There was a problem with the signal generator and pre-amplifier. They were accidentally shorted when a faulty 5V regulator was being installed to help mitigate displacement noise. While an identical spare pre-amplifier is on-hand, a new signal generator will not arrive from China until March 21 at the earliest.
Big end RADIUS: Bd = wb = 0.30 m
Small end RADIUS: Sd = ws = 0.15 m
Height: L = 0.26 m
Thank you.
Can you also calculate the Q?
There was a problem with the signal generator and pre-amplifier. They were accidentally shorted when a faulty 5V regulator was being installed to help mitigate displacement noise. While an identical spare pre-amplifier is on-hand, a new signal generator will not arrive from China until March 21 at the earliest.
How awful, my condolences.
You sure the generator's dead? Did it get over-voltaged? Perhaps it can be fixed. If you'd like some assistance with troubleshooting it, perhaps we can find a schematic and I or others can give you tips, as you could probably get replacement parts in a day or two, if you can do surface mount stuff.
You sure the generator's dead? Did it get over-voltaged? Perhaps it can be fixed. If you'd like some assistance with troubleshooting it, perhaps we can find a schematic and I or others can give you tips, as you could probably get replacement parts in a day or two, if you can do surface mount stuff.
It was over-voltaged. Fortunately it is only a $60 component, but it takes a while to replace. I can get a picture of the board later tomorrow. This is the unit:
https://tinyurl.com/zkcay2g
There was a problem with the signal generator and pre-amplifier. They were accidentally shorted when a faulty 5V regulator was being installed to help mitigate displacement noise. While an identical spare pre-amplifier is on-hand, a new signal generator will not arrive from China until March 21 at the earliest.
These things happen.
Maybe this is a good moment to ask if you've designed an experimental protocol yet. Given the nature of what you are investigating, you should be aiming for "5 sigma" results. That's never easy and it requires a very thoughtful, methodical approach. Building the apparatus is only half the battle. You've also got to work out -- and meticulously document -- how to perform an experiment that will stand the kind of scrutiny your work will face.
Once you have used your methodical approach to eliminate all possible sources of noise and experimental artifact, I am 99% certain your signal will entirely disappear.
I'm 100% certain that you are incorrect and I'm 100% certain Jaime will prove Roger is correct.
Real world coal face experimental data trumps arm chair wannabe theory every time.
"Real world coal face experimental data trumps arm chair wannabe theory every time."This is true and yet until you have credible real world data that can be independently verified, even your 100% certainty.., reduces to arm chair wannabe arm waving....
And again NO ONE has yet to present an entirely credible, as in provable theory of operation. Some good work has been presented on several models to describe and sometimes predict results, but even those wait for credible data to back up the arm waving...
Phil, I have been, or was optimistic enough to have contributed a significant (from my perspective) monetary contribution to the DIY process. Of late I am somewhat less optimistic. Likely the result of the dry spell in shared credible data...
Still everyone must recognize that there are contributors to these discussions on both sides of the debate, both practical and theoretical. In fact as far as developing testing systems that may prove one way or the other the skeptics may have provided more productive feed back and suggestions than their counterparts.
Once you have used your methodical approach to eliminate all possible sources of noise and experimental artifact, I am 99% certain your signal will entirely disappear.
I'm 100% certain that you are incorrect and I'm 100% certain Jaime will prove Roger is correct.
Real world coal face experimental data trumps arm chair wannabe theory every time.
"Real world coal face experimental data trumps arm chair wannabe theory every time."
This is true and yet until you have credible real world data that can be independently verified, even your 100% certainty.., reduces to arm chair wannabe arm waving....
And again NO ONE has yet to present an entirely credible, as in provable theory of operation. Some good work has been presented on several models to describe and sometimes predict results, but even those wait for credible data to back up the arm waving...
Phil, I have been, or was optimistic enough to have contributed a significant (from my perspective) monetary contribution to the DIY process. Of late I am somewhat less optimistic. Likely the result of the dry spell in shared credible data...
Still everyone must recognize that there are contributors to these discussions on both sides of the debate, both practical and theoretical. In fact as far as developing testing systems that may prove one way or the other the skeptics may have provided more productive feed back and suggestions than their counterparts.
I agree with this statement. I think the debate got more civil lately. But I am more affraid now that there are elements that want to push the research on one way or the other to prove their opinion. I can not see that as a scientific neutral approach, that was preferd by both sides. I think we should return to that approach.
Once you have used your methodical approach to eliminate all possible sources of noise and experimental artifact, I am 99% certain your signal will entirely disappear.
I'm 100% certain that you are incorrect and I'm 100% certain Jaime will prove Roger is correct.
Real world coal face experimental data trumps arm chair wannabe theory every time.
"Real world coal face experimental data trumps arm chair wannabe theory every time."
This is true and yet until you have credible real world data that can be independently verified, even your 100% certainty.., reduces to arm chair wannabe arm waving....
And again NO ONE has yet to present an entirely credible, as in provable theory of operation. Some good work has been presented on several models to describe and sometimes predict results, but even those wait for credible data to back up the arm waving...
Phil, I have been, or was optimistic enough to have contributed a significant (from my perspective) monetary contribution to the DIY process. Of late I am somewhat less optimistic. Likely the result of the dry spell in shared credible data...
Still everyone must recognize that there are contributors to these discussions on both sides of the debate, both practical and theoretical. In fact as far as developing testing systems that may prove one way or the other the skeptics may have provided more productive feed back and suggestions than their counterparts.
Jamie's replication will be the 14th I know of. I have no doubt that if Jamie follows Roger's breadcrumbs (Roger has been in contact with Jamie) he will be successful.
How many more replications will it take for the fence sitters to decide to get off the fence and accept that the accumulation of experimental data clearly shows the EmDrive does work as Roger has been claiming since 1999?
One thing for sure is Jamie's data will upset almost all of the theory guys here who have yet to use their theories to explain and produce force calculations to back them up for the outstanding work Paul did while he was at EW. Which BTW the way destroys the guys who claims:
It's Lorentz.
It's thermal CG shift.
Both of which can clearly be see to have be involved but are also clearly not the cause of the force generated by the EmDrive.
You sure the generator's dead? Did it get over-voltaged? Perhaps it can be fixed. If you'd like some assistance with troubleshooting it, perhaps we can find a schematic and I or others can give you tips, as you could probably get replacement parts in a day or two, if you can do surface mount stuff.
It was over-voltaged. Fortunately it is only a $60 component, but it takes a while to replace. I can get a picture of the board later tomorrow. This is the unit: https://tinyurl.com/zkcay2g
Found it here (seems to be the same)
http://preview.tinyurl.com/jjlwb63 maybe Amazon could ship it faster ?
[edit]
another possible source
http://preview.tinyurl.com/zj8bbxc[edit #2]
Sounds like that critter is also known as NWT4000
http://preview.tinyurl.com/zpfjaf9http://preview.tinyurl.com/zmgt48eand searching with that name, sounds like there are other critters and schematics too (so it may be possible to revive your dead box)
Found it here (seems to be the same) http://preview.tinyurl.com/jjlwb63 maybe Amazon could ship it faster ?
[edit]
another possible source
http://preview.tinyurl.com/zj8bbxc
[edit #2]
Sounds like that critter is also known as NWT4000
http://preview.tinyurl.com/zpfjaf9
http://preview.tinyurl.com/zmgt48e
and searching with that name, sounds like there are other critters and schematics too (so it may be possible to revive your dead box)
That's it, but through Amazon the delivery date is even worse: Arrives between Mar. 24 - April 3.
Price is also $30 higher. I was able to pick up another unit for $59 through ebay. The seller removed the listing after I purchased another unit. Perhaps they did so to increase the price to be in-line with the $89 I see elsewhere. But I have a tracking number, so the purchase went through.
I can also use the miniVNATiny as a signal generator, but it requires two pre-amps since its output is -20dBm compared to -5dBm of the chinese signal generator. So I should be able to resume some testing by Thursday. Albeit much more carefully as the VNA is a $600 piece of equipment!
What is interesting to note is that eq. (14) predicts more force for very pointed frustra whose big end radii are greater that their heights.
This is strange as Pr Igor Kaporin of the Russian Academy of Science found an optimization relation using the very same equation 14 in McCulloch's paper, which I pointed out in this post, showing an optimized frustum according to MiHsC would on the contrary be either extremely long, or not so long but very pointy, and always with a height greater than its big diameter.
Is there a flaw in McCulloch's equations, where diameters would have sometimes been mistaken for radii, screwing the optimized geometry predictions (a similar mistake between radii and diameters was found in Tajmar's paper about the EmDrive)? This is interesting; where is the issue?
I have done the math in WolframAlpha (I'm lazy), and I report here the results.
McCulloch's equation (14) is: F = -6PQL/c * ( 1 / (L+4*ws) - 1 / (L+4wb) ).
We have theree independent variables: L (length), ws (small radius), and wb (big radius).
The fact that wb and ws are radii is clear from the way McCulloch introduces equations (8) and (9), on page 4 of his paper:
[...] it is necessary to calculate the average distance between a photon at either end
and the walls. We can do this in a simplified manner by looking at this distance
along six orthogonal directions. Looking at Figure 1 and the distance from the
centre of the left-hand end plate (denoted P) to the walls: these six directions
are: to the left where the distance between P and the wall is zero, to the right
where it is L, up and down and into and out of the page where it is ws.
He then computes the average distance the Unruh waves have to resonate in at the narrow end as (0 + L + 4*ws)/6, i.e. the mean of the six degrees of movement: left, right, up, down, forward, backward.
Obvoiusly, P is the centre of the plate, and owing to the plate being round, the distance from the center to the side walls, up and down and into and out of the page, is ws. This, for me, means that ws clearly represents the radius of the small end, since from the center of the plate you can go ws meters up or down, i.e. the diameter is 2*ws. Similar reasoning can be applied for the big end case. Clearly, the figure is mislabeled, most probably table 1 too.
Coming back to the Kaporin optimization of equation (14), I note the product 6PQ/c = a, L=x, ws=y, wb=z, then F= -a*x * ( 1 / (x+4*y) - 1 / (x+4*z) ). We are interested in computing dF/dx.
Wolfram alpha gives the following rresult: dF/dx = - (4 * a (y - z) (x^2 - 16 y z))/((x + 4 y)^2 (x + 4 z)^2).
The derivative has the following roots:
a=0 (zero power or no resonance, because a=6PQ/c),
y=z (a cylinder -> ws=wb),
x^2=16xy (the result of Kaporin, i.e. L^2=16 * wb * ws, or L = 4 * (wb*ws)^0.5).
Therefore, we agree that Kaporin's optimization is ok.
Coming back to my graph, i have Sd/L (or ws/L) on the ordinate and Bd/L (or wb/L) on the abscissa.
Then from Kaporin's formula, I take L = 4 * sqrt(wb*ws) and divide by L.
Thus, I get 1=4*sqrt(wb/L * ws/L).
Now I make the variable change to coincide with my axes ws/L = y and wb/L=x. Putting this in the above equation, I get 1=4*sqrt(x*y). Taking the square, I get 1=16*x*y, or y=1/(16*x), which is a hyperbola. This is the line of optimum force, according to Kaporin.
Overimposing the plot of the hyperbola on my graph (note that the axes are the same), confirms that the force is higher for pointed frustra whose big end radii are larger that their length.
This is also intuitive to see from eq (14): F = -6PQL/c * ( 1 / (L+4*ws) - 1 / (L+4wb) ).
The overall force is negative (i.e. small end leading). Thus, on the one hand, we should strive to make the 1 / (L+4*ws) factor as big as possible, i.e. small ws. The second factor - 1 / (L+4wb) , on the other hand, is negative and ideally the denominator should be as big as possible to minimize its influence, hence wb should be as big as possible.
The optimal geometry coming out of this result is similar to https://en.wikipedia.org/wiki/Asian_conical_hat.
@X_RaY Would you be willing to conduct a set of simulations to compare optimized vs unoptimized geometry?
Edit: added eq (14) predicted force figure.
Thank you for this explanation. I asked Mike McCulloch directly with a view to get a clearer picture of it, yet…
Hi Mike,
Can you please quickly have a look at this NSF post and confirm whether ws and wb are small & big end radii or diameters of the EmDrive end plates in you MiHsC equations (including the Kaporin optimisation)?
Regards,
Ws and wb are diameters (widths). My reasoning was that the Unruh waves would exist for all six directions and fit into the space available (width) in each direction, so the up-down waves for example would fit the diameter.
Regards,
Mike
Found it here (seems to be the same) http://preview.tinyurl.com/jjlwb63 maybe Amazon could ship it faster ?
[edit]
another possible source
http://preview.tinyurl.com/zj8bbxc
[edit #2]
Sounds like that critter is also known as NWT4000
http://preview.tinyurl.com/zpfjaf9
http://preview.tinyurl.com/zmgt48e
and searching with that name, sounds like there are other critters and schematics too (so it may be possible to revive your dead box)
That's it, but through Amazon the delivery date is even worse: Arrives between Mar. 24 - April 3.
Price is also $30 higher. I was able to pick up another unit for $59 through ebay. The seller removed the listing after I purchased another unit. Perhaps they did so to increase the price to be in-line with the $89 I see elsewhere. But I have a tracking number, so the purchase went through.
I can also use the miniVNATiny as a signal generator, but it requires two pre-amps since its output is -20dBm compared to -5dBm of the chinese signal generator. So I should be able to resume some testing by Thursday. Albeit much more carefully as the VNA is a $600 piece of equipment!
Just a note; while it isn't my cup of coffee (at least not nowadays), the critter also comes as a kit (the kit includes the raw board w/o enclosure and a little more) or, willing to do so, one may even build one by himself (herself - hello SeeShells
) if wanted
Anyhow, let's wait for the replacement to come in, and I hope you'll be able to reduce the noise ... uh, given that the NWT4000 shows some noise, I wonder if it may... well, pick the schematics and have a look at the tests made on the 'net; maybe they'll be of some help (not sure but since you're at it...)
Monomorphic,
Apologies if this has already been discussed, I could not find the following mentioned:
Are your Li-Ion batteries protected or unprotected?
The protection circuitry might be a source of unwanted interference.
Monomorphic,
Apologies if this has already been discussed, I could not find the following mentioned:
Are your Li-Ion batteries protected or unprotected?
The protection circuitry might be a source of unwanted interference.
The 18650 Li-ion batteries are protected. The single 4000mAh Li-Po battery is not protected. So I can try both to see if there is a difference. But since that chemistry is different, it may be worth picking up three unprotected 18650.
Here is the board. There is no obviously fried part like on the other component. The board will communicate with a computer, but the LED will not light and the board shows as offline on their app.
Here is the board. There is no obviously fried part like on the other component. The board will communicate with a computer, but the LED will not light and the board shows as offline on their app.
The board seens to expose a TTL serial interface; please, power the board and check the gnd/+5v pins to see if they're powered, if not then probably the USB part is powered by USB but the voltage regulator for the circuit (powered by the power cable) was fried and in such a case you'll have to check it
Here is the board. There is no obviously fried part like on the other component. The board will communicate with a computer, but the LED will not light and the board shows as offline on their app.
The board seens to expose a TTL serial interface; please, power the board and check the gnd/+5v pins to see if they're powered, if not then probably the USB part is powered by USB but the voltage regulator for the circuit (powered by the power cable) was fried and in such a case you'll have to check it
This version does not have a power cable. Power is provided via USB at ~5V. This six pin power supply marked ASL3C seems a little suspect. 4.8V on one side and then 3.3V on the other. I'm not sure if that is normal for this board. Looks like I may be able to buy more. Wish the spec sheet was available. This is all I can find:
http://www.yoycart.com/Product/20612595391/The_6_foot_IC_10B8_73D42_R24_ASF9D_power/EDIT: 3.3V seems to be a standard used today for chip to chip com. So maybe not.
Silly me ... and I even looked at the board
sorry. As for schematics block ones are easy to find, but I doubt they'll help <sigh>
Maybe some amateur radio forum has some infos since it sounds like the original project came from there
Sorry to have overlooked the power stuff, now I'll wear the dunce cap and go sit in a corner
sorry. As for schematics block ones are easy to find, but I doubt they'll help <sigh>