-
#1040 by rfmwguy on 28 Dec, 2016 00:55
-
I would argue it is bad to have a less sensitive torsion assembly, especially with lower power levels. Rather than dampening by the wire, oil bath and paddle can be added. Higher mass and stronger wire cannot easily have dampening reduced. Best to start out under-dampened then add as needed. Make sense?
That would be a total waste of mass to suggest a ceramic plate as a torsion beam.I would also suggest that using melamine laminated fiberboard as an experimental platform may be problematic. It is extremely moisture and temperature sensitive, and will cold flow under its own weight. Since you have apparently spent tens of thousands of dollars so far, may I suggest an Invar, quartz, or Schott glass/ceramic platform (obtainable from a flat surface cooktop) rather than the material you propose?
This project is supposed to be cheap (about $500) because its goal is an easy EmDrive replication platform. Thus in order to keep the price at its lowest level, what alternative type of wood would you recommend?
For a device to measure micronewton forces, I wouldn't recommend wood of any kind. For those in the US, I would recommend going to the local dump, and snagging a glass surface cooktop. The glass ceramic surface has a coefficient of thermal expansion near zero (better than quartz), and is insensitive to moisture. Carbon fiber extrusions, if properly selected, would be another good choice, but not free of cost unless your dump is better than mine.
Doing good science on the cheap is an art. The art involves a grasp of all of the sciences. Some call this magic. Others call it obvious.
Seal the hardwood beam, isolate it from direct heating and use an ir sensor to monitor beam temp during power on.
Get some positive results? Then move to improving a test stand. Keep mass as low as possible, a thinner torsion wire is always better. Heavier assembly? Thicker wire.
What is meant by "waste of mass"? In the case of a torsion pendulum, increased mass can only slow (provide a first order low pass filter response) to the pendulum. This may be good, it may be bad, depending on the required response. Using a material that changes mass under condtions which are uncontrollable for the home experimenter (wood, or any other material) that expands and contracts under variations in atmospheric pressure, temperature, and humidity, is ludicrous. -
#1041 by rq3 on 28 Dec, 2016 01:02
-
That would be a total waste of mass to suggest a ceramic plate as a torsion beam.I would also suggest that using melamine laminated fiberboard as an experimental platform may be problematic. It is extremely moisture and temperature sensitive, and will cold flow under its own weight. Since you have apparently spent tens of thousands of dollars so far, may I suggest an Invar, quartz, or Schott glass/ceramic platform (obtainable from a flat surface cooktop) rather than the material you propose?
This project is supposed to be cheap (about $500) because its goal is an easy EmDrive replication platform. Thus in order to keep the price at its lowest level, what alternative type of wood would you recommend?
For a device to measure micronewton forces, I wouldn't recommend wood of any kind. For those in the US, I would recommend going to the local dump, and snagging a glass surface cooktop. The glass ceramic surface has a coefficient of thermal expansion near zero (better than quartz), and is insensitive to moisture. Carbon fiber extrusions, if properly selected, would be another good choice, but not free of cost unless your dump is better than mine.
Doing good science on the cheap is an art. The art involves a grasp of all of the sciences. Some call this magic. Others call it obvious.
Seal the hardwood beam, isolate it from direct heating and use an ir sensor to monitor beam temp during power on.
Get some positive results? Then move to improving a test stand. Keep mass as low as possible, a thinner torsion wire is always better. Heavier assembly? Thicker wire.
What is meant by "waste of mass"? In the case of a torsion pendulum, increased mass can only slow (provide a first order low pass filter response) to the pendulum. This may be good, it may be bad, depending on the required response. Using a material that changes mass under condtions which are uncontrollable for the home experimenter (wood, or any other material) that expands and contracts under variations in atmospheric pressure, temperature, and humidity, is ludicrous.
Wood has another benefit as being easy to work with. Screw anything to it? just use a screw. You do not even need to drill a hole first.
Then I'd suggest you make the entire thing out of Play-Do, including the frustum and microwave source. Are you folks after cheap and easy, or are you after replicatable results? Pick one. Real science takes time. Time is money. Science requires money. The corollary that good science is expensive is not necessarily true, but you have to be clever enough to know what to pay for to get results that sceptics will accept.
Using wood on a micronewton torsion pendulum places your results, positive or negative, in the realm of laughable. -
#1042 by rq3 on 28 Dec, 2016 01:18
-
That would be a total waste of mass to suggest a ceramic plate as a torsion beam.I would also suggest that using melamine laminated fiberboard as an experimental platform may be problematic. It is extremely moisture and temperature sensitive, and will cold flow under its own weight. Since you have apparently spent tens of thousands of dollars so far, may I suggest an Invar, quartz, or Schott glass/ceramic platform (obtainable from a flat surface cooktop) rather than the material you propose?
This project is supposed to be cheap (about $500) because its goal is an easy EmDrive replication platform. Thus in order to keep the price at its lowest level, what alternative type of wood would you recommend?
For a device to measure micronewton forces, I wouldn't recommend wood of any kind. For those in the US, I would recommend going to the local dump, and snagging a glass surface cooktop. The glass ceramic surface has a coefficient of thermal expansion near zero (better than quartz), and is insensitive to moisture. Carbon fiber extrusions, if properly selected, would be another good choice, but not free of cost unless your dump is better than mine.
Doing good science on the cheap is an art. The art involves a grasp of all of the sciences. Some call this magic. Others call it obvious.
Seal the hardwood beam, isolate it from direct heating and use an ir sensor to monitor beam temp during power on.
Get some positive results? Then move to improving a test stand. Keep mass as low as possible, a thinner torsion wire is always better. Heavier assembly? Thicker wire.
Isolate it from direct heating? How? And for how long? Monitor its temperature? And then what? Construct an algorithm that "corrects" your data based on a perhaps non-linear thermal response? It sounds like you have lost track of the effect you are trying to measure.
Seal the wood? With what? Are you going to characterise the moisture response of the "sealed" wood? How will you do that? How can I replicate it? Why would I bother, since the errors you impose by your methodology make the entire experiment pointless. I'm trying to measure micronewton forces caused by an unexplainable method. Wood? I don't think so. -
#1043 by rq3 on 28 Dec, 2016 01:24
-
I would argue it is bad to have a less sensitive torsion assembly, especially with lower power levels. Rather than dampening by the wire, oil bath and paddle can be added. Higher mass and stronger wire cannot easily have dampening reduced. Best to start out under-dampened then add as needed. Make sense?
That would be a total waste of mass to suggest a ceramic plate as a torsion beam.I would also suggest that using melamine laminated fiberboard as an experimental platform may be problematic. It is extremely moisture and temperature sensitive, and will cold flow under its own weight. Since you have apparently spent tens of thousands of dollars so far, may I suggest an Invar, quartz, or Schott glass/ceramic platform (obtainable from a flat surface cooktop) rather than the material you propose?
This project is supposed to be cheap (about $500) because its goal is an easy EmDrive replication platform. Thus in order to keep the price at its lowest level, what alternative type of wood would you recommend?
For a device to measure micronewton forces, I wouldn't recommend wood of any kind. For those in the US, I would recommend going to the local dump, and snagging a glass surface cooktop. The glass ceramic surface has a coefficient of thermal expansion near zero (better than quartz), and is insensitive to moisture. Carbon fiber extrusions, if properly selected, would be another good choice, but not free of cost unless your dump is better than mine.
Doing good science on the cheap is an art. The art involves a grasp of all of the sciences. Some call this magic. Others call it obvious.
Seal the hardwood beam, isolate it from direct heating and use an ir sensor to monitor beam temp during power on.
Get some positive results? Then move to improving a test stand. Keep mass as low as possible, a thinner torsion wire is always better. Heavier assembly? Thicker wire.
What is meant by "waste of mass"? In the case of a torsion pendulum, increased mass can only slow (provide a first order low pass filter response) to the pendulum. This may be good, it may be bad, depending on the required response. Using a material that changes mass under condtions which are uncontrollable for the home experimenter (wood, or any other material) that expands and contracts under variations in atmospheric pressure, temperature, and humidity, is ludicrous.
No, it doesn't make sense. Increased mass on a torsion balance influences only the reponse time, not the "sensitivity", unless the expected response is above the bandwidth of the pendulum itself. For a constant thrust Emdrive, the pendulum mass is not relevant. For a pulse thrust Emdrive, the pendulum is critical.
-
#1044 by xyzzy on 28 Dec, 2016 01:41
-
I think the PLL system EW used in their 2016 looks good -- though Mr. March said he needed to tune it constantly. Maybe some improvement over their system can be made.
Well, that was basically my first proposal from a little while ago. In order to use a PLL efficiently, one has to match the phase angle of the PLL output to the resonator standing wave, to keep the power factor close to unity and the power transfer in the right direction. EW did have an adjustment device for that, but it was mechanically tuned and could only be preset in advance, no phase tracking in real time. Getting rid of the mechanical delay line and synthesizing the carrier phase electronically via an I/Q modulator would make this quantity amenable to realtime electronic control. Also using a mode-selective antenna geometry can help prevent the PLL from jumping and locking to the wrong mode. -
#1045 by PotomacNeuron on 28 Dec, 2016 02:08
-
I would argue it is bad to have a less sensitive torsion assembly, especially with lower power levels. Rather than dampening by the wire, oil bath and paddle can be added. Higher mass and stronger wire cannot easily have dampening reduced. Best to start out under-dampened then add as needed. Make sense?
No, it doesn't make sense. Increased mass on a torsion balance influences only the reponse time, not the "sensitivity", unless the expected response is above the bandwidth of the pendulum itself. For a constant thrust Emdrive, the pendulum mass is not relevant. For a pulse thrust Emdrive, the pendulum is critical.
If the torsion balance is heavier, my instinct tells me that it will be less sensitive. Besides being slower, the rotary elastic coefficient of the piano wire will increase (being stiffer). -
#1046 by PotomacNeuron on 28 Dec, 2016 02:15
-
That would be a total waste of mass to suggest a ceramic plate as a torsion beam.
Seal the hardwood beam, isolate it from direct heating and use an ir sensor to monitor beam temp during power on.
Get some positive results? Then move to improving a test stand. Keep mass as low as possible, a thinner torsion wire is always better. Heavier assembly? Thicker wire.
Isolate it from direct heating? How? And for how long? Monitor its temperature? And then what? Construct an algorithm that "corrects" your data based on a perhaps non-linear thermal response? It sounds like you have lost track of the effect you are trying to measure.
Seal the wood? With what? Are you going to characterise the moisture response of the "sealed" wood? How will you do that? How can I replicate it? Why would I bother, since the errors you impose by your methodology make the entire experiment pointless. I'm trying to measure micronewton forces caused by an unexplainable method. Wood? I don't think so.
Wood is good. According to Wikipedia, Cavendish used wooden arms and measured 0.174 micro Newtons.
"The apparatus constructed by Cavendish was a torsion balance made of a six-foot (1.8 m) wooden rod suspended from a wire, ..."
"Cavendish's equipment was remarkably sensitive for its time.[9] The force involved in twisting the torsion balance was very small, 1.74×10^−7 N,..." -
#1047 by xyzzy on 28 Dec, 2016 02:24
-
You describe a microwave carrier with low level amplitude modulation. A good spectrum analyzer will see this already as an artifact of the line powered supply, unless the microwave source is strictly battery powered.
While correct in principle, this should not be a significant practical problem. Of course there will be some AM from a mains powered amplifier, but it will be small and concentrated at low integer multiples of 50 (60) Hz. This won't interfere much. In my proposed resonance tracking system the modulation signal is provided by a dedicated oscillator, one that is under control of the experimenter, therefore supposedly stable and with reasonably precise frequency chosen preferably not in the direct vicinity of 50 or 60 Hz or their strongest harmonics. Because the modulation frequency is known and fixed, the receiver part can have bandpass filters tuned to it, so it can be designed not to let mains interference through to the level meters and the controller. Also the intended amplitude modulation is not really low level. To get something like 5% power into each sidetone, one would need quite a sizeable modulation depth. Likely something on the order of a 30% modulation index. In a well designed experiment, this should be well above any mains induced unintended modulation that usually happens. -
#1048 by WarpTech on 28 Dec, 2016 02:46
-
Isolate it from direct heating? How? And for how long? Monitor its temperature? And then what? Construct an algorithm that "corrects" your data based on a perhaps non-linear thermal response? It sounds like you have lost track of the effect you are trying to measure.
Seal the wood? With what? Are you going to characterise the moisture response of the "sealed" wood? How will you do that? How can I replicate it? Why would I bother, since the errors you impose by your methodology make the entire experiment pointless. I'm trying to measure micronewton forces caused by an unexplainable method. Wood? I don't think so.
I can understand your points, and they are well made. I have built everything from hydroplane racing boats to guitars out of wood. Sealing wood from moister and humidity is no problem. A coat of boat sealer and some varnish will do the trick. There are also modern acrylic resins intended for cars, that would well insulate the wood from high temperatures, but nothing will prevent it from expanding or contracting with temperature.
However, if the point of the test rig is to "rotate AND accelerate", while being video recorded, I defy anyone to demonstrate a wooden pendulum that will do this, no matter how much the temperature fluctuates. In other words, if the EmDrive actually works, then this level of precision is unnecessary. DIY'ers are not really expecting 1 or 2 uN of thrust. They are expecting a great deal more than that. I would be more concerned with it's susceptibility to vortex air currents from heat rising.
-
#1049 by rfmwguy on 28 Dec, 2016 03:25
-
The thrust duration of diy emdrives is far from constant and the torque resistance of thicker wire and higher rotational mass makes no sense when first starting a build. Low inertial resistance is critical in early development stages. A builder should focus on the drive first and then make improvements to a test stand later.
It's what I and other builders have done with limited budgets. Just passing along some practical hands on knowledge. The drive is what I'd focus on initially, make some observational tests on a basic test stand and if positive, make the commitment for a more advanced stand. Ask Paul, shell or Monomorphic. Each has built out test stands only after initial observational tests on simpler stands. -
#1050 by Giovanni DS on 28 Dec, 2016 05:02
-
Using that setup the thermal expansion could not be mistaken for a thrust signature, it will go 90° compared to the expect thrust direction. It will just unbalance the setup on one side.
I would more be concerned about heated air flow, does the laptop have lateral grids? it could create an air flow from the bottom grids to the side grids. A cheap low power MCU would be better, data could be sent wirelessly to a computer. -
#1051 by TheTraveller on 28 Dec, 2016 05:19
-
A more up-to-date setup attached.
You need a rod or wood plank mounted vertically in the middle of the horizontal plank (like a up-side-down T) , and then mount the piano wire on the top tip of the rod or plank. Otherwise the balance is not stable and will flip over.
There will be a wire bridle from both sides of the plank, with the apex above the thruster height. At the apex the piano wire to the roof beam will be attached. Bridle should be approx 300mm above the plank upper surface. Piano wire will be >2m in length. As attached.
Plank bottom will be approx 100mm above the bare brick floor. Will have wooden 100mm blocks at each side to sit the plank on while doing work and freq tuning. Of course removed when doing rotation runs.
I'm building a strip line forward and reflected power monitor that should have almost 0 dB insertion loss and can be left in place during rotation testing:
http://vk5ajl.com/projects/swrmeter.php
If so will add that to the circuit as well as an 4 channel USB scope
http://www.ebay.com/itm/6254BC-4-CH-1GSa-s-250Mhz-Bandwidth-PC-Based-USB-Digital-Storage-Oscilloscope-/291868160217?hash=item43f4b2b0d9:g:xb4AAOSwdIFX0S2r
for the laptop that can display the forward, reflected power as well as the voltage and current being drawn from the battery pack.
The SWR "meter" you quote was built by a guy with minimal understanding of RF engineering. A pure RF short or open will result in an infinite standing wave ratio. Achieving a pure RF short or open while maintaining system impedance is not a trivial exercise. I would also suggest that using melamine laminated fiberboard as an experimental platform may be problematic. It is extremely moisture and temperature sensitive, and will cold flow under its own weight. Since you have apparently spent tens of thousands of dollars so far, may I suggest an Invar, quartz, or Schott glass/ceramic platform (obtainable from a flat surface cooktop) rather than the material you propose?
Rq3,
For sure this is not a lab quality build. But that is not needed. The reflected power sensor will be used with a digital meter and the PC driven freq gen to find the lowest reflected power freq for the rotary test. It will also be used when tuning the antenna position.
Remember the thruster build cost, including the test rig, is limited to $500, so there is a process to go through to find what can be made or sourced from EBay to achieve the 2mN of force at <=$500 cost.
I may provide the reflected power sensor unless I can find one on EBay that can be left in circuit. -
#1052 by TheTraveller on 28 Dec, 2016 05:24
-
That would be a total waste of mass to suggest a ceramic plate as a torsion beam.
Seal the hardwood beam, isolate it from direct heating and use an ir sensor to monitor beam temp during power on.
Get some positive results? Then move to improving a test stand. Keep mass as low as possible, a thinner torsion wire is always better. Heavier assembly? Thicker wire.
Isolate it from direct heating? How? And for how long? Monitor its temperature? And then what? Construct an algorithm that "corrects" your data based on a perhaps non-linear thermal response? It sounds like you have lost track of the effect you are trying to measure.
Seal the wood? With what? Are you going to characterise the moisture response of the "sealed" wood? How will you do that? How can I replicate it? Why would I bother, since the errors you impose by your methodology make the entire experiment pointless. I'm trying to measure micronewton forces caused by an unexplainable method. Wood? I don't think so.
Wood is good. According to Wikipedia, Cavendish used wooden arms and measured 0.174 micro Newtons.
"The apparatus constructed by Cavendish was a torsion balance made of a six-foot (1.8 m) wooden rod suspended from a wire, ..."
"Cavendish's equipment was remarkably sensitive for its time.[9] The force involved in twisting the torsion balance was very small, 1.74×10^−7 N,..."
Expected force is ~2mN. Attached are the rotational data
-
#1053 by TheTraveller on 28 Dec, 2016 05:35
-
Using wood on a micronewton torsion pendulum places your results, positive or negative, in the realm of laughable.
Welcome to the world of DIY garage experimenters.
8W of Rf even if it all escaped the thruster would have little effect on causing rotation of the test rig.
So please explain how any escaped Rf will cause rotation?
Remember there will be 12 experimental test runs, some of which should not produce any rotation. Those run should prove the electronics and the heat they product can't cause rotation. This of course includes the laptop with it's SSD and no fan.
1) CCW both end plate holes plugged
2) CCW big end plate hole unplugged, small plugged
3) CCW small end plate hole unplugged, big plugged
4) CCW both holes unplugged
5) CW both end plate holes plugged
6) CW big end plate hole unplugged, small plugged
7) CW small end plate hole unplugged, big plugged
8) CW both holes unplugged
9) small end pointing up
10) small end pointing down
11) small end pointing to centre
12) small end pointing away from centre
Please check my rotation calculations as it shows the amount of "other" torque that would need to be generated by whatever to cause the same rotation as the force from the thruster.
So please contribute to the discussion but do it while showing how something else on the test rig can generate the effect of 2mN applied 0.5m from the centre of the rotation point. -
#1054 by TheTraveller on 28 Dec, 2016 05:39
-
I can understand your points, and they are well made. I have built everything from hydroplane racing boats to guitars out of wood. Sealing wood from moister and humidity is no problem. A coat of boat sealer and some varnish will do the trick. There are also modern acrylic resins intended for cars, that would well insulate the wood from high temperatures, but nothing will prevent it from expanding or contracting with temperature.
The wooden plank is covered on all 6 sides by white laminex. There is no exposed wood.
-
#1055 by TheTraveller on 28 Dec, 2016 05:45
-
Seal the wood? With what? Are you going to characterise the moisture response of the "sealed" wood? How will you do that? How can I replicate it? Why would I bother, since the errors you impose by your methodology make the entire experiment pointless. I'm trying to measure micronewton forces caused by an unexplainable method. Wood? I don't think so.
Please can we stay on track? The expected force is 2mN of force. The book shelf is covered on all 6 sides by white laminex. There are 12 experimental runs planned that should lay to bed any concerns about other thrust sources being able to generate 2mN of force that acts on the end of a 0.5m lever arm. -
#1056 by TheTraveller on 28 Dec, 2016 05:49
-
What is meant by "waste of mass"? In the case of a torsion pendulum, increased mass can only slow (provide a first order low pass filter response) to the pendulum. This may be good, it may be bad, depending on the required response. Using a material that changes mass under condtions which are uncontrollable for the home experimenter (wood, or any other material) that expands and contracts under variations in atmospheric pressure, temperature, and humidity, is ludicrous.
Idea is to generate at least 180 deg of rotation and maybe many rotations. Depends on the piano wire and how much energy is absorbed as it twists.
Might just throw in a few fishing line ball bearing swivels to allow the piano wire to spin a bit. -
#1057 by TheTraveller on 28 Dec, 2016 05:57
-
The original choice of wood is to avoid any chance of overlooked ground loop. Re-introducing aluminium may lead to the same problem as EW's 2014 paper had : DC ground loop. Wood is not a bad choice.
If wood is preferred, I would go with a reinforced structure instead of a thin plank. The reinforced strucutre will not bend down at the ends due to gravity.
Personally, I would go with an extruded aluminum rectangular bar - which is what I used in my build. A piece long enough will run you $50 at your local metal supplier.
Keep it KISS as a LOT of DIYers are going to be duplicating. The laminex covered 1.2 x 0.2m book shelf and a simple 2 point bridle to the very thin piano wire are all that are required.
Thruster weighs 2kg, laptop about the same. Whole test rig should mass under 8kg and cost under around $20.
Moderators please note every post I have made was in reply to another. -
#1058 by TheTraveller on 28 Dec, 2016 06:02
-
The effect you all are looking for is NOT minimum VSWR for a particular microwave input, its maximum FORCE for a particular microwave input. As I've said ad nauseum, why does no one tune for that? Once the phase locked loop is in place, making it a force locked loop is trivial. Can no-one think outside the box?
1st you tune for min VSWR as without that there is no max force to tune for. The sequence of tuning is important. -
#1059 by Monomorphic on 28 Dec, 2016 10:11
-
Bit of an update on the piano wire & bookshelf torsion balance that will be used with the $500 KISS EmDrive test rig.
What signal generator are you going to use? The inexpensive chinese signal generator has -3dB of output. You will need a pre-amp to drive the 8W amp with 10dB. The windfreak signal generator has enough output to drive the amp by itself, but it is $600. If you have another signal generator in mind, please let me know so I can pick one up!