Author Topic: EM Drive Developments - related to space flight applications - Thread 3  (Read 1878950 times)

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
The doughty DIYer has not yet stepped forward with a plan to build a frustum out of transparent aluminium.  8)

I'm told this stuff actually exists. But resistivity...

Offline TheTraveller

I am also concerned that this emdrive seems to have been thrown together. I'd have proved the magnetron first.

I don't understand why there's not more effort given towards the production of a device with higher Q. I am not asking for a superconductor cavity, but can it at least be polished? not hand-soldered? Made of decently pure copper?

Using a magnetron as the Rf source, the Q needs to be around 50 so the cavity input bandwidth can accept the wide frequency range the magnetron outputs.

The higher Q, the narrower the cavity input bandwidth. For Qs of say 50,000, the Rf source needs to be narrow band and it needs to have the ability to automatically adjust the frequency to stay inside the narrow cavity input bandwidth.

Only needs to be polished on the inside surfaces.

In my cavity build, not only will the inside surfaces be polished to a mirror like finish but the cavity will be sealed and filed with N2 at 1/2 atmo pressure. Don't want all my hard work doing the polishing end up being oxidised by the oxygen inside an air filled cavity. The cavity will also have internal pressure monitoring and data logging.

Here I note Tajmar did observe a Q drop from 48.8 to 20.3 over the course of his experiments due to the internal surfaces becoming oxidised. So internal oxidation is a serious issue.
« Last Edit: 07/29/2015 05:58 AM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
I am also concerned that this emdrive seems to have been thrown together. I'd have proved the magnetron first.

I don't understand why there's not more effort given towards the production of a device with higher Q. I am not asking for a superconductor cavity, but can it at least be polished? not hand-soldered? Made of decently pure copper?

Using a magnetron as the Rf source, the Q needs to be around 50 so the cavity input bandwidth can accept the wide frequency range the magnetron outputs.

The higher Q, the narrower the cavity input bandwidth. For Qs of say 50,000, the Rf source needs to be narrow band and it needs to have the ability to automatically adjust the frequency to stay inside the narrow cavity input bandwidth.
What regulation algorithm do you plan on using? P? PI? PD? PID?
I'm guessing pure 'P' (proportional)

Offline TheTraveller

TT, the reason I did not include Shawyer's formula here, is that unlike the three I mentioned above that all have a common basic force, which is the difference between two energy states, divided by the length. Shawyer uses the difference between two forces. I have yet to resolve his equation to be "similar" to the other 3.
Thanks.

Todd,

For Force to be generated, the cavity must operate at resonance and for that to happen physical length & end plate diameters, effective cavity guide wavelength, each end plate guide wavelength and external driving frequency are all involved.

While the above may not seem to be in F = (2 P Q Df) / they are very much there.
« Last Edit: 07/29/2015 06:11 AM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline tchernik

  • Full Member
  • **
  • Posts: 251
  • Liked: 298
  • Likes Given: 583
By the way, here you can find legit non-paywalled links to the papers presented by Martin Tajmar in PDF format at AIAA 2015, among others.

http://tu-dresden.de/die_tu_dresden/fakultaeten/fakultaet_maschinenwesen/ilr/rfs/forschung/folder.2007-08-21.5231434330/ag_raumfahrtantriebe/breakthrough_propulsion_physics

P.S.

I don't know how to make a correct hyperlink. Hopefully this doesn't break the page format.

Offline R.W. Keyes

  • Member
  • Posts: 67
  • Massachusetts
  • Liked: 42
  • Likes Given: 42
I am also concerned that this emdrive seems to have been thrown together. I'd have proved the magnetron first.

I don't understand why there's not more effort given towards the production of a device with higher Q. I am not asking for a superconductor cavity, but can it at least be polished? not hand-soldered? Made of decently pure copper?

Using a magnetron as the Rf source, the Q needs to be around 50 so the cavity input bandwidth can accept the wide frequency range the magnetron outputs.

The higher Q, the narrower the cavity input bandwidth. For Qs of say 50,000, the Rf source needs to be narrow band and it needs to have the ability to automatically adjust the frequency to stay inside the narrow cavity input bandwidth.

Only needs to be polished on the inside surfaces.

In my cavity build, not only will the inside surfaces be polished to a mirror like finish but the cavity will be sealed and filed with N2 at 1/2 atmo pressure. Don't want all my hard work doing the polishing end up being oxidised by the oxygen inside an air filled cavity. The cavity will also have internal pressure monitoring and data logging.

Here I note Tajmar did observe a Q drop from 48.8 to 20.3 over the course of his experiments due to the internal surfaces becoming oxidised. So internal oxidation is a serious issue.

I understand the relationship between Q and bandwidth. A previous post here provided a paper detailing ways to narrow the bandwidth of a magnetron. Even if that was not possible, how difficult/expensive is it to get a more narrow band RF source? I'd almost expect that a place like TU Dresden would have such things at hand already. If not, I would expect that the university could shake loose one or two thousand euro for the equipment, or if so, to be able to rent or borrow it.

I assume that the device was not polished, nor was very pure copper used because it was not mentioned in the text. You are wise to polish your model, and to protect it against corrosion.

Offline TheTraveller

I am also concerned that this emdrive seems to have been thrown together. I'd have proved the magnetron first.

I don't understand why there's not more effort given towards the production of a device with higher Q. I am not asking for a superconductor cavity, but can it at least be polished? not hand-soldered? Made of decently pure copper?

Using a magnetron as the Rf source, the Q needs to be around 50 so the cavity input bandwidth can accept the wide frequency range the magnetron outputs.

The higher Q, the narrower the cavity input bandwidth. For Qs of say 50,000, the Rf source needs to be narrow band and it needs to have the ability to automatically adjust the frequency to stay inside the narrow cavity input bandwidth.
What regulation algorithm do you plan on using? P? PI? PD? PID?
I'm guessing pure 'P' (proportional)

Is really very simple.

The real time VSWR output of the 100W Rf amp will be monitored in real time. Every 50ms or so (settable time) the freq will be adjusted +1kHz and the VSWR compared to what it was before, If it has dropped, another +1kHz frequency alteration will be done until it increases, then back off the last adj. If the 1st +1kHz step increased the VSWR, the freq will be dropped until the lowest VSWR is found.

Of course there will be some hysteresis involved and will include code to deal with that effect once I have quantified it.

With a projected unloaded cavity Q of 100k, the -3dB side freqs are +- 12.5kHz, so this VSWR or return loss or reflected power tuning system should be fairly sensitive to +-1kHz freq steps.
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline deltaMass

  • Full Member
  • ****
  • Posts: 955
  • A Brit in California
  • Liked: 671
  • Likes Given: 275
Yes, it's the simplest - 'P' at constant gain.
« Last Edit: 07/29/2015 06:31 AM by deltaMass »

Offline TheTraveller

I am also concerned that this emdrive seems to have been thrown together. I'd have proved the magnetron first.

I don't understand why there's not more effort given towards the production of a device with higher Q. I am not asking for a superconductor cavity, but can it at least be polished? not hand-soldered? Made of decently pure copper?

Using a magnetron as the Rf source, the Q needs to be around 50 so the cavity input bandwidth can accept the wide frequency range the magnetron outputs.

The higher Q, the narrower the cavity input bandwidth. For Qs of say 50,000, the Rf source needs to be narrow band and it needs to have the ability to automatically adjust the frequency to stay inside the narrow cavity input bandwidth.

Only needs to be polished on the inside surfaces.

In my cavity build, not only will the inside surfaces be polished to a mirror like finish but the cavity will be sealed and filed with N2 at 1/2 atmo pressure. Don't want all my hard work doing the polishing end up being oxidised by the oxygen inside an air filled cavity. The cavity will also have internal pressure monitoring and data logging.

Here I note Tajmar did observe a Q drop from 48.8 to 20.3 over the course of his experiments due to the internal surfaces becoming oxidised. So internal oxidation is a serious issue.

I understand the relationship between Q and bandwidth. A previous post here provided a paper detailing ways to narrow the bandwidth of a magnetron. Even if that was not possible, how difficult/expensive is it to get a more narrow band RF source? I'd almost expect that a place like TU Dresden would have such things at hand already. If not, I would expect that the university could shake loose one or two thousand euro for the equipment, or if so, to be able to rent or borrow it.

I assume that the device was not polished, nor was very pure copper used because it was not mentioned in the text. You are wise to polish your model, and to protect it against corrosion.

Polish advice was from Shawyer, who I'm in email contact with. Doing the N2 fill was my idea to protect my polishing work.

I have sourced a wide band 100W Rf amp at around $450US. By using the real time reflected power output I'm developing an automatic system to track the lowest reflected power, to track any cavity changes from thermal heating. This should ensure the max amount of the Rf amps power will be inside the cavity at all times.

Another nice ferature of the Rf amp is I can adjust the output power from 79mWs to 100Ws in 32 x 1dBm steps, which is good for doing experimental tuning and impedance matching at 79mWs without stressing the amp at 100Ws.

Coax of choice is EcoFlex15Plus. VERY low loss at 2.45GHz and can handle 350W at that freq so OK for my tests.
« Last Edit: 07/29/2015 06:53 AM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline TheTraveller

By the way, here you can find legit non-paywalled links to the papers presented by Martin Tajmar in PDF format at AIAA 2015, among others.

http://tu-dresden.de/die_tu_dresden/fakultaeten/fakultaet_maschinenwesen/ilr/rfs/forschung/folder.2007-08-21.5231434330/ag_raumfahrtantriebe/breakthrough_propulsion_physics

P.S.

I don't know how to make a correct hyperlink. Hopefully this doesn't break the page format.

Nice find. Thanks for posting it.

Direct download link for the Tajmar EMDrive paper:
http://tu-dresden.de/die_tu_dresden/fakultaeten/fakultaet_maschinenwesen/ilr/rfs/forschung/folder.2007-08-21.5231434330/ag_raumfahrtantriebe/JPC%20-%20Direct%20Thrust%20Measurements%20of%20an%20EM%20Drive%20and%20Evaluation%20of%20Possible%20Side-Effects.pdf

Many interesting papers there to read.
« Last Edit: 07/29/2015 07:17 AM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline R.W. Keyes

  • Member
  • Posts: 67
  • Massachusetts
  • Liked: 42
  • Likes Given: 42

I have sourced a wide band 100W Rf amp at around $450US. By using the real time reflected power output I'm developing an automatic system to track the lowest reflected power, to track any cavity changes from thermal heating. This should ensure the max amount of the Rf amps power will be inside the cavity at all times.

Another nice ferature of the Rf amp is I can adjust the output power from 79mWs to 100Ws in 32 x 1dBm steps, which is good for doing experimental tuning and impedance matching at 79mWs without stressing the amp at 100Ws.

Coax of choice is EcoFlex15Plus. VERY low loss at 2.45GHz and can handle 350W at that freq so OK for my tests.

Sounds like a good amplifier. Can you tell me where to get one? For when my money comes in. How does Ecoflex compare to LMR400? By the way, I have a bunch of unused LMR400 4' and 6' cables with N-Male connectors on them, if anyone is looking.

Offline TheTraveller


I have sourced a wide band 100W Rf amp at around $450US. By using the real time reflected power output I'm developing an automatic system to track the lowest reflected power, to track any cavity changes from thermal heating. This should ensure the max amount of the Rf amps power will be inside the cavity at all times.

Another nice ferature of the Rf amp is I can adjust the output power from 79mWs to 100Ws in 32 x 1dBm steps, which is good for doing experimental tuning and impedance matching at 79mWs without stressing the amp at 100Ws.

Coax of choice is EcoFlex15Plus. VERY low loss at 2.45GHz and can handle 350W at that freq so OK for my tests.

Sounds like a good amplifier. Can you tell me where to get one? For when my money comes in. How does Ecoflex compare to LMR400? By the way, I have a bunch of unused LMR400 4' and 6' cables with N-Male connectors on them, if anyone is looking.

At 2.5GHZ:

LMR400 22.2 dB loss / 100 mtr, max pwr 330W
EcoFlex15Plus: 14.9 dB loss / 100 mtr, max pwr 350W

Very similar.

As for the 100W amp, a friend of mine who does manuf and sourcing in China found it.
http://www.alibaba.com/product-detail/100w-High-Power-Broadband-Power-amplifier_60125990298.html
You may need to register with Alibaba to see the link.
« Last Edit: 07/29/2015 07:33 AM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline frobnicat

  • Full Member
  • ****
  • Posts: 518
  • Liked: 500
  • Likes Given: 151
...
-JK

Just some suggestions...

TO THE MEEPERS; This is WHY we need an accurate simulation that shows what happens when the power is turned off. If the losses are small, the resonant wave will persist for some period of time until all the energy decays. We need to know what that decay time really is. In fact, I would predict that the thrust goes up the faster it decays. Thrust persisting after it is turned off is to be expected. The question is, how long does it take for the internal stored energy to dissipate? The higher the Q, the longer it will take to discharge.

DIY's; It would be a good idea to monitor what is going on inside the frustum at all times. Such that you can determine the difference between a hot frustum, and one that is still charged with resonant energy. If there is persistent force after the power is turned off but there is nothing going on inside it, then you know you have an artifact. Just because it's hot doesn't mean it's an artifact. The temperature will follow the power dissipation. As it declines so will the temperature. The two are well correlated, so the fact that the thrust persists and decays with the temperature is to be expected. This alone is not an indication of an artifact. The artifact would be only IF there is no energy inside the frustum doing the work.

How do you reconcile this line of argument with the (way above) 6 orders of magnitude difference in time constants between decaying resonant EM energy (about 1΅s for high Q, probably less than 100ns at Q=50) and apparent lingering "thrust" decaying time >>1s ?

Quote
Still, I would be hard pressed to believe that thermal radiation can exert forces at least an order of magnitude larger than a photon rocket.

True but even in a vacuum there can be other thermal effects, thermally driven out gassing as mentioned, centre of mass displacements, thermally induced differential stress (buckling...), change in the flexure bearings stiffness (shifting balance equilibrium point)...

Frankly, what kind of mechanism can "sponge up" energy and release it at such time constants > 1s after power off ? Heat for sure, electrostatic charge maybe, remanent magnetization (but that would tend to get stuck, not decay), what else ? Certainly not anything involving circulating currents or bouncing waves (in the absence of superconductivity). If you disagree please put forth rough magnitude values to your argument, because I'm tired of this EM charge/discharge time explanation that's used again and again each time experiments show soft transients and stuck signals when we expect prompt step up and prompt down back to baseline from anything on/off EM related.

If a 50s powered drive can really thrust for 500s then integrate that in the thrust/power accordingly (=total_momentum/total_energy) and see if it isn't above the Q times photon rocket (that appears as a limit in some of the formulas).

Quote
Regarding arcing, breakdown, etc. Today's video from @rfmwguy did not appear to me to show arching. Shell said she saw it, but I think that what looks like arcing is actually his laser on the thermal probe reflecting off the magnetron. I did not see arcing. Perhaps @rfmwguy will update us in the morning regarding this.
Todd

Invest in small portable ozone monitors ? I see it starts about 600$ ...
« Last Edit: 07/29/2015 10:40 AM by frobnicat »

Offline TheTraveller

Tajmar's Fig 5b which shows a summary of the small end Up, small end Down and small end Horizontal might be seen to be making an incorrect conclusion.

I took the Down curve and inverted it to show the negative Forces generated in the 2nd attachment. What we see is the lingering, after power off Force being generated in each direction. Buoyancy could not do this as it would only generate lift and not generate a lingering downward Force.

There does appear to be buoyant in effect as the Up force is greater than the Down Force by 229uNs or a bouyancy Force of 115uN applied to a +505uN Up Force and to a - 505uN Down Force.

Why both the Up and Down Forces lingered after power off is a question to be answered.
« Last Edit: 07/29/2015 11:20 AM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
...How do you reconcile this line of argument with the (way above) 6 orders of magnitude difference in time constants between decaying resonant EM energy (about 1΅s for high Q, probably less than 100ns at Q=50) and apparent lingering "thrust" decaying time >>1s ?...
Tajmar's Fig 5b which shows a summary of the small end Up, small end Down and small end Horizontal might be seen to be making an incorrect conclusion.

I took the Down curve and inverted it to show the negative Forces generated in the 2nd attachment. What we see is the lingering, after power off Force being generated in each direction. Buoyancy could not do this as it would only generate lift and not generate a lingering downward Force.

There does appear to be buoyant in effect as the Up force is greater than the Down Force by 229uNs or a bouyancy Force of 115uN applied to a +505uN Up Force and to a - 505uN Down Force.

Why both the Up and Down Forces lingered after power off is a question to be answered.
I question the elasticity of the scale as being responsible for not springing back.  We should not assume that the scale acts like a perfect elastic body.

To analyze why the scale is not acting as perfect elastic we would need to analyze the Sartorius AX224 scale.

« Last Edit: 07/29/2015 11:46 AM by Rodal »

Offline TheTraveller

Email from Roger Shawyer just arrived giving author access to his official peer reviewed paper and a video presentation on the right side which mentions the Tajmar results.

Quote
Hi

Now that Martins paper has been presented I will release free access to the peer reviewed version of my IAC14 paper, which was actually first published a couple of weeks ago. This author’s guest link includes access to a 5 minute audio/slide presentation which I updated to include Martins test data.

http://authors.elsevier.com/a/1RQaGLWHFbB5c

I believe my copyright allows distribution of the link via websites, forums etc, so feel free to share this if you want.

Best regards
Roger
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline TheTraveller

I question the elasticity of the scale as being responsible for not springing back.  We should not assume that the scale acts like a perfect elastic body.

To analyze why the scale is not acting as perfect elastic we would need to analyze the Sartorius AX224 scale.

If scale hysteresis was responsible, would also expect the original Force generation measurement to be effected.
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2165
  • Liked: 2681
  • Likes Given: 1124
(Reposting for posterity with the direct links to the paper removed:)

I see three pieces of evidence from Tajmar's paper that high power RF effects such as corona breakdown, multipaction (or simple out-gassing) could be incinerating the materials in the cavity and generating particles thus creating the observed thrust:

1) For the latest results, the fact that the thrust continues to exist after the removal of RF power and correlates well to temperature indicates to me that particle generation is due to thermal effects (such as burning an adhesive).

From the paper:
"The implementation of all isolation methods (thermal, magnetic, air circulation block) resulted in the cleanest measurement with an expected behavior such that the thrust appeared after turn-on, then steadily increased until power turn off. It then remained there and slowly decreased as the EMDrive cooled down. "

2) The second piece of evidence comes from my suggestion here: http://forum.nasaspaceflight.com/index.php?topic=36313.msg1367663#msg1367663 that the cavity be disassembled and inspected for the damaging effects of the above phenomenon to verify if they are occurring or not.

It seems this was done and the damage was found:
" Indeed we measured that our Q factor was reduced to only 20.3 – probably due to the fact that our inner surfaces were now much more oxidized compared to the start of our test campaign after a visual inspection. "
The visual evidence confirms that the effects I've mentioned are occurring.

3) Further, taking a look at the thermal imager pictures in Figure 3 it appears the seam of the cylindrical cavity is the hottest point which is where you would expect these effects to occur as was also suggested here http://forum.nasaspaceflight.com/index.php?topic=36313.msg1367663#msg1367663

-JK

Just some suggestions...

TO THE MEEPERS; This is WHY we need an accurate simulation that shows what happens when the power is turned off. If the losses are small, the resonant wave will persist for some period of time until all the energy decays. We need to know what that decay time really is. In fact, I would predict that the thrust goes up the faster it decays. Thrust persisting after it is turned off is to be expected. The question is, how long does it take for the internal stored energy to dissipate? The higher the Q, the longer it will take to discharge.

DIY's; It would be a good idea to monitor what is going on inside the frustum at all times. Such that you can determine the difference between a hot frustum, and one that is still charged with resonant energy. If there is persistent force after the power is turned off but there is nothing going on inside it, then you know you have an artifact. Just because it's hot doesn't mean it's an artifact. The temperature will follow the power dissipation. As it declines so will the temperature. The two are well correlated, so the fact that the thrust persists and decays with the temperature is to be expected. This alone is not an indication of an artifact. The artifact would be only IF there is no energy inside the frustum doing the work. Still, I would be hard pressed to believe that thermal radiation can exert forces at least an order of magnitude larger than a photon rocket.

Regarding arcing, breakdown, etc. Today's video from @rfmwguy did not appear to me to show arching. Shell said she saw it, but I think that what looks like arcing is actually his laser on the thermal probe reflecting off the magnetron. I did not see arcing. Perhaps @rfmwguy will update us in the morning regarding this.
Todd
Good eye warp...the ir pointer was the culprit. It bounced around the magnetron's top magnet.

Best I can tell after visual inspection thru the copper mesh, there were no hotspots or arcs. I will do more testing this week, building up on power duration. Ideally, I want a full power 5 minute run.

Best to take that in steps...:/

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6045
  • Likes Given: 5325
I question the elasticity of the scale as being responsible for not springing back.  We should not assume that the scale acts like a perfect elastic body.

To analyze why the scale is not acting as perfect elastic we would need to analyze the Sartorius AX224 scale.

If scale hysteresis was responsible, would also expect the original Force generation measurement to be effected.

Not necessarily.  There are many other forms of imperfect elasticity that do not involve hysteresis. 
Need an engineering analysis of the scale itself instead of preconceptions.
I have not found an engineering description of this actual scale in the web to perform an analysis.  Just general low-level operating manuals.

« Last Edit: 07/29/2015 11:58 AM by Rodal »

Offline TheTraveller

Not necessarily.  There are many other forms of imperfect elasticity that do not involve hysteresis. 
Need an engineering analysis of the scale itself instead of preconceptions.
I have not found an engineering description of this actual scale in the web to perform an analysis.  Just general low-level operating manuals.

Strange for bidirectional scale stiction to release as the cavity cooled down but maybe that is the reason.

Each of the measured 15s on runs were not repeated for 600s, to allow the magnetron to cool down and for the lingering Force to decay back to baseline. While this really does sound like heat buoyancy, that can't be the case for the down test as there is no way I can see buoyancy would cause a lingering down Force.
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.”
Herman Melville, Moby Dick

Tags: