piezo ceramic project

[hoola]

today I got in the new 10KV transformers and it began with seeing that they were encased in a hard clear thick epoxy, preventing me from altering them in the manner I had considered. They need a large (.47mfd) capacitor across the primary to get any useful current to transfer through them, indicating a much too low primary winding impedance, but they are an improvement over the previous smaller ones. With the addition of the .47mfd cap, current transfer peaks at 170 KHZ, close to the needed range. I will begin to assemble the new system with those, continue to look online for a more suitable transformer, and will explore the idea of winding my own. 

Edited January 8, 2022 by hoola

[hoola]

today I went back to the little transformers and decided to utilize the fact that they are easy to take apart and did so. I removed the primary bobbin and took out the windings, of which there are two in parallel, each only about 3 1/2' long. A second primary winding was of lighter gauge and was used for a feedback line to the spark generator kit the transformer was designed for. I unwrapped a light enamel wire from a small AC motor field coil and wound a single primary of about 40' in length, all I could fit onto the small bobbin and be able to fit it back into the ferrite core. The results are promising, having a good secondary output response throughout the range required and seems to have a good enough input impedance match. I will proceed to re-wind four more of the minis and use them for the next series of tests. The measured DC  resistance of the new primary is 3.9 ohms.

Edited January 8, 2022 by hoola

[hoola]

i must have hit it way lucky with my first try to wind a new primary, as primary #2 one was rather weak in  response test, not responding until driven over 200K.  The #1 primary weighs exactly 3milligrams  (including bobbin which weighs .367 milligrams) with the scotch tape holding the winding to the core. I swapped the #1 primary onto the core and secondary of the #2 transformer, and the results were still positive, eliminating the core or secondary as a failure point on the #2 primary. The #2 primary weighed 3.21 milligrams, so I trimmed off about 5' of wire and got a reading of just under 3 milligrams without the tape. Amazingly, this slight removal of wire has made the installed  #2 primary almost identical to #1 in response in the sweet spot of the  100-200K range.The quick resolve of the problem may have more to do than simple weight (using that as a measure of overall length) such as layering techniques and steadyness of tension, but I appear to have 2 functioning transformers so far.

Edited January 9, 2022 by hoola

[hoola]

Having finished all five transformers, they do stimulate the elements about as well as one would expect, and of inadequate levels.  Given their small size they quickly saturate under load, so I have gone back to direct stimulation from the transistors. I will increase the supply voltage to them from the 700v now used in increments to see how much voltage they can safely handle. In the meantime I am researching transformer winding techniques and will acquire ferrite cores and .1mm wire and begin test winds. I will inquire with the seller on the potted tranformers  to see if  he could request the manufacturer supply the same transformers without completing the final potting step.

[hoola]

 I have considered that a double power supply of both + and -  might prove advantageous. The positive one drives a conventional setup to one side of the stack, with corresponding output transistor from the negative supply driving the other end of the stack. This could deliver a AC signal pulse of an increased voltage as is delivered by the single positive supply only, to it's simple ground reference. Initial tests of the this system seems promising, having used a phase splitter to run a push-pull type drive to the two output transistors. If this system continues to give promise, I will forgo the transformer work for the present and continue to build up the double supply idea. This will require 2 output transistors per channel with an associated phase splitter. The common ground of the stack will have to be eliminated and each piezo will be driven independently by two outputs. This hopefully allow the actual driven signal to  each element to meet or exceed 1200v p-p while keeping the individual output transistors below their 800v p-p limitation. Each driving signal is no larger than a single transistor's output, but due to being driven in overall phase additive, by opposing voltage supplies, there should be a significant additive effect. ideally, a PNP complement of the nte 165 might be used, but there seems no high voltage PNP transistors readily available (hence the need for a phase splitter), so I have used a conventional nte165, by putting the negative supply on the emitter, with the collector to ground with the standard size load resistor of 100k, which seems to work fine in today's inital hookup.

Edited March 3, 2022 by hoola

[hoola]

while in the process of building the phase inverter module, a new 4017 board is also being built. All signal processing and control of the main 125 khz signal originating from the wavetek generator will be in one chassis, which will feed a second  chassis containing the output transistors and related power resistors, itself driven by the dual power supply. The balance arm will be shorter than before, with an overall length of 8" or so as to fit inside a minimum sized vacuum chamber containing the arm only, hooked up by gas tight feed-through capacitors from the driver chassis. If I achieve what appears to be thrust for a second sustained time, the vacuum chamber will be evacuated to falsify or show evidence of any success of the concept.

[hoola]

the dual power supply idea seems more promising as I have been occasionally placing high value resistors across the test stack inputs to gauge the approximate P-P voltage. Up till now a rather weak 1/8" or so of arcing was seen. Upon completing the biasing and drive level adjustments of the phase inverter circuit,  running at 125khz, the test resistor exploded, then the remnants burned and smoked. I see this as a positive sign overall, but may be nearing the limits that the particular piezos i have been using can operate at.

[hoola]

the new 4017 and phase inverter boards are installed and working correctly. The power drive module is completed and affixed to the balance arm support structure, but have not powered it up yet, as the arm is still being rebuilt and I don't want to run the module unloaded. Each piezo end will be independently driven and without the common ground, so ten wires per side. I am using the previous arrangement of the two stacks mounted at arm ends, inverted from one another and being run in parallel. The double HV supply is completed and working well.

Edited March 13, 2022 by hoola

[hoola]

today I got the entire setup powered at 3/4 voltage only, as there is a  chattering feedback in the output module (even with no drive going to it) which increases as voltage goes towards full, and 2 of the outputs are not drawing  proper current, but the ones that are are at the desired temperature. So, no grave errors were found with shorts and obvious wiring errors. I was curious about if I switched the drives around  going into the output module, as one delivers a positive going pulse and the other is it's negative, and i was wondering if the two output stages of the final drive would be better driven by neg sig. input to neg.  outputs and positive to positive, or driving with opposite polarities  I have two DIN plugs coming from the drive mod, so an easy test by a simple swap. I did hear that there was a difference in apparent drive levels, but haven't followed up on that as of yet. I have not hooked the end piezos back up as signal sources to the scopes, and I will go back to trying load down an end piezo with a resistor to remove energy from that end of the stack, therefor reducing the reflection of the phononic waves, analogous to lowering SWR  in a transmitter antenna, perhaps enhancing an energy imbalance within the stack.

Edited March 14, 2022 by hoola

[hoola]

the negative going pulse inputs drive the positively powered outputs more effectively,  and the positives drive the negatives with the same result, not as what i expected

Edited March 15, 2022 by hoola

[hoola]

today. tests ended with 3 seriesed 500V electrolytics in the +power supply shorting soon after an arc occurred. The most DC across them is 1200V, so I think high stack pulses fed back through the output module, and from there reflected back up the supply lines to damage the power supply.  I will install  series inductors on both the positive and negative supply leads as a possible preventative of further problems.  Before the power supply failed, I had figured out the chattering feedback issue, and corrected an oversight in the bias circuitry, so now all the output transistors have an appropriate temperature.

Edited March 15, 2022 by hoola

[hoola]

the power supply has been fixed for some time and the present setup using transistor drivers  instead of the unreliable tubes is better as far as P-P drive to the stack and with no fear of arcing, having only a 2KV max DC potential across any one element. I have occasional readings in the 2-3 milligram range, but nothing remarkable. I will continue to monitor the output daily,  and report anything of interest.

[hoola]

I have been considering the idea of making two more identical devices, running concurrently and adjusted as close as possible with the same drive freqs. and delay intervals. These "twins" would be adjusted for identical scope traces and monitored for any duplications in the weight changes. They would be two single stack arrangements, as the waveforms on the double stacks are much less predictable and less able to remain fixed on a stable set of resonances. This I determine due to the crosstalk between the double stacks setup since they are run in direct parallel, and in non buffered connection. Not only are the traces more unstable, but less "pretty" and don't have the character generator feature of the single stack arrangement.

Edited April 8, 2022 by hoola

[hoola]

 along with possible determination of environmental factors affecting the stacks concurrently, another aspect I have considered is to see if there is any interactions between the stacks as related to their proximity and relative angles, that are not from normal emf or simple acoustic transfers.

Edited April 9, 2022 by hoola

[hoola]

I have determined to a near certainty that the apparent thrust of last November was due to audio ultrasonics emitted from the stacks and not due to a "real" thrust from the mechanism. I do have an idea that stacking the elements directly touching and stimulating with a higher potential, using a two contact arrangement , one on each end. The potential across each element will be divided by the number of  elements, since they will be in a simple series circuit, (connected by their faces) so the needed voltage will be quite high, needing upwards to 3KV across each piezo.  The central nylon rod arrangement will still be used to constrain the stack as before. The physical movement of the end piezos will be the accumulated movement of all of them. I will use a large mass at one end of the stack, as in the Woodward scheme, to cause an imbalance of physical movements in stack ends. So far I have tried one, then two in series with the established voltage used of approx 2KV, which divided within the series arrangement is too low to expect even a false ultrasonic output. To have any chance of achieving any output, real or false, will require reverting back to the tube drive, with the higher voltages that they can handle vs. the horizontal output transistors, and attempt to drive a stack of 3 piezos with a 10Kv p-p signal to attempt to attain 3kv p-p across each element.  I have determined that 3kv is the maximum voltage  across each piezo before arcing will occur.

[hoola]

I have considered that the poly insulators separating the elements within the stack are a limiting factor in transference of shock, and therefore are a limiting factor in it's peak value as the wave travels through it. The washers currently used are a little over 1 mm thick and rather soft in density. It seems that a replacement of the washers with either a more solid substance, such as a ceramic of the same size, or by using thinner poly washers, or perhaps cutting ones from thin mylar sheeting. The thinner washers idea will be tried before proceeding with the direct contact approach as described in the previous posting.

[hoola]

as to how i presume that the observed thrust is from ultrasonics, I simply placed a solid flat surface below the right side stack, and observed the weight gain in the scale positioned below the left stack. But, the first time I did this a few days earlier, i noted a weight decrease, as if the surface was somehow attracting the right stack, and not repelling it as expected. I proceeded to use the existing setup for the "conjoined" dual piezo arrangement and hooked it up to piezo #1 feed line, and turned off the other four. To my surprise, the #1 output was dead, and quickly found that with building the new 4017 board, I had mistakenly switched the firing arrangement. I rewired the board and made the firing order correct, and then place the solid surface under the right side stack, and the weight in the scale at the opposing end, did now increase, as I had originally expected it to do. How the first trial of checking for acoustics could have displayed an attractive force is puzzling. I will reverse the normal scan order to attempt to replicate this attractive behavior the ultrasonics seemed to produce, or find it was a measurement error and was meaningless. I know little of ultrasonics behavior and it perhaps certain phase relations with a nearby surface can cause and attractive force. Anyone with knowledge of this sort behavior being possible, please advise.

[hoola]

 a few comments on the project, since I have falsified any previous scale readings or physical movements of the arm assy. as to having "real" thrust.  One presumed limitation on any positive results seems due to the waveform of the stimulating signal, which is somewhat sineosoidal. That limitation was not intended, but seems inevitable with normal, transistor or tube drivers, plus the inherent inductance of the elements themselves. The signal input is a pure square wave from the 4017,  but with the circuity that I have used,  the intantaneous travel, or wave energy instilled into the elements is limited, and this peak speed limit could be a crucial factor for limiting thrust, simply due to this "rounding" effect. This seems a limiting factor in the devices used by the Woodward team, as they use a transformer drive, which presumably delivers a rather pure sinosoidal drive. Another issue is the materials used, in that a nanomaterial that has a quicker response time to stimulation is needed other than rather heavy piezos, and has a low inductance. The insulating material between elements likewise needs to be a nanomaterial that exibits a "phononic diode" effect, in that it allows physical impulse to transferred in one direction easily, and offer a high impedance in the other, thus limiting back reaction. The question of how it even could work at all is perhaps that  the acceleration of the elements must be fast enough to "interfere" with the virtual particle pairs of space/time, and under such conditions, become a physical medium capable for mass to interact with. I do think I have developed some miniscule actual thrust, but only slightly above the plank limit, as an analogy to waving your hand might create gravity waves, but only just above plank limit.

Edited June 30, 2022 by hoola

[hoola]

I found a site that describes a paper in the August 2020 issue of Science Advances that asserts the successful development of an "acoustic diode". This can be googled at "nonreciprocal surface acoustic wave propagation via magneto-rotation". They specifically describe the material as a direct analogue to the diode in standard electronics, and suggest it's usefulness in medical testing, specifically sonograms. The need to inhibit the reverse direction of the acoustic energies in the woodward thruster mechanism could make this development of primary interest, should it be adaptable to it's particular requirements. The woodward thruster uses a 44khz signal, and that is perhaps in the range used in sonogram scans.

[hoola]

an earlier article dated  October 2010 in     researchgate.net/publication/47535210_An_acoustic_rectifier     states that rectification was observed by "coupling a superlattice with a layer of ultrasound contrast agent microbubble suspension".

Edited August 14, 2022 by hoola

[hoola]

a quick search of the sonogram spectrum states that frequencies in the 2-18 mhz range are normally used.

Edited August 14, 2022 by hoola

[hoola]

While the idea of coming up with an efficient phononic diode is quite beyond mine (or perhaps anyone at this point) to design one specifically adapted to the Woodward device,  it does occur that a half way measure might be tried. The stacks as I have designed them are done so to maximize the transfer the shocks as efficiently as possible, and the simple expedient of simply introducing a resistance to shock transfer might have some positive effect. The underlying  theory is that the shocks are "helped along" by the sequential stimulations, whereas the echos are not as the directional pulse trains are paused after each sequence finished. If the  hard poly washers now used are replaced with  rubber washers and the overall stack torque is somewhat lowered, the inevitable loss of transfer might favor a certain directionality, and produce an inefficient  pseudo-phononic diode effect. It seems interesting now that I do recall Mr. Woodward remarking that no thrust was seen at all until a single rubber washer was introduced into his stacks. It seems possible that by replacing the 5 hard plastic washers which I use to electrically isolate the elements from one another, with viton rubber or some such material, this will preform the duty of electrical insulation and provide the inherent loss of impulse transfer to perhaps work to some advantage in my situation.

[hoola]

just today a youtube video "Mike McCullough - Quantized Inertia & The Horizon Drive - You Tube"   is online that describes his quantized inertia idea, and then covers how it might relate to the Shawyer microwave device, the Woodward mechanism, as well as other ideas that seem to describe propellantless thrust from something as simple as capacitor plates giving a thrust from cathode to anode in a DC circuit. I remember reading something about the cap plates concept back in the 70s, I think in Omni magazine, perhaps in it's "antimatter" section.

[hoola]

a July 24, 2022  youtube video     Roger Shawyer - Em Drive: NASA's "impossible" Fuel-Less Thruster     may be of interest to some.

[hoola]

it has ocurred that the asymmetrical  weights used by the Woodward team at each end of their stacks may have a better alternative. The weights are used, to my understanding, so as to have the piezos something to react against to give directionality to the travelling waves and selective damping. If I have normally not used such things, and have preferred to simply amplify what can be started with the sequential scan method I use, and accept that inherent limitation of having "nothing to push against". I have taken the first end piezo, normally used as a turn around and a scope sensor source, and shorted across it. This I think accomplishes what a passive weight would do, as used elsewhere. As an analogy to a common loud speaker, if you thump it, it rings and can be easily pushed in and out. If you short across the speaker terminals, the cone becomes quite stiff and naturally resists any physical movements. The piezos are like speakers in a sense, so shorting one at the front of the stack seems to offer a "variable mass reactor" acting as a stiff surface for the first active element to react against and damps out any remnant vibrations as the scan proceeds along and as they travel back and forth. The stack end piezo is now the only scope signal source. I have measured the voltage appearing across it, and have found that a 1.2K resistor cuts that voltage in half, so I figure that offers the lowest "SWR" at today's first attempt at the new idea. This is to dump that energy into the resistor so that less is reflected back, hence raising the efficiency of any possible thrust production. The stack setup now is piezo #1 shorted for variable mass reactor,  piezos #2-#6 driven elements, #7 as scope sensor and dummy load.

Edited September 9, 2022 by hoola