-
#2060 by rfmwguy on 23 Sep, 2015 18:28
-
I'm going to do everything I can to design the test in ambient air. I know its not the best, but for DIY types, vac chambers, especially large ones, are a no go. Working on a plan to negate lift but only running the tests at a preheated temp. Whether thats a heater underneath or fire the magnetron up for warm-up has not been decided.
Regardless, my goal is to beat the lift characteristics which I think can be quantified and removed from the test data. Thats my hope anyway...
Was thinking of the mini-EMDrive that might fit into a small vacuum chamber (though I wonder if the electronics involved can run in vacuum). The forces there are small enough that they're using a laser interferometer to measure it.
For NSF-1701, I think the strange behavior of the lift might be related to air escaping from around the magnetron heatsink at random intervals. That said, I keep going back to that quantum reflection of photons paper. There were some mentions of redshift (not surprising, bouncing around photons should produce redshift). I wonder if I'm misunderstanding and some redshift is the result of the quantum mirror "powering up" and starting to reflect photons before resonance is reached. That would explain why peak thrust is reached at a frequency slightly higher than the resonant frequency of the frustum -- the waves are being redshifted down into the resonant frequency.
... At worst you'll end up with more data to help characterize the thermal lift.
Thats my hope. When I report the thermal "settling" time after power off, think people will be surprised at the persistence. It was late last night. Had no idea it would go in for minutes. This probably tells me heating below the magnetron has stopped after power-off and the persistence is from the mag shedding heat very slowly. Not to mention lift continued upwards after power-off. However, I can't claim this to be a fact yet, its just unexpected.
As far as fans are concerned Shell, I'm gonna try and avoid any shawyer-like pumps and fans, although I might try and blow air across it perpendicular to rise.
-
#2061 by SeeShells on 23 Sep, 2015 18:57
-
I see what your saying on the RS pumps and fans but his were not strategically placed and he didn't quantify them. Even placing a air flow over the top, the maggie will still create a thermal pressure high and low that will draw the frustum upwards.I'm going to do everything I can to design the test in ambient air. I know its not the best, but for DIY types, vac chambers, especially large ones, are a no go. Working on a plan to negate lift but only running the tests at a preheated temp. Whether thats a heater underneath or fire the magnetron up for warm-up has not been decided.
Regardless, my goal is to beat the lift characteristics which I think can be quantified and removed from the test data. Thats my hope anyway...
Was thinking of the mini-EMDrive that might fit into a small vacuum chamber (though I wonder if the electronics involved can run in vacuum). The forces there are small enough that they're using a laser interferometer to measure it.
For NSF-1701, I think the strange behavior of the lift might be related to air escaping from around the magnetron heatsink at random intervals. That said, I keep going back to that quantum reflection of photons paper. There were some mentions of redshift (not surprising, bouncing around photons should produce redshift). I wonder if I'm misunderstanding and some redshift is the result of the quantum mirror "powering up" and starting to reflect photons before resonance is reached. That would explain why peak thrust is reached at a frequency slightly higher than the resonant frequency of the frustum -- the waves are being redshifted down into the resonant frequency.
... At worst you'll end up with more data to help characterize the thermal lift.
Thats my hope. When I report the thermal "settling" time after power off, think people will be surprised at the persistence. It was late last night. Had no idea it would go in for minutes. This probably tells me heating below the magnetron has stopped after power-off and the persistence is from the mag shedding heat very slowly. Not to mention lift continued upwards after power-off. However, I can't claim this to be a fact yet, its just unexpected.
As far as fans are concerned Shell, I'm gonna try and avoid any shawyer-like pumps and fans, although I might try and blow air across it perpendicular to rise.
I thought it world be a interesting way to turn a vertical thermal chaotic effect into a easier quantifiable rotational one and rid the upwards rising thermal plume from your maggie. In other words turn your vertical rising component into a rotational which your measurements can't see anyway. Am I just blowing hot air here?
It would be easy to test with a couple of PC fans to see the difference in vertical measurements. -
#2062 by lmbfan on 23 Sep, 2015 19:21
-
Unless Aero has changed something recently, all the MEEP simulations have air or vacuum inside and outside the copper walls. Radiation could be seen propagating outside the frustum if there were any.
-
#2063 by RERT on 23 Sep, 2015 19:37
-
Rfmwguy -
If thermal effects operated on a timescale much longer than thrust effects, you could proceed as follows.
Cycle the magnetron on/off quite quickly, but for a long time. If the cycle time is much shorter than the response time for thermal effects, you should reach a situation where the system is always close to constant thermal effect, but with a superimposed cyclical thrust. It would be easy to pick out the corresponding cyclical component in the response. If that cyclical signal reverses when you invert the frustrum, you would have some evidence for thrust.
One fly in the ointment might be that the response time for the beam is too long to allow the rapid magnetron cycling to be seen. The method would only work if the cycle time and beam and thermal response times all fit right. At the very least you would have to pick the cycle time judiciously, and then get a little lucky with the thermal inertia.
No originality from me here, this is more or less what you did on your first test, though the cycling then was more to do with thermal management than creating a signal to extract.
R. -
#2064 by rfmwguy on 23 Sep, 2015 19:56
-
Rfmwguy -
Thanks, very good thoughts. You are right, I did the 30% for 5 minute testing and it brought the mag up to temp and maintained it around 170 deg C. The beam actually starts rising fairly quickly. This surprised me since I have a 4x4 inch, 3 inch deep oil pool for the dampener.
If thermal effects operated on a timescale much longer than thrust effects, you could proceed as follows.
Cycle the magnetron on/off quite quickly, but for a long time. If the cycle time is much shorter than the response time for thermal effects, you should reach a situation where the system is always close to constant thermal effect, but with a superimposed cyclical thrust. It would be easy to pick out the corresponding cyclical component in the response. If that cyclical signal reverses when you invert the frustrum, you would have some evidence for thrust.
One fly in the ointment might be that the response time for the beam is too long to allow the rapid magnetron cycling to be seen. The method would only work if the cycle time and beam and thermal response times all fit right. At the very least you would have to pick the cycle time judiciously, and then get a little lucky with the thermal inertia.
No originality from me here, this is more or less what you did on your first test, though the cycling then was more to do with thermal management than creating a signal to extract.
R.
I did use a stainless shower drain cover (no, my wife said I had to buy one) which has many rectangular holes allowing oil flow. I had a solid circular piece at one time and it retarded movement way too much. -
#2065 by jmossman on 23 Sep, 2015 21:08
-
I'm going to do everything I can to design the test in ambient air. I know its not the best, but for DIY types, vac chambers, especially large ones, are a no go. Working on a plan to negate lift but only running the tests at a preheated temp. Whether thats a heater underneath or fire the magnetron up for warm-up has not been decided.
Regardless, my goal is to beat the lift characteristics which I think can be quantified and removed from the test data. Thats my hope anyway...
Congratulations on the continued characterization work!
Perhaps another optimization to consider is what modifications to the magnetron housing might improve passive heat dissipation? Various Intel CPU heatsinks and such have been mentioned in the past; I wonder if affixing some thermal grease and some extra heatsink fins (copper preferably, rather than aluminum) would help heat dissipation?
Pictures of your magnetron seem to suggest ?aluminum? heat sink fins.... compared to the aggressive copper fin counts available in modern CPU and chipset heatsinks, I suspect there's some low-hanging fruit available. Also seems like that magnetron was designed to be mounted with those straight heatsink fins oriented vertically (perhaps to allow a chimney effect to improve passive cooling?). I suspect the magnetron heatsink fins won't be nearly as effective when oriented horizontally and stacked on top of each other.
P.S. Caution regarding potential hazardous substances within the magnetron/housing should probably be restated for the benefit of lurkers. Drilling holes into a magnetron is *NOT* recommended; the resulting air-born powders can be very dangerous. Instead, I was envisioning use of thermal grease and securing with cable clamps/ties and/or some high-temp RTV in the corners.
EDIT: fixed NSF-1701 youtube video link
EDIT: +Caution against drilling into a magnetron
-
#2066 by aero on 23 Sep, 2015 21:20
-
Unless Aero has changed something recently, all the MEEP simulations have air or vacuum inside and outside the copper walls. Radiation could be seen propagating outside the frustum if there were any.
I've been running air filled frustums lately - can't remember when the last time I ran vacuum filled but these runs are most assuredly air filled. -
#2067 by rfmwguy on 23 Sep, 2015 21:59
-
Thanks! Yep, Berrylium Oxide ceramic is nasty stuff...not drilling or fiddling about with it.I'm going to do everything I can to design the test in ambient air. I know its not the best, but for DIY types, vac chambers, especially large ones, are a no go. Working on a plan to negate lift but only running the tests at a preheated temp. Whether thats a heater underneath or fire the magnetron up for warm-up has not been decided.
Regardless, my goal is to beat the lift characteristics which I think can be quantified and removed from the test data. Thats my hope anyway...
Congratulations on the continued characterization work!
Perhaps another optimization to consider is what modifications to the magnetron housing might improve passive heat dissipation? Various Intel CPU heatsinks and such have been mentioned in the past; I wonder if affixing some thermal grease and some extra heatsink fins (copper preferably, rather than aluminum) would help heat dissipation?
Pictures of your magnetron seem to suggest ?aluminum? heat sink fins.... compared to the aggressive copper fin counts available in modern CPU and chipset heatsinks, I suspect there's some low-hanging fruit available. Also seems like that magnetron was designed to be mounted with those straight heatsink fins oriented vertically (perhaps to allow a chimney effect to improve passive cooling?). I suspect the magnetron heatsink fins won't be nearly as effective when oriented horizontally and stacked on top of each other.
P.S. Caution regarding potential hazardous substances within the magnetron/housing should probably be restated for the benefit of lurkers. Drilling holes into a magnetron is *NOT* recommended; the resulting air-born powders can be very dangerous. Instead, I was envisioning use of thermal grease and securing with cable clamps/ties and/or some high-temp RTV in the corners.
EDIT: fixed NSF-1701 youtube video link
EDIT: +Caution against drilling into a magnetron
Good observation. I did notice the fins (yokes) were aligned vertically in the microwave box just as you suspected. This would make a better chimney for sure, but not possible with my config. -
#2068 by aceshigh on 24 Sep, 2015 05:59
-
Thanks! Yep, Berrylium Oxide ceramic is nasty stuff...not drilling or fiddling about with it.
plus thorium mixed with tungsten... (in the filament) -
#2069 by RotoSequence on 24 Sep, 2015 06:20
-
Good observation. I did notice the fins (yokes) were aligned vertically in the microwave box just as you suspected. This would make a better chimney for sure, but not possible with my config.
You could spring for a 90 degree waveguide and position the magnetron closer to the fulcrum, but I have no idea what that would do to your weight (or fiscal) budget. -
#2070 by Fugudaddy on 24 Sep, 2015 15:25
-
Okay so this may well be a wrong-headed idea, but what is the reason for using a solid tube attached to the side of the frustum to deliver EM energy? If the main cause for lift is the heat generated by the EM generator (such as in nsf-1701), then if that is instead connected to the main frustum via something with some 'give' to it, it may not deliver as 'clean' of microwaves (if that matters), but it should then eliminate thermal lift caused by the EM generation system. No?
Ronald
-
#2071 by madmax on 24 Sep, 2015 15:55
-
Ronald, until you asked this, I think I had assumed (perhaps wrongly) that the heating that was interfering with measurement was the frustrum. But if it's the magnetron only that is heating, couldn't that be isolated from the frustrum / measurement apparatus?
Or is there heating of the frustrum as well?
Been lurking for quite a while, but decided to make an account when I saw this.
-
#2072 by rfmwguy on 24 Sep, 2015 16:46
-
NSF-1701 update - tonight, I will record and upload video of Flight Test 2A. It will be a long one, will do a walkaround of upgraded test stand, the flight test itself and a screen capture of the data, so I'll keep the video rolling then add the screen capture video at the end. Let me say this...I spent the morning live testing and determined how to stabilize lift, it is a magnetron preheat to 170 degrees C. Stay tuned, you will not be disappointed. My results will be posted on NSF first.
-
#2073 by SeeShells on 24 Sep, 2015 23:07
-
Areo and I got the dual opposing waveguides working, wasn't sure what we would see but I hoped for a symmetrical mode generation and there was. Taking a slices of time through the CE frustum it is apparent that there are at least 2 modes overlaying each other.
-
#2074 by ThinkerX on 25 Sep, 2015 00:35
-
Quote
Areo and I got the dual opposing waveguides working, wasn't sure what we would see but I hoped for a symmetrical mode generation and there was. Taking a slices of time through the CE frustum it is apparent that there are at least 2 modes overlaying each other.
Going from the overly colorful image...
...'forward' thrust being dispersed 'sideways' instead of bouncing clear to the opposite end of the frustum? -
#2075 by rfmwguy on 25 Sep, 2015 01:08
-
NSF-1701 Update - Below is the walkaround I did this afternoon on the improved test stand in preparation for Flight Test 2B tonight. Only glitch I have is the old, slow PC just can't handle screen recording software without pushing the lame microprocessor to max...rather than re-doing the computer (it works fine for data logging) I will simply record the video monitor. I will also record the data in the software and will make it available as a separate attachment. So here you go, the results of 3 months of building to get to this point. New test stand walkaround:
Flight Test 2B coming later tonight. -
#2076 by rfmwguy on 25 Sep, 2015 01:54
-
NSF-1701 Update - I have just finished Flight Test 2B at 9:45 PM EST September 24th, 2015 and am uploading the video now. While I was not able to screen capture due to a slow computer, I taped the video monitor and recorded the serial data which I will make available as an attachment soon. I would like to tell you the results now, but I think you should watch the video as the results unfold. I'll post the video link here after the upload...about 90 minutes from now. I am prepared to draw a conclusion about my Electromagnetic Engine Experiment. Some may like it, some may not. Get ready for either.
-
#2077 by TheTraveller on 25 Sep, 2015 04:04
-
Recent email corro with Roger Shawyer.
There is a lot of real engineering information and solid answers in this exchange.
The difference between the loaded and unloaded Q Roger has shared with an ideal injection match is new information and is why I believe SPR are now doing pulsed injection as after the Rf injection period is over (and the associated injection caused phase distortion) the cavity is them able to form a non phase distorted resonant condition, doubling the cavity Q (going from loaded Q to unloaded Q) and Force generation.
Roger makes this clear in his statement about the Flight Thruster and how the Force generated matches the doubled unloaded Q and not the 1/2 lower loaded, Rf injection period Q.
So as Roger explains in his latest patent application, the main Force is generated AFTER the Rf injection is switched off. Why? Because after the Rf pulse stops, the cavity Q doubles as it can now operate in an unloaded mode.
Nice breadcrumbs Roger. You have answered a lot of my questions.
Thanks.
> Hi Roger,
>
> I'm home again, recovering from the effects of the super bug I picked up in
> hospital. Have a deep wound, caused by the super bug, where one of the
> robot arms entered by abdomen. Like a 2nd belly button. Docs want me to
> take it very gentle until it starts to naturally regrow. They don't want to
> close it up just yet, just to be sure to not close with some of the
> superbug bacteria still active inside. So for now I have a patch over my
> 2nd belly button and daily nurse visits.
>
> So some time before I'm allowed in my workshop. But not wasting the time.
>
> Most of the software for the Raspberry PI 2B based control & monitoring
> system has been written. Will have a high power 5GHz WiFi based USB
> connection for control & data logging from the rotary table to my laptop.
>
> Point of discussion with others, NSF and myself. How do you / SPR measure
> frustum bandwidth / Q? Do you use bandwidth at 3dB away from max S11 rtn
> loss dB freq (as Prof Yang does) or 3dBs from 0 dB ref level as
> Eagleworks / NSF do or do you guys use another method? Would prefer to
> measure Q as you guys do. Makes a massive difference in bandwidth, Q &
> projected Force generation.
> Hi TT
>
> Using a network analyser to determine the Q of a high Q cavity via S11 is
> quick but requires a good understanding of the effect of scan speed and
> detector bandwidth or the results can be wildly inaccurate.
>
> For the delivery data of space qualified equipment, a calibrated procedure
> using separate signal source and power meter was usually specified. This is
> the technique we have always used at SPR.
>
> The signal source is put on a slow sweep to allow a dwell time of at least
> 10X time constant at each measurement point. The internal cavity power is
> measured using a cavity wall mounted detector measuring at least 30dB down
> (a very short probe!) and a wide detection bandwidth. The data is processed
> to measure the bandwidth 3dB down from max power level for at least two
> scans, one in each direction. (they must agree or there is a drift error in
> the measurement).
>
> Hope this helps.
>
> Best regards
>
> Roger
> Hi Roger,
>
> Thanks.
>
> So you prefer 2 port S21 over 1 port S11 or is S11 ok if the scan speed is
> VERY slow?
>
> I do understand the need for a SLOW scan and the frustum fill time / TC
> factor. What you shared is in line with my understanding.
>
> May I share this information on NSF cause there are a few arm chair
> "experts" that are strongly saying to measure bandwidth 3dB from the rtn
> loss 0dB reference level, which to me is madness.
> Hi TT
>
> S11 measurements are complicated if a highly tuned input circuit is used. I have seen published measurements that are clearly the Q of the input circuit only.
>
> Note that all Q measurements are Loaded Q measurements, and strictly speaking, if a perfect match is achieved, the actual unloaded Q is twice the measured Q.
>
> There are endless papers on the complications of Q measurement which is why smart customers are very careful about deliverable test data.
>
> The thrust predicted by the Thrust equation assumes unloaded Q. The measured thrust for the Flight thruster was very close to the predicted thrust when twice the measured Q was used in the equation.
>
> The design was successfully sold to Boeing on the measured thrust data.
>
> Feel free to share my comments on your forum.
>
> Best regards
>
> Roger
-
#2078 by rfmwguy on 25 Sep, 2015 05:02
-
NSF-1701 Update - Flight Test 2B is now available to view. The results speak for themselves. I uploaded raw, unedited camcorder video right from the camera. Nothing was done to it. Some might want to download it and enhance it for better viewing on your PC.
Here's where I want to say that I thank everyone who has given me positive support and critiques along the way. Special thanks to Doc Rodal and Chris Bergen for providing this forum which has been an invaluable resource. Thanks to all my fellow builders for never giving up hope.
But enough with my late night ramblings, see the video and evidence for yourself. It is completely uploaded in HD and is now ready to view:
-
#2079 by rfmwguy on 25 Sep, 2015 05:52
-
I guess I can say now that after more than a year in planning, building and testing, I've proven to myself that there is something to this and we need to continue to pursue it. I consider the test positive with a force against thermal lift of approximately 18 mg average. Not every mag fire up gave that reading since often it simply held back lift, but I believe this is a good approximation. This should help others scale their test setups to be able to measure this small amount of force against a much stronger thermal lift. Preheat and go for it
