mach thrusters

Discuss life, the universe, and everything with other members of this site. Get to know your fellow polywell enthusiasts.

Moderators: tonybarry, MSimon

kunkmiester
Posts: 873
Joined: Thu Mar 12, 2009 3:51 pm
Contact:

mach thrusters

Postby kunkmiester » Mon Dec 28, 2009 4:15 am

http://www.cphonx.net/weffect/alt.php
A bunch of resources there, including a power point talking about how to make a "flux capacitor."

Some of it is useful, like PPTs made for talking to laymen, so they're not as complicated and skip some of the math. Wish they had some notes though.

I did finally find some useful stuff on the actual mechanics--how one is built, and expected to work, without diving into the physics. It doesn't answer all of my questions though.

If I build a "flux capacitor" as described in the power point of that name, how do I drive it? I'm guessing it can be scaled up to some extent, to possibly produce more power, but that does little good if I can't run it.

Is there something that will explain how to build the amp system and controls in a way that an electrical engineering student can build it?
Evil is evil, no matter how small

kunkmiester
Posts: 873
Joined: Thu Mar 12, 2009 3:51 pm
Contact:

Postby kunkmiester » Sun Dec 12, 2010 5:49 am

Thought I'd poke this, since there's a bit more activity on the ME thread. It's still something that interests me though there may be a few months before I have spare cash to spend on it. There's a fair amount about how MLTs work, and documentation on the experiments done by others, but info useful for a science fair type project is hard to pick out.
Evil is evil, no matter how small

djolds1
Posts: 1296
Joined: Fri Jul 13, 2007 8:03 am

Postby djolds1 » Tue Dec 14, 2010 2:35 am

kunkmiester wrote:Thought I'd poke this, since there's a bit more activity on the ME thread. It's still something that interests me though there may be a few months before I have spare cash to spend on it. There's a fair amount about how MLTs work, and documentation on the experiments done by others, but info useful for a science fair type project is hard to pick out.
You might consider working at the margins, instead of immediately diving into the core experiments. IIRC, Paul March commented sometime back that the capacitors he & Woodward were using were commercial, and not optimal for M-E experimentation.
Vae Victis

Diogenes
Posts: 6958
Joined: Mon Jun 15, 2009 3:33 pm

Postby Diogenes » Wed Dec 15, 2010 9:47 pm

djolds1 wrote:
kunkmiester wrote:Thought I'd poke this, since there's a bit more activity on the ME thread. It's still something that interests me though there may be a few months before I have spare cash to spend on it. There's a fair amount about how MLTs work, and documentation on the experiments done by others, but info useful for a science fair type project is hard to pick out.
You might consider working at the margins, instead of immediately diving into the core experiments. IIRC, Paul March commented sometime back that the capacitors he & Woodward were using were commercial, and not optimal for M-E experimentation.


This is what I was thinking. The two most knowledgeable and professional people of which I know who are working on this, were having a difficult time getting appropriate dielectric material. What are the chances that a johnny come lately could resolve this problem if the masters can't?

While we're on the topic, I recall someone saying they could get some exceptionally high quality capacitor material for Paul March to work with. I wonder what ever happened with that?

krenshala
Posts: 913
Joined: Wed Jul 16, 2008 4:20 pm
Location: Austin, TX, NorAm, Sol III

Postby krenshala » Wed Dec 15, 2010 11:42 pm

Well, I've seen different people that didn't know each other where one had something the other needed. Let the right person know what you are looking for and you might just "suddenly" find the appropriate material, even if you have had incredibly poor luck sourcing it so far.

I'm not saying that is the case here, but you can't rule out the possibility just because someone happens to be the new guy. ;)

GIThruster
Posts: 4686
Joined: Tue May 25, 2010 8:17 pm

Postby GIThruster » Sun Dec 19, 2010 9:05 am

Kunk, the questions you're asking have a series of complex answers. For instance, the ppt on MLT construction is very good. I built a series of about 8 for Jim Woodward a couple years ago. The trouble with them is that at the time that thruster was designed, we did not appreciate the importance in generating "bulk acceleration" in the dielectric. We were just accelerating the mobile ion in the BaTiO3 rather than the entire lattice, which is what's necessary. That's why that design produced such disappointing results (although it did produce thrust more than an order magnitude above the noise floor of the ARC Lite it was tested on.)

Long story short, if you want your time spent to be well rewarded, you probably don't want to use an old design. There are however many other possibilities.

You can write Paul March, tell him your skill level and ask him how he thinks you can contribute, what sort of project he'd recommend, etc. For myself, just guessing I'd say either build an MLT and spin it in order to generate the bulk acceleration, or build an MLT with a highly piezoactive dielectric like PZT. Paul is favoring the former. I'm favoring the latter just because the bulk accelerations are much higher, in the tens of millions of gees, using PZT.

Another option is build a UFG instead. That's simpler in many ways. Fact is, I would work with any EE student if he thinks he can build the power system to drive my UFG Thin design. Just saying though, it's a real bear because I have this custom dielectric from Japan with a perfect form factor, a k of 30,000 and that has been characterized as having a flat response to 500 Mhz. Trouble is, the wafer has such fangdangeous capacitance and only takes 50V, that is it VERY hard to drive. I was working with a PhD EE on this until he found he just does not have the time to build the power system the way he thought he would. However, I still have the uber-dielectric, the PZT discs it is intended to be sandwiched between, etc. and my ex-partner in this still has my Mettler H20. Just needs serious EE work to complete the project. . .probably is much more complex than a student can manage. It was a challenge for an Ivy League PhD at the top of his field.

FYI, building the thrusters is always easier than building the power system to run them. Odd as it may seem, the biggest challenges are in the standard EE work.

Another good option might be to do a replication of the rotator experiment. That experiment demonstrates M-E without generating thrust, so in it you avoid all the difficulties of having a thrust stand, vacuum, etc. It's relatively easy compared to building and testing a thruster. In like fashion, you might choose to be the first to build a GravInertial Levitation Chamber (GILC). Like the rotator, it demonstrates M-E without having to take measurements with a thrust stand. It does however use vacuum, just inside a test tube.

There are plenty of other options. Don't run out and buy an expensive dielectric. My shop is full of hundreds of the caps used over the years and I'd be happy to send you some if you decide that's what you want to work with. Shoot me a private note, or one to Paul, and we'll see how we can get you connected. If you want to work on an MLT, best first choice is certainly Paul. And as his student in ME work the last half decade I can say without qualification that he is an excellent teacher.
"Courage is not just a virtue, but the form of every virtue at the testing point." C. S. Lewis

GIThruster
Posts: 4686
Joined: Tue May 25, 2010 8:17 pm

Postby GIThruster » Sun Dec 19, 2010 9:44 am

Diogenes wrote:While we're on the topic, I recall someone saying they could get some exceptionally high quality capacitor material for Paul March to work with. I wonder what ever happened with that?


Paul already as a very low loss dielectric built inside an MLT. He just needs the time to run it. Since he's leaving his job at Johnson Space Center in a couple weeks, I expect he'll find the time soon. If he gets results or not, I think his plan is not only to run his MLT, but to set it up to spin in order to get bulk acceleration. (His current project design dates back before we recognized the import of bulk acceleration as mentioned in my post directly above.)

There are a handful of uber-dielectrics I've found over the last couple years. Truthfully, the biggest impediment to the work right now is skilled EE labor. There are lots of designs, dielectrics and options on offer for anyone who has the skill and time to spend.
"Courage is not just a virtue, but the form of every virtue at the testing point." C. S. Lewis

TDPerk
Posts: 976
Joined: Mon Jul 30, 2007 12:55 pm
Location: Northern Shen. Valley, VA
Contact:

Postby TDPerk » Sun Dec 19, 2010 12:25 pm

GIThruster wrote:Odd as it may seem, the biggest challenges are in the standard EE work.


You have a frequency domain characterization of the waveform (voltage and current) to be presented at the capacitor leads, and know what freq (freq range) you want to drive it at?

Oh, if pulsed is okay, what cycle time and pulse width do you need? Or do you need continuous?
molon labe
montani semper liberi
para fides paternae patria

Diogenes
Posts: 6958
Joined: Mon Jun 15, 2009 3:33 pm

Postby Diogenes » Sun Dec 19, 2010 6:26 pm

TDPerk wrote:
GIThruster wrote:Odd as it may seem, the biggest challenges are in the standard EE work.


You have a frequency domain characterization of the waveform (voltage and current) to be presented at the capacitor leads, and know what freq (freq range) you want to drive it at?

Oh, if pulsed is okay, what cycle time and pulse width do you need? Or do you need continuous?


I'm pretty sure this thing has to be driven with a sine wave, but you are asking questions similar to what I wanted to know.

GIThruster, What are the drive characteristics you need? I saw the 500 mhz, but what kind of power level/Impedance are you needing?

I'm having trouble understanding why this is such a difficult task. I don't see distortion or linearity being much of a problem, and a direct coupled amplifier should be easily capable of achieving the frequency range you are talking about. For an experiment, I don't see efficiency as being an issue. It seems to me that the more important goal is to establish the experimental proof that the concept works well enough to give you a signal far enough out of the noise level to convince skeptics.

I would very much like to know what your specs are for driving this thing. Direct coupled amplifiers are pretty straightforward, especially nowadays with modern Op-Amps.

kunkmiester
Posts: 873
Joined: Thu Mar 12, 2009 3:51 pm
Contact:

Postby kunkmiester » Mon Dec 20, 2010 4:46 am

I would very much like to know what your specs are for driving this thing. Direct coupled amplifiers are pretty straightforward, especially nowadays with modern Op-Amps.

That's what I was thinking. Generating high frequencies isn't a big deal, the big trouble is getting things timed right--pushing when the charge is what it should be, and pulling at the other end of the cycle. I was thinking some of this would be off the shelf.

Timing is more important for a stacked thruster--I was thinking to simply stack them up until the combined thrust is above the higher noise level of a simpler setup. I wasn't planning on just jumping straight into that though, replicating the rotator experiment sounds like a good start.
Evil is evil, no matter how small

GIThruster
Posts: 4686
Joined: Tue May 25, 2010 8:17 pm

Postby GIThruster » Mon Dec 20, 2010 9:41 am

I'm impressed but a little shocked to read the answers above.

Just saying again--if an Ivy League PhD EE notes it takes huge amounts of time to do the power systems necessary, Peeps who only barely know EE, really need to get a grip.

This stuff is REALLY HARD TO DO.

But. . .certainly we're happy to have onboard anyone who has the interest, skill and time to give, and there are certainly easier tasks than building a power system for the UFG Thin. The rotator replication, the GILC, other things, are easier to manage, but you still have to have the time available to do what is always--HARD WORK.

Write me.
"Courage is not just a virtue, but the form of every virtue at the testing point." C. S. Lewis

TDPerk
Posts: 976
Joined: Mon Jul 30, 2007 12:55 pm
Location: Northern Shen. Valley, VA
Contact:

Postby TDPerk » Mon Dec 20, 2010 11:27 am

"Just saying again--if an Ivy League PhD EE notes it takes huge amounts of time to do the power systems necessary, Peeps who only barely know EE, really need to get a grip."

Yeah. Yeah. Yeah. What are the drive characteristics please? Can't hurt to have them out there, can it?
molon labe

montani semper liberi

para fides paternae patria

GIThruster
Posts: 4686
Joined: Tue May 25, 2010 8:17 pm

Postby GIThruster » Mon Dec 20, 2010 6:26 pm

There's no way to answer your question because it's a conceptually confused question. Each experiment has different characteristics. For instance, the MLT has not one but two sine waves (actually saw tooth is preferred) that are 90* out of phase, while the UFG has a single, combined 1+2 omega wave--certainly NOT sine. Generally, MLT's are designed to operate in resonance, meaning their capacitive and inductive elements balance, but until you choose a dielectric, you can't know what values you're talking about. Other possible experiments can be done with things like PMN-PT, which has the highest k value of any material I know of, and the highest electro-mechanical linking coefficient (which then gives the highest bulk accelerations), but it does not have a 1 omega action, so it needs to be driven with pulsed DC to provide a 1 omega action. That requires a VERY different setup than what we've seen so far, but DC has a host of serious advantages. Some experiments require phase locking, others do not. Just as some require an extremely sensitive thrust stand or balance, and methods to remove spurious effects, such as vacuum to remove ionic wind and thermal issues, and shielding to rule out EM coupling. Other experiments like the rotator don't need any of these things.

So you can see, there is no one single answer to such a general question, and if there were, you would certainly want it from an engineer, not a philosopher. There is no "short cut" or "easy way" to answer these very complex questions, just because they were offered in a simplistic manner.
"Courage is not just a virtue, but the form of every virtue at the testing point." C. S. Lewis

GIThruster
Posts: 4686
Joined: Tue May 25, 2010 8:17 pm

Postby GIThruster » Mon Dec 20, 2010 6:47 pm

Diogenes wrote:GIThruster, What are the drive characteristics you need? I saw the 500 mhz, but what kind of power level/Impedance are you needing?


In answer to your specific question as regards the UFG Thin, the Thin thruster is based upon the notion of mechanically resonating a custom ALTAS wafer at about 1 Mhz, with a 1+2 omega action, meaning it will be resonating at about 2 Mhz part of its action. It's an ultra thin wafer (10 mils thick), 1" (25.4mm square) made from the material the 50V caps here are made from:

http://www.tecdia.com/us/product/hf/pdf ... ide_02.pdf

Now if you do the math, you'll find the ALTAS wafer ought to have truly fantastical capacitance, but it needs to be measured in the lab and that has not yet been done. (It will be about 560,000 pF.) In short, you're talking about driving truly amazing capacitance at about 2 Mhz with a very specific waveform. Bruce (the PhD who was originally working on this) chose to do this with DC in part because I think he believes it is not even possible to do this with AC--the standoff of the material is way too low--just 50V.

BTW, it's not just the wafer. You also have to drive the two PZT discs, in phase with the wafer and 180* out of phase with each other, with just the 4 contacts of the assembled stack.

I hope that answers your question. I can't really do better than that because I am not an engineer. :-) I can tell you, it will require about 10 Kw DC amp and a lot of VERY fancy engineering. It also has to be designed as very broadband, because the stack needs to be frequency swept across a broad area in order to find the assembled stack's natural resonance, which could be as high as 2 Mhz but is likely to be very much lower, because of the positive mass fluctuations slowing the stack.

Impedance matching, broadband, with fantastic currents and very little voltage--you can see we did indeed shift the vast bulk of the engineering from the thruster to the power system in order to optimize the function of the thruster. At 1 Mhz, the stacks ought to produce about 1 micron displacements in the 1 omega, and so the bulk aceleration is on the order of 2 million gees.
"Courage is not just a virtue, but the form of every virtue at the testing point." C. S. Lewis

Diogenes
Posts: 6958
Joined: Mon Jun 15, 2009 3:33 pm

Postby Diogenes » Tue Dec 21, 2010 12:08 am

GIThruster wrote:
Diogenes wrote:GIThruster, What are the drive characteristics you need? I saw the 500 mhz, but what kind of power level/Impedance are you needing?


In answer to your specific question as regards the UFG Thin, the Thin thruster is based upon the notion of mechanically resonating a custom ALTAS wafer at about 1 Mhz, with a 1+2 omega action, meaning it will be resonating at about 2 Mhz part of its action. It's an ultra thin wafer (10 mils thick), 1" (25.4mm square) made from the material the 50V caps here are made from:

http://www.tecdia.com/us/product/hf/pdf ... ide_02.pdf

Now if you do the math, you'll find the ALTAS wafer ought to have truly fantastical capacitance, but it needs to be measured in the lab and that has not yet been done. (It will be about 560,000 pF.) In short, you're talking about driving truly amazing capacitance at about 2 Mhz with a very specific waveform. Bruce (the PhD who was originally working on this) chose to do this with DC in part because I think he believes it is not even possible to do this with AC--the standoff of the material is way too low--just 50V.

BTW, it's not just the wafer. You also have to drive the two PZT discs, in phase with the wafer and 180* out of phase with each other, with just the 4 contacts of the assembled stack.


I'm a bit confused. From what I recall reading, the Mechanical transfer characteristics have been a big problem with any sort of Stacked design. I was thinking that your plates need to be chemically bonded to your dielectric to reduce this problem as much as possible.


GIThruster wrote:I hope that answers your question. I can't really do better than that because I am not an engineer. :-) I can tell you, it will require about 10 Kw DC amp and a lot of VERY fancy engineering.



Not necessarily. Unless I am misunderstanding something, your drive characteristics could be implemented with an ordinary DC coupled design with perhaps some negative feedback correction to improve the waveform as needed. The reference wave shape could be created with a digital to analog converter output from a PC. Waveforms of any shape you want can easily be produced in this manner, up to several megahertz.

What is really important is whether or not you can apply the Signal you want to the load, not whether or not you do it efficiently. I suspect that with a PC doing "On the fly" error correction by varying the drive characteristics in accordance with the negative feedback it's getting from the output, all those difficult complexities of amplifier design become not as difficult.

GIThruster wrote: It also has to be designed as very broadband, because the stack needs to be frequency swept across a broad area in order to find the assembled stack's natural resonance, which could be as high as 2 Mhz but is likely to be very much lower, because of the positive mass fluctuations slowing the stack.






I suspected that's what the large bandwidth was intended for and why this might be thought of as complicated by a PHD. The impedance will change as you change frequency, and so your amplifier will experience a change in it's loading characteristics as you attempt to find the "Sweet spot."

I still don't see it as a deal killer. Impedance matching is mostly important to insure that power transfers efficiently. You can still transfer power with an impedance mismatch, you will just be dealing with losses. Again, as efficiency ought not to be of primary concern, I am thinking that a big enough heat sink would tolerate any mismatch. Another issue might be eliminating harmonics and parasitics. I'm thinking that if the leads between the drivers and the capacitor are kept short enough, this shouldn't be much of a problem at these low frequencies.


GIThruster wrote:Impedance matching, broadband, with fantastic currents and very little voltage--you can see we did indeed shift the vast bulk of the engineering from the thruster to the power system in order to optimize the function of the thruster. At 1 Mhz, the stacks ought to produce about 1 micron displacements in the 1 omega, and so the bulk aceleration is on the order of 2 million gees.



Yeah, if your drive characteristics require a 200 amp current through your 1" capacitor, it's probably going to impart motion to something at least once! Maybe not the way you had in mind. :)


Anyway, you've given me more to think about. At least I have some notion. A 200 amp current with a 50 volt drive sets some limits that I can work with. The output drivers will likely have to be an array, but I don't think this will present any special problems. Corrective feedback can perform wonders.


Return to “General”

Who is online

Users browsing this forum: No registered users and 7 guests