A or mA) that go through the 3 wires and orientations with clear labeling, I or anyone knowing Ohm's and Kirchhoff's law can design the circuit powered on any low power source of tension at 12VDC or so (wild guess).
Any force from the magnetron supply wires could be due to electrostatic effects. Twisting two conductors that have an opposite high voltage potential is effectively creating a capacitor. This will make the wires stiffer, resulting in a force as the wires try to adopt a straighter configuration, due to electrostatic attraction of the two oppositely charged wires. The fact there are 3 wires is immaterial in this consideration. There are only 2 electric charges, + and -.
There might also be some magnetic component of this force so maybe testing with just the current as you suggested will isolate that component.
-
#3062
by
SeeShells
on 07 Jun, 2016 21:36
-
.../...
If you consider the supply is delivering .25 A @ 4kV, a load resistance of 16 kOhms would substitute for the magnetron. If you used 100 X 180 Ohm 2 - 10 Watt metal film resistors all wired in series that should work for short duration tests. Mount them on an FR4 card on standoffs inside a sealed box. A metal box is ok if everything is rigidly mounted and well insulated.
The heating of the lead wires are due to current through them, not tension. So if you want to get the component of measurement (assuming linear additivity of contributing components) due to this thermal effect alone it's best to avoid heating anything else by dumping hundreds of W to some dummy load, because at same dissipated power the dummy load will have a different impact on measurements : we won't know if (how much of) the measure is different because of a thermal effect on the wire or different thermal effect of the charge at the end of the wires.
I think it would be better to mimic the same current pattern (if it is known, even only roughly) in the wires only, by short circuiting the 3 wires (where they normally feed the magnetron) and feeding them by a current generator circuit (in place of feeding them with the normal output of the HV power supply). There is a little complication because there is 3 wires so 3 currents to set but I'm sure this is doable with low power resistors and not involving high tensions.
Magnetron power supply off (not wired), magnetron off (not wired), same wattage dissipated in wires only, check with thermal measurements that wires heating is comparable, read measurements induced by this heating alone. Only constraint, devise and build a specific 3 wires current generator. How much of the kW or so that goes through the wires actually is dissipated in the wires ? A few watts at most : this is the required power for this current generator. If someone can tell approximately the 3 intensities (in A or mA) that go through the 3 wires and orientations with clear labeling, I or anyone knowing Ohm's and Kirchhoff's law can design the circuit powered on any low power source of tension at 12VDC or so (wild guess).
Good post but I hope you don't mind if I add a little video to it.
Shell
Good video.
-
#3063
by
frobnicat
on 07 Jun, 2016 22:58
-
Any force from the magnetron supply wires could be due to electrostatic effects. Twisting two conductors that have an opposite high voltage potential is effectively creating a capacitor. This will make the wires stiffer, resulting in a force as the wires try to adopt a straighter configuration, due to electrostatic attraction of the two oppositely charged wires. The fact there are 3 wires is immaterial in this consideration. There are only 2 electric charges, + and -.
True, I forgot the electrostatic factor that might be significant at those voltages and low measured forces... more used to electromagnetic actuators than electrostatic ones.
There might also be some magnetic component of this force so maybe testing with just the current as you suggested will isolate that component.
Magnetic component (that is clearly visible when playing with loosely wound wires and amperes) and thermal expansion of wires and change in stiffness of the plastic insulator (that also will change force when there is remaining stress already present in the bending and twisting between fixed part and pendulum's arm). So my proposition is still good for these 3 aspects and is probably cheap and certainly safe. Intuitively I'd say that thermal/mechanical aspects (expansion and stiffness change) are prevailing over both electromagnetic and electrostatic, but anyway it would be good to know if thermal/mechanical + electromagnetic (electrostatic aside) is already in the ballpark of what is measured or somehow below. If it's below, electrostatic can be further evaluated separately by leaving the connexions open (magnetron side) and using the magnetron's generator to reach usual operating tensions (probably a bit higher because unloaded but still in the ballpark). That would give us a separate evaluation for the order of magnitude of systematic error due to thermal/mechanical+electromagnetic and for electrostatic.
Using a dummy load electrically equivalent to the magnetron would have the advantage of operating the wires in same conditions both in tension and current at the same time, but to repeat : we won't know if (how much of) the measure is different because of a thermal (or magnetic or electrostatic) effect on the wire or different thermal effect of the charge at the end of the wires, the dummy load would certainly have a different way of generating thermal convection than a powered magnetron + fed frustum, so this would not really isolate a given set of error source with certainty.
-
#3064
by
jay343
on 08 Jun, 2016 00:25
-
Hello everyone! I've been monitoring this topic with utter fascination, and I have a couple questions for you...
Could a microwave dummy load be connected inside the frustum, in such a way that the heat it generates is conducted to the surface via heat sink fins, to produce a thermal pattern on the outside that is similar to the pattern generated when the frustum is operated normally? (if you can call this normal
) If so, then maybe you could produce some tests that isolate only the effect of the microwave energy itself?
-
#3065
by
dblack
on 08 Jun, 2016 00:40
-
Couldn't the overall effect of 'wire stiffening' on the measurement be identified by leaving the wires exactly where they are, but turning the frustrum 180 degrees, and re-measuring?
-
#3066
by
jay343
on 08 Jun, 2016 00:50
-
Has anyone considered using a klystron or a TWT to produce the microwave energy? I think these can be found on the surplus market relatively cheap. Together with a microwave signal generator, you could fully explore the frequency/thrust relationship without the various resonate modes of the frustum influencing the stability of the microwave power generator.
-
#3067
by
SeeShells
on 08 Jun, 2016 01:35
-
I think monomorphic had a great idea by moving the wires to the other side of the top support bar and connecting the the large end of the frustum then up to the magnetron. If thermal issues exist on the frustum magnetron wires they will move the frustum the other direction.
Shell
-
#3068
by
mikegem
on 08 Jun, 2016 07:49
-
I want to build a high voltage dummy load that I can swap out the frustum for. My magnetron is 600W
Would a couple of these in parallel work? http://www.digikey.com/product-detail/en/FVE030020E10R0KE/FVE300-10-ND
Re-reading your post this morning, I see you may have been asking about a dummy load for the magnetron HV power supply, not the magnetron output. Sorry if I missed your meaning.
If you want a dummy load just for the HV supply, the wire wound resistors will be fine at the supply's 120Hz half wave output. The only other question is resistor voltage rating with respect to interwinding breakdown. I'd experiment. Put them inside a grounded metal perf box, energize them and look for arcs/smoke, etc. The pic in the data sheet looks like a coarse winding pitch, meaning there's significant insulation between adjacent winds. That's good for preventing breakdown.
Maybe I'm missing something here but the wires to the magnetron that are twisted together that he wants to measure are ~4,000V.
The resistors will make magic smoke if you put them in this line to replace the magnetron. 
Shell
I figured he could use a high enough value of resistor by connecting several in series to get the load he needs, which would draw 600W from the 4kV HV supply (assuming 100% mag efficiency).
600W would be 0.15 Amps at 4kV. The total resistance needed would be ~ 27K. He could get acceptably close by series-connecting (6) 4.7KΩ 100W resistors. And fan cooling them. Each resistor would see less than a 1kV drop, and they'd be operating within their rated power dissipation. They'd be OK if actively cooled.
-
#3069
by
Mulletron
on 08 Jun, 2016 10:57
-
-
#3070
by
R.W. Keyes
on 08 Jun, 2016 12:04
-
Things were going well until somehow it turned into a discussion of a Creator and consciousness 
http://www.akamaiuniversity.us/PJST11_1_173.pdf
Ah yes indeed I think this creator/consciousness would have been best left for a separate paper.
It really put my sleep-deprived mind off-kilter.
-
#3071
by
RERT
on 08 Jun, 2016 12:51
-
Couldn't the overall effect of 'wire stiffening' on the measurement be identified by leaving the wires exactly where they are, but turning the frustrum 180 degrees, and re-measuring?
Three cheers for this: seems much easier than making a dummy load which will accurately mimic the thermal and power cable effects of the frustrum without introducing RF into it.
R.
-
#3072
by
SeeShells
on 08 Jun, 2016 14:51
-
There are many ways to define anomalous forces. One is to rotate the drive 90
O onto the center line of the beam and the wires under the beam to the magnetron.
Maybe off line today as we are losing power while they upgrade our local power lines.
Shell
-
#3073
by
venir
on 08 Jun, 2016 15:29
-
-
#3074
by
FattyLumpkin
on 08 Jun, 2016 22:32
-
Monomorphic, X_Ray---A question for you (if you would be so kind): am still not sure about geometry for given shapes of CRCs: Monomorphic you indicated that a great number of triangles were required to create the 1st below. ?: how do you know the geometry for the cavity is "correct"? I recognize that the "design" appears to be that of 2nd and 3rd below, but their (Cannae's) papers and patents display a somewhat different geometry 4th below. While I understand that Cannae purports to "create" a "bias" in the QV, I don't understand how this might be achieved by geometries that are nearly 100% symmetrical. Would you please comment and if possible perform a sim of shape #4. I can calculate dimensions if you need them. (Additionally, the images 1-3 do not appear to differ from RC's associated with a particle accelerators)! Flummoxed FL
-
#3075
by
X_RaY
on 09 Jun, 2016 05:36
-
Monomorphic, X_Ray---A question for you (if you would be so kind): am still not sure about geometry for given shapes of CRCs: Monomorphic you indicated that a great number of triangles were required to create the 1st below. ?: how do you know the geometry for the cavity is "correct"? I recognize that the "design" appears to be that of 2nd and 3rd below, but their (Cannae's) papers and patents display a somewhat different geometry 4th below. While I understand that Cannae purports to "create" a "bias" in the QV, I don't understand how this might be achieved by geometries that are nearly 100% symmetrical. Would you please comment and if possible perform a sim of shape #4. I can calculate dimensions if you need them. (Additionally, the images 1-3 do not appear to differ from RC's associated with a particle accelerators)! Flummoxed FL
As stated in the NASA-EW paper they have tested two of these cavities. One of them as an empty dummy the other one with the slots inside at one sidewall. As far as I remember, the result was that thrust signal was dependent on the PTFE filled section of there antenna feed.
For me it looks like Monomorphic had focused on the not slotted version, maybe to take a look to the field patterns inside of this resonator type in general.
Can't tell anything about the overall dimensions used by Monomorphic.
-
#3076
by
SeeShells
on 09 Jun, 2016 05:52
-
New Electronics for Drive, currently in build.
Off to bed.
Shell
-
#3077
by
FattyLumpkin
on 09 Jun, 2016 06:53
-
X_Ray, I am going to have to go back and check, I don't remember a non or un-slotted version, but I do remember that they reversed the direction and still tested some "thrust".
The question still remains however, that the images EWL produced/provided lacks nearly any asymmetry whatsoever IMO, (image 1) indeed minimal at best, where Cannae's purported internal shape/geometry is quite different than it's shape on the outside.(image 2) I'd point out again that that what EWL published for it's sim of Cannae shows little or no apparent geometric asymmetry, and looks like the resonant cavity(s) of a particle accelerator. (image 3)
This point is difficult to make, and while I agree that a certain amount of subtlety is involved in these concepts, I would suggest something seems quite "wrong" where the aforementioned representation is concerned.
Addendum: OK , went back and checked: indeed both "slotted" and "un-slotted" were tested , and in both directions. But where is the circumferentially located cavity in the sim? See image following post
-
#3078
by
FattyLumpkin
on 09 Jun, 2016 07:10
-
-
#3079
by
Mulletron
on 09 Jun, 2016 12:18
-
This reference is an analysis of the two below:
www.physics.princeton.edu/~mcdonald/examples/tuval.pdfTime saver:
Hence, there can be no momentum gain for the case of a time-dependent circuit plus a permanent magnet, contrary to the claim in [2].
There can be a gain of mechanical momentum in the case of two time-dependent circuits,
each of whose time dependence has nonzero odd-order time derivatives, of order 3 or higher.
Yet, we expect by conservation of momentum that a system initial at rest cannot spontaneously
take on momentum. Indeed, we expect that the gain of mechanical momentum of
the system is offset by an equal and opposite electromagnetic field momentum. In particular,
we note that a system of two circuits can be regarded as a pair of magnetic dipoles, which
form a magnetic quadrupole with a nonzero, time-dependent quadrupole moment. As noted in sec. 71 of [49], a quadrupole moment emits radiation if its 3rd time derivative is nonzero.
Hence, we learn that if the 3rd-order integral is nonzero in the expression (17) for the gain in
mechanical momentum by the system, then this momentum is compensated by the emission
of momentum in the form of quadrupole radiation (and if the nth-order integral is nonzero,
then there exists nth-order multipole radiation of momentum).
Of course, as the system emits momentum in its multipole radiation, it also emits energy,
such that only a finite amount of energy and momentum can be radiated. Hence, the gain of
mechanical momentum by the system is finite, just as in rocket propulsion based on chemical
reactions.
In sum, the system of time-dependent circuits can serve as an “electromagnetic spaceship,”
with (weak) propulsion via the reaction to the radiation of momentum due to the
time-varying quadrupole (or higher) moment of the system).
http://arxiv.org/find/physics/1/au:+Tuval_M/0/1/0/all/0/1
) If so, then maybe you could produce some tests that isolate only the effect of the microwave energy itself?
It really put my sleep-deprived mind off-kilter.