As previously mentioned I would find a way of solving the mathematical
wall I ran in. Since then I've completely re-written the simulation
program twice, each using a unique method. The latest simulation is
far improved in that it incorporates the realism of near and far
electromagnetic fields. Again there were new problems, but not as
difficult. Today there's light at the end of the tunnel as I've
though of a new method that should accurately work. Hopefully within
a week we'll see if the new method works.
Even though this caused a huge delay in what seems several months, it
should result in far more accurate simulations.
Regards,
Paul Lowrance
That is good news! Hopefully it will lead to a working device
smoothly and quickly... :-)
Eldarion
Hi Rob,
I'll find out in a few days if the idea works, but it seems likely it
will work. If it works then the simulation should display all relevant
magnetic effects except coercively. Coercively will require perhaps
three weeks of programming. In a nutshell, the simulation is presently
limited to planar dimensions, which basically means the magnetic field
is equal at the sides of the magnetic dipole moment as compared to the
front. For that reason the magnetic material would always form
anti-ferromagnetic internal structure with no applied field; i.e., no
remanence. In real magnetic dipole moments the magnetic field in front
is twice as compared to the sides, which is why magnetic materials want
to remain saturated. In real life it's ambient temperature that
prevents the saturated core from remaining fully saturated. That's
another reason most magnetic cores are like switches at low
temperatures-- high remanence.
> Cooling of the environment would need to be dramatic if it was heat
> being converted to energy.
If we're talking about a small resistor then there's dramatic
temperature change. Compare that to a large resistor, say the size of a
basketball, while blowing air over it and you'll see it requires tens
kilowatts just to appreciably increase the temperature. Air blowing
through the fins of a regular size car radiator can easily dissipate
over 100,000 watts of continuous power.
> What are you writing the simulation software in?
Microsoft Visual C++
> What would be good is if you can prove that the load on the secondary
> has less effect on the primary when there is a magnet in place(core
> centre). Without the magnet it should behave like a standard
> transformer.
I can hardly wait to begin simulating the MEG! I'm fairly certain it
will be difficult extracting avalanche energy in the form of
electricity. I've already had time to play with simulated avalanches.
It's really cool, but so far it's apparent magnetic material is
extremely proficient at capturing it's own energy. Also from what I've
seen in the few avalanche tests so far it appears coil timing is very
critical. If coil reversal is not just precise then it's difficult to
collect some of the avalanche extra energy. Of course I was just
simulating a single wire wound toroid-- far from a MEG. Producing extra
energy in the form of heat from magnetic entropy is one thing (as in the
case of my MCE experiments), but robbing such energy from the core to
the point of creating a self-running machine while producing extra "free
energy" is another. I'm betting there are far better designs than MEG.
Although according to Naudin, Bearden, and Patrick Stephen the MEG
works. I would agree, but the only way I can see the MEG working is if
the magnetic field from the permanent magnet matches the core and
magnetic material with correct coil timing. That's why it seems so
important to precisely replicate Naudins MEGv2.1. There are still a few
issues, such as exactly what magnet did Naudin use. Also, was Naudins
AMCC 320 core annealed longitudinally or traverse, or perhaps no
annealing. Hopefully the simulation will find a design that's good
enough to work on common magnetic cores such as silicon iron. :-)
Simulations are valuable because you can precisely isolate all
parameters. You can see what's causing what, so there's no guessing
games involved. If I add a magnet or increase the magnets strength or
do this or that then I can track all the avalanche energies throughout
the entire core, focus on certain areas, track where energy is
transferred, etc. etc.
Regards,
Paul Lowrance
Gerry,
So far it appears the magnets I showed you on ebay are the best unless
you are willing to spend mega bucks for NbFeB. Can find the following
magnets any cheaper -->
You would need about 6 of them. This guy will probably give you a
discount if you buy 6.
I wouldn't spend much $ on bearings! Bearing friction will probably be
insignificant compared to brush friction. Brushes is will either kill
or make your N-machine.
I've spent more time studying the N-machine. Dr Kincheloe, a Professor
of Electrical Engineering, performed detailed measurements of Bruce's
N-machine. Rumors (from thugs) have it that Dr. Kincheloe (a Professor
of Electrical Engineering) made mistakes with the simple measurements
that any electronics 101 student could perform. As far as I can tell
there's no official retraction from Dr. Kincheloe. If anything, the
thugs threatened Kincheloe and his family.
Anyway, after studying Dr. Kincheloe detailed measurements it's apparent
that the brushes are extremely inefficient. They cause huge drags.
They heat up to extreme temperatures, which increases their resistance.
The N-machine requires extremely low resistance, and therefore to
achieve such low resistance you'll need to apply a lot of pressure on
the bushes.
I'm wondering if brush technology has improved over the past decade.
IMHO most of the focus should be on brushes.
Here are the past two emails I sent, which you probably did not get -->
Email #1:
---
Gerry,
From now on all my emails to you will start with a number. This will be
the 21st email I've sent with this particular email, which is why I'm
starting with number 21.
Gary Trump wrote:
> I just got a call from my friend out west tonight, the one I
contacted about
> getting the sample coil for you, he is looking into getting it and just
> called to let me know. He understands what the coil is for and is
interested
> in what it may be used for, but I did not explain much to him other
then it
> could be an important part of a project. His friend, Rob, who I also
know is
> looking into building the gray motor so he knows what it's all about
anyway.
Sounds good.
> I am checking into the magnet info you sent me on E mail, it looks
good, I
> wish the ID was smaller as it would be a little stronger, but all in
all it
> seems like what they used in the mark up on the file you sent me. I will
> just look for some parts and see what I can line up, I may think about
> ordering the magnets.
Actually you want a large ID. There will be hardly any voltage caused
by the inner diameter since it will be traveling at much lower
velocity. Also, the ID significantly reduces the overall magnetic
field. Ideally you want a long and narrow magnet to get higher fields.
Don't you think it would simply to fill it with aluminum? Not sure how
expensive, but Aluminum is a lot cheaper than ceramic magnet. :-)
I spoke with Andrew Mount today. I'll email you the details on our next
email exchange.
Regards,
Paul Lowrance
---
Email #2:
---
Gary Trump wrote:
> Paul,
>
> I have not seen any ring magnets on E-bay of the size we are talking
about.
> I hope to have some replays back today form the other companies I
sent some
> request to for pricing.
Gary,
I think you are missing or not reading all my emails. I sent you the
following email 4 days ago -->
-------
Gerry,
I did a little research for you. It seems NbFeB cost ~20 times more
than C8 (Ceramic 8) magnets. Forget the NbFeB!!! Too expensive. For
the same cost you could build a much much much larger N-machine with
C8's or C5's.
Here is a 6" x 3" x 1" NbFeB for $146 -->
Long url (make sure url does not wrap)
http://cgi.ebay.com/N42-6-X3-X1-NdFeB-Rare-Earth-Neodymium-Magnet_W0QQitemZ320013609687QQihZ011QQcategoryZ121837QQtcZphotoQQcmdZViewItem
Also same dealer sells a large C8 6" x 4" x 1" for $10 a peace -->
http://cgi.ebay.com/Bargain-10PC-C8-6X4X1-Ceramic-Magnet-Ferrite-Magnets_W0QQitemZ7609230070QQihZ017QQcategoryZ29402QQtcZphotoQQcmdZViewItem
Don't buy anything yet, LOL. Still lot of things to learn and decide
before buying anything. :-)
The same dealer sells 7.87" C5 diameter ring magnets by 0.787" wide. -->
http://cgi.ebay.com/Huge-Ceramic-Ring-Magnets-OD7-87-mm-X-ID3-39-X-0-787_W0QQitemZ320069745973QQihZ011QQcategoryZ29402QQtcZphotoQQcmdZViewItem
That's not close to 1 foot, but it's pretty big as far as ring magnets
go. It cost $29. You would need about 6 of those for a cost of $174 +
S&H. He charges $11.68 to ship each one, but maybe you can get a deal
if you buy 6. Let's say S&H is $10 x 6 = $60. That's a total of $174 +
$60 = $234. That's still kind of expensive. I think it would work if
you could get it spinning at least 12K rpms, but you would need to
encase the entire N-machine to keep the magnets from exploding. Perhaps
the next best task is to ask your friend what's the ***absolute***
maximum rpm he could balance an 8" diameter x 5" wide machine. Try to
find out what it would take to balance it at such rpms. Can he do it for
free? :-)
Paul
-------
---
I finally did get your E-mails, sorry my server messed up, hopefully
from now on I should not have anymore issues.
http://cgi.ebay.com/Huge-Ceramic-Ring-Magnets-OD7-87-mm-X-ID3-39-X-0-...
I contacted that person on E-bay, he had the OD @ 7.87 MM and not
inches, but finally after two E-mails to him he finally said it was a
typo and the magnets are in inches. He will give me a 5% discount so I
will get 6 of them.I am not sure how strong they are, but I am sure
they will do the job. I understand the issue with the bearings, it
would be a project later on as it would take some testing to make them
work anyway.
I agree, the brushes must be a big concern, we are asking a lot of
questions about the types of brushes and also on the low resistance.
Hopefully we can get something resolved soon so we can try it after we
get all the parts together. I seen a 1/2 thick aluminum disk on E-
bay, It may be a good way to go as it did not seem like it was very
expensive. We will pick up some aluminum round stock for the main
shaft. I wanted to first find out about what size ID the magnets were
going to be, but now I know that so next step is to get the aluminum
shaft. We are thinking of making the aluminum shaft about 24" long so
as to keep the motor at least 18" from the closest magnet, "as you
suggested".
Gerry
On Feb 5, 3:23 pm, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
> http://cgi.ebay.com/Huge-Ceramic-Ring-Magnets-OD7-87-mm-X-ID3-39-X-0-...
> ---Gary Trump wrote:
>
> > Paul,
> >
> > I have not seen any ring magnets on E-bay of the size we are talking
> about.
> > I hope to have some replays back today form the other companies I
> sent some
> > request to for pricing.
>
> Gary,
>
> I think you are missing or not reading all my emails. I sent you the
> following email 4 days ago -->
>
> -------
> Gerry,
>
> I did a little research for you. It seems NbFeB cost ~20 times more
> than C8 (Ceramic 8) magnets. Forget the NbFeB!!! Too expensive. For
> the same cost you could build a much much much larger N-machine with
> C8's or C5's.
>
> Here is a 6" x 3" x 1" NbFeB for $146 -->
> Long url (make sure url does not wrap)http://cgi.ebay.com/N42-6-X3-X1-NdFeB-Rare-Earth-Neodymium-Magnet_W0Q...
>
> Also same dealer sells a large C8 6" x 4" x 1" for $10 a peace -->http://cgi.ebay.com/Bargain-10PC-C8-6X4X1-Ceramic-Magnet-Ferrite-Magn...
>
> Don't buy anything yet, LOL. Still lot of things to learn and decide
> before buying anything. :-)
>
> The same dealer sells 7.87" C5 diameter ring magnets by 0.787" wide. -->http://cgi.ebay.com/Huge-Ceramic-Ring-Magnets-OD7-87-mm-X-ID3-39-X-0-...
That sounds good. Is that 5% discount also include S&H. :-) Sounds
like 5% on price, but you should get a big discount on S&H, no? How
does that price compare with other magnet manufacturers?
Personally I would just use ball bearings-- nothing fancy. Again, the
concerning issue is the electrical contact brushes. Ordinary brushes
will generate massive amounts of heat from friction at 12K to 13K rpms,
but you'll need those rpms. I don't know much of anything about
brushes. Although I'm certain there are electrically conductive
lubricants. There must be something better and affordable than plain
old graphite brushes? Twenty years ago mercury contacts used be the
best, but very expensive and difficult to design, so forget about
mercury. With new technology there must be something better.
Another concern is G-forces. At 13K rmps the magnets would normally
shatter. So you'll need to tightly encase the magnets. I'm thinking
cheap Aluminum will work, but how thick?
Still a lot of details to work out before buying parts, no?
Regards,
Paul Lowrance
Maybe we should make a special posting for the N motor so this is not
intermingled in the wrong topic?
I like your suggestion on encasing the magnets in aluminum, that RPM
is a lot and even though the magnets in Ceramic are a little more
solid. We need to keep in mind that the center aluminum disk will need
to have a spot for a brush contact on the top and bottom as shown in
the drawing. Maybe some sort of wrap with epoxy or some other type of
wrap over the magnets would work, anyway very good food for thought.
Talking to my friend who will balance the unit, he said that I need to
make sure that the unit is less than 60 lbs as that is the maximum for
the balancer. The aluminum bar and plate should not be that heavy.
Using bearings will be fine to start with, we can machine the OD of
the aluminum bar to fit the bearings. We have a lathe so that part is
not a problem. I can have oilers on the bearings so they will be well
lubricated if needed. I may look into ordering the round plate of
aluminum tomorrow.
I ordered the magnets tonight, the price I got was $24.56 shipping and
the cost of each magnet was $27.54 which comes to $189.80 with
shipping.The price to get the magnets from China was, shipping alone
at $35.00 per magnet. Later on when things progress it may be a good
thing to get a larger quantity from China, but not at this stage.
I thought of weighting on getting any parts, but really I need to get
going on this, it will take enough time to do things as it is. I am
like you, all of my testing and parts will come out of my pocket, I am
not about to ask for any mopney even from my friend who is very
interested in this, he will donate his time and machine work for me as
well as help on the balancing part, which is really a lot in itself.
Regards
Gerry
On Feb 5, 6:54 pm, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
Trump wrote:
> Paul,
>
> Maybe we should make a special posting for the N motor so this is not
> intermingled in the wrong topic?
So far our replies are going to the EnergyMover group under the title,
"N-machine replication"
Since you're having email problems you could check the web page for
updates -->
http://groups.google.com/group/energymover/t/aed26d1aeb22ff78
> I like your suggestion on encasing the magnets in aluminum, that RPM
> is a lot and even though the magnets in Ceramic are a little more
> solid. We need to keep in mind that the center aluminum disk will need
> to have a spot for a brush contact on the top and bottom as shown in
> the drawing. Maybe some sort of wrap with epoxy or some other type of
> wrap over the magnets would work, anyway very good food for thought.
> Talking to my friend who will balance the unit, he said that I need to
> make sure that the unit is less than 60 lbs as that is the maximum for
> the balancer. The aluminum bar and plate should not be that heavy.
On occasion I see a T.V. show where designers use some type of carbon
fiber composite material that has at least 5 times the
strength-to-weight ratio as metals. Perhaps you could post the question
in the appropriate google groups usenet-- http://deja.com I would agree
it would be difficult finding a hollow Al disk the precise size to
encase the magnets, lol. The only problem is you'll need to leave the
center Al disk exposed for the brush since that will be one of the
electrical terminals and the other being the center Al bar. If you post
the question then mention your design details-- rpm, diameter, etc.
Just tell them it's a high current generator. No need to alarm them by
claiming a "free energy" machine. Here are some videos of carbon
composite at google videos -->
http://video.google.com/videoplay?docid=8801536764731290932&q=%22Carbon+Fiber%22&hl=en
Here's info on carbon fiber -->
http://en.wikipedia.org/wiki/Carbon_fiber
See the picture how carbon fiber is cloth woven material.
Other info
http://www.sheldonbrown.com/rinard/carbonqa.htm
If carbon fiber is expensive then there must be a cheaper replacement.
So here's what we have so far. Six C5 magnets-- 3 on each side of the
Al disk. The C5 magnets are 7.87" think, but you'll need to verify that
when you get the magnets, *if* those are the magnets you're going to
buy. So the magnets are 7.87" OD. The Al disk is sandwiched between
the magnets. Perhaps it is better to make the Al disk stick out about
one inch, which means the magnet would be 8.78" in diameter. I'm not
certain, but that might help secure the magnets with whatever you're
going to wrap them in. The extra 1" Al might provide some extra surface
to bond to. Now that I think of it, I don't see it any other way.
Although there's nothing magic about 1". Actually, we're talking about
diameter, not radius, so 1" would only provide 1/2" of Al sticking out.
I guess 1/2" is fine.
OK, one 8.78" Al disk sandwiched between 6 7.87" C5 magnets, with a
3.39" Al bar going through the center. I created and uploaded a picture
for you -->
http://groups.google.com/group/energymover/web/n-machine.gif
Also here's a text picture -->
XXXAXXXXXXXXXXXXXXX
CCCACCCXXXXXXXXXXXX
CCCACCCXXXXXXXXXXXX
CCCACCCXXXXXXXXXXXX
AAAAAAAXXXXXXXXXXXB
AAAAAAAAAAAAAAAAAAB
AAAAAAAXXXXXXXXXXXB
CCCACCCXXXXXXXXXXXX
CCCACCCXXXXXXXXXXXX
CCCACCCXXXXXXXXXXXX
XXXAXXXXXXXXXXXXXXX
XXXBXXXXXXXXXXXXXXX
X = empty space
A = Aluminum disk
C = C5 magnets
B = Electrically conducting Brush
Notice to two brushes (blue). Depending which direction you rotate the
N-machine, one will be + and the other -. So you will place a low
resistance resistor and thick copper wire between these two brushes.
Notice how long Al rod is. IMHO this is very important. We know from
Bruce Depalma's gyro experiments there's something happening to
space/inertia/mass in and near a rotating mass. IMHO it is important
for the returning electricity to stay as far apart from the rotating
n-machine as possible. I'm *guessing* one or two feet away from the
N-machine is sufficient.
Notice how everything rotates, except the ball bearings. Also the
resistor and thick copper wire that's connected to the brushes will not
rotate of course.
> Using bearings will be fine to start with, we can machine the OD of
> the aluminum bar to fit the bearings. We have a lathe so that part is
> not a problem. I can have oilers on the bearings so they will be well
> lubricated if needed. I may look into ordering the round plate of
> aluminum tomorrow.
>
>
>
> I ordered the magnets tonight, the price I got was $24.56 shipping and
> the cost of each magnet was $27.54 which comes to $189.80 with
> shipping.The price to get the magnets from China was, shipping alone
> at $35.00 per magnet. Later on when things progress it may be a good
> thing to get a larger quantity from China, but not at this stage.
Sounds good, but don't forget the most important aspect ***by far***,
the brushes. I've read report after report how the brushes can either
make or kill an N-machine. The amount of wasted heat caused by friction
will be difficult to overcome. Here's what the brushes need to excel at -->
1. Extremely low friction.
2. Extremely low resistance.
I'm uncertain which of the two above is most important, but they're both
probably more critical than even the magnet strength. Have you
researched the latest electrical brush technology?
Regards,
Paul Lowrance
I have not researched the electrical brush technology enough yet, I
have just asked some questions to some people that should have an idea
on what would be the best way to do it. One person whom I talked to
said it would be much better if the outer aluminum disk would have
notches in it like a gap, that would reduce the heat and drag
efficiency's a lot. Understanding what you are saying about the drag
and heat, I feel that I have a lot to learn on the best way to
approach the matter.
"Have you researched the latest electrical brush technology? "
I like the idea of using the carbon fiber is cloth, again that seems
to be a very worthwhile choice, I am just concerned on any flexing
during the rotation, I have worked around that stuff in the past, but
I can ask some questions. Maybe I can get a large piece of heat shrink
and just put it over the magnets and shrink it down? Many options to
look into, that is whey I felt like I needed to get going on getting
the parts. Time will pass fast and I want to have some items together
so I can work on things. Even the mounting plate that will hold the
unit, I thought of getting a piece of aluminum flat stock and the
putting some rubber insulators between the bearings and the unit just
to isolate the unit from the mounting plate. Hopefully the magnets
will be here in a few days so I can check them out and then get the
aluminum disk, I may order that today if the boss will let he :)
I have two electric motors, one of them is DC, but not sure how much
HP it will take to run the machine so I will keep my options open in
that area. I can get other motors as they seem inexpensive around
here.
Regards
Gerry
On Feb 6, 11:13 am, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
> Hi Trump,
>
> Trump wrote:
>
> > Paul,
> >
> > Maybe we should make a special posting for the N motor so this is not
> > intermingled in the wrong topic?
>
> So far our replies are going to the EnergyMover group under the title,
> "N-machine replication"
>
> Since you're having email problems you could check the web page for
> updates -->
>
> http://groups.google.com/group/energymover/t/aed26d1aeb22ff78
>
> > I like your suggestion on encasing the magnets in aluminum, that RPM
> > is a lot and even though the magnets in Ceramic are a little more
> > solid. We need to keep in mind that the center aluminum disk will need
> > to have a spot for a brush contact on the top and bottom as shown in
> > the drawing. Maybe some sort of wrap with epoxy or some other type of
> > wrap over the magnets would work, anyway very good food for thought.
> > Talking to my friend who will balance the unit, he said that I need to
> > make sure that the unit is less than 60 lbs as that is the maximum for
> > the balancer. The aluminum bar and plate should not be that heavy.
>
> On occasion I see a T.V. show where designers use some type of carbon
> fiber composite material that has at least 5 times the
> strength-to-weight ratio as metals. Perhaps you could post the question
> in the appropriate google groups usenet--http://deja.com I would agree
> it would be difficult finding a hollow Al disk the precise size to
> encase the magnets, lol. The only problem is you'll need to leave the
> center Al disk exposed for the brush since that will be one of the
> electrical terminals and the other being the center Al bar. If you post
> the question then mention your design details-- rpm, diameter, etc.
> Just tell them it's a high current generator. No need to alarm them by
> claiming a "free energy" machine. Here are some videos of carbon
> composite at google videos -->
>
> http://video.google.com/videoplay?docid=8801536764731290932&q=%22Carb...
>
> http://video.google.com/url?vidurl=http%3A%2F%2Fwww.youtube.com%2Fwat...
>
> http://video.google.com/url?vidurl=http%3A%2F%2Fwww.youtube.com%2Fwat...
>
> http://video.google.com/url?vidurl=http%3A%2F%2Fwww.youtube.com%2Fwat...
>
> Here's info on carbon fiber -->
>
> http://en.wikipedia.org/wiki/Carbon_fiber
>
> See the picture how carbon fiber is cloth woven material.
>
> Other infohttp://www.sheldonbrown.com/rinard/carbonqa.htm
I don't know about that! I'm guestimating this machine at 13K rpm could
be close to 10000 G's! So what ever you use should be able to hold say
250000 pounds. You don't want $190 worth of magnets going down the drain.
Did you see the image in my previous post?
http://groups.google.com/group/energymover/web/n-machine.gif
> I have two electric motors, one of them is DC, but not sure how much
> HP it will take to run the machine so I will keep my options open in
> that area. I can get other motors as they seem inexpensive around
> here.
I'm wondering if AC motors are more efficient. I saw a new 1/2 DC motor
for $20 at the store last week. DC motors have a lot of friction due to
brushes.
Paul
Did you already buy the magnets?
Paul Lowrance
I did see your "gif "on the motor with the colors. I will make sure
that the aluminum disk is larger around then the magnets by at least
1".All of the info like that I keep in a file so I can reference it.
The 10,000 G's with 250,000 pounds? that will take a good deal of
thought now. I will try and get some calculations on things. Really
who knows what the RPM will need to be, but also I understand that the
unit will need to be able to go to that RPM if needed. I can look
around as I know that there are a lot of brushless AC motors on the
market, maybe I can find one that will fit the needs. I will just
guess at the HP of the motor as I am not sure what will be needed, I
am going to look for at least a 1/2 hp motor. we have some thoughts of
gearing already to that is a plus.
I am not going to skimp on this project, I want to built a unit that
has a chance to do what we want it to do. My budget may stretch the
timing out a little, but I will do what I can to keep the project
going smoothly, with your help Paul.
I was away from the house most of the day today so I did not get a
chance to post anything on the brushes, but hopefully I can
tomorrow.
Regards,
Gerry
On Feb 6, 12:23 pm, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
> Trump wrote:
>
> > Paul,
> >
> > I have not researched the electrical brush technology enough yet, I
> > have just asked some questions to some people that should have an idea
> > on what would be the best way to do it. One person whom I talked to
> > said it would be much better if the outer aluminum disk would have
> > notches in it like a gap, that would reduce the heat and drag
> > efficiency's a lot. Understanding what you are saying about the drag
> > and heat, I feel that I have a lot to learn on the best way to
> > approach the matter.
> >
> > "Have you researched the latest electrical brush technology? "
> >
> > I like the idea of using the carbon fiber is cloth, again that seems
> > to be a very worthwhile choice, I am just concerned on any flexing
> > during the rotation, I have worked around that stuff in the past, but
> > I can ask some questions. Maybe I can get a large piece of heat shrink
> > and just put it over the magnets and shrink it down?
>
> I don't know about that! I'm guestimating this machine at 13K rpm could
> be close to 10000 G's! So what ever you use should be able to hold say
> 250000 pounds. You don't want $190 worth of magnets going down the drain.
>
> Did you see the image in my previous post?http://groups.google.com/group/energymover/web/n-machine.gif
Yes I did purchase the Magnets today. I should get a response on when
they will arrive in a day or so. I will try also to get the aluminum
disk ordered yet tonight if I can or first thing tomorrow. I am
going to do a search on some aluminum bar stock also.
Regards
Gerry
On Feb 6, 1:09 pm, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
Trump wrote:
> Paul,
>
> I did see your "gif "on the motor with the colors. I will make sure
> that the aluminum disk is larger around then the magnets by at least
> 1".All of the info like that I keep in a file so I can reference it.
> The 10,000 G's with 250,000 pounds? that will take a good deal of
> thought now. I will try and get some calculations on things.
Gerry,
My guesstimate was just a rough figure for the average G force. I just
did a quick calculation of the magnets outer rim (7.87") and at 13,341
rpm's there will be 20,000 G's!!! Of course the G force on magnet
material at half the distance from center to outer edge would be 1/2,
which is 10,000 G's.
So it seems very important that you find someone who knows how to
calculate material strengths, stresses, brittleness, etc. to make sure
your machine holds together. The magnets outer rim would be traveling
at 312 mph, roughly half the speed of sound. Sounds scary, but it's
doable. I'm not sure an outer Al casing can hold the weight of the
magnets. The magnets would normally fly outward at such force and
therefore the outer casing will need to hold all that weight inward. A
mechanical engineer or someone with this type of experience will just
have to go through the numbers. I'm sure some qualified engineer would
freely calculate this for you, but just make sure they are qualified.
Perhaps you could get a second opinion. :-)
Regards,
Paul Lowrance
The one drawing that I have shows the center disk being 3/4" thick, if
I go to a 1" thick or 1 1/2" thick disk will that be ok. I can order
both the 1" and 1 1/2 " disk, but I was not sure if going over the
3/4" would hurt the machine or not?
I think we have the magnet protection worked out, I talked to my
friend tonight and we have come to a resolution on how to cope with
that issue.
My friend is a Mechanical Engineer, he said that I will need to find
the exact weight that we are spinning to determine the force at a
specified RPM.
If I can, I will try and take some digital pictures of the machine
when I get to that point. At this point I am just trying to get some
parts in and then we will go into the machining part and then
assembly, this will take some time and I hope that things go smoothly.
The statement that Paul has made on this N machine is true, there is
no proof that this machine will produce Free Energy, it is upon my own
free willingness to try and see what this machine will do, if it does
nothing then so be it, if it does work then so be it again.
Regards,
Gerry
On Feb 7, 9:57 am, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
Gerry,
I'll have to take a look at that in FEMM. 1.5" thick Al? That sounds
think, but I'll see how much it could decrease the voltage.
Paul
We must also realize that the Magnets are a little over 3/4" thick at
0.787" and we are running three of them on each side and not two. I
will wait for your response on this before I order anything.
Gerry
On Feb 7, 10:14 pm, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
As far as I can tell, the theory behind the N-machine is rotating mass
distorts space/time. This is evident in various experiments performed
by Bruce and others. Therefore, we should try and keep the Aluminum
disk as thin as possible so long as it's resistance is low enough. The
only reason I suggested Al is because it's cheap. There's one major
drawback to Al. Al is not very dense. So I have no idea how that will
affect the over-unity aspect, which is why I wouldn't recommend thick Al
disk. Bruce used copper disk, not Al. Copper is 3.3 times more mass
per area than Aluminum. Initially I would have never suggest you buy a
copper disk because I thought copper was zillions time more expensive,
but I just looked at a present prices of metals -->
http://www.wmwa.net/metal_prices.htm
It says the present price of Copper is $2.4 / lb and Al is $1.27 / lb.
Is it true that??? That's only 1.89 time more expensive. ***--> Better
yet, look at lead. Lead is nearly half the price of Al at $0.74 /
lb??? Wow, this is entirely your decision, but if lead is cheaper than
I would use a lead disk hands down because lead is very dense!!! Really
I can't say what causes the over-unity, or least the measurements I've
read about claim over-unity, but it appears rotating mass affects
inertia and/or space/time. If true then lead should have a much bigger
over-unity effect than Al.
The only requirement is you cannot use magnetic material for the
conducting disk, which brings up another topic I wanted to discuss. I
would recommend you place spacers between the conducting disk and
magnets. Here's another updated N-machine picture I drew -->
http://groups.google.com/group/energymover/web/n-machine2.gif
The green is an electrical insulator such as epoxy. The idea is
appreciably separate the magnet from the electrical current. IMHO this
is very important. If the magnet was right up against (hugging) the
conducting disk then the electrical current in the disk would greatly
affect the magnet. This could easily realign the magnet that's close to
the current. What you could do is place a piece of cheap clear cooking
plastic in addition to epoxy. You know the plastic used to cover food.
This would guarantee they are electrically insulated and the epoxy would
keep things from moving around.
So, the price of lead is indeed cheaper than Aluminum. Here's a good
idea. Lead has a very low melting point. While in high-school my
friend and I used to go to gas stations and collect free lead tire
weights, melt them down in a cheap skillet on the stove and pour them in
molds. So you could buy some cheap lead and melt it in a cheap skillet
well ventilated area to make lead disks. Just make sure you don't cook
food in that skillet afterwards. :-) Out here they sell cheap skillets
for 99 cents.
There's only one issue. The brush cannot make contact with lead for
obvious reasons. So you'll need to place an Aluminum outer ring. In the
above image you'll see there's an outer aluminum ring in the center disk.
Don't worry about the low resistance of lead. The resistance of the
lead disk will be far lower than the brush and load resistance.
What do you think? If this is what you like and when you ready I'll
calculate the numbers.
Regards
Let me think on that message a little, I will get back to you. I am
glad that we can work things out like this prior to me getting the
wrong materials. The magnets will be here tomorrow. My fear right now
is the "weight" Led is the heavy metal, I maybe approaching an issue
on balancing the unit, they can only balance a weight of I think 65
lbs or less.
Whatever it takes to make this work I am willing to do. I see the one
picture shows copper as the center disk, but let me chew on the
thought of led a little. I will hold off on getting any disk material
right now. I need to make sure of the materials I am getting, it may
not be easy to change materials after the machine is together. I am
looking at some brushless motors to turn the machine with right now.
My friend and I have talked about the the bearings in depth, they will
need to be oil bearings and not grease bearings, but we have already
figured out a method to do that part.
Do you have any negative thoughts on the magnets I am getting? size or
width? The C5 magnets are 7.87" thick? and using three of them per
side? The original motor that I seen used two magnets on each side and
they were 3/4" thick, maybe mine will be stronger?
Would your idea be to use the LED for the shaft also, or another
material like aluminum?
Regards
Gerry
On Feb 8, 2:09 pm, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
Gerry,
We'll have to see how heavy everything is. I'm thinking the lead disk
will be relatively light compared to the 6 magnets.
If you look at the second image I posted you should get an idea what
part is lead (gray) and what's aluminum (red).
Any research on reducing brush electrical resistance and friction?
Paul
Regards
Gerry
On Feb 8, 8:45 pm, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
Exactly. That's expected. The reason is because a 1" x 1" has a much
higher length to width ratio of 1. Also that's exactly why I did not
want you to buy magnets with smaller ID because they would have been
even weaker.
According to FEMM the field of just one of your magnets should be 225
Gauss (2.35" from outer edge) to 2500 Gauss (outer edge). That's very weak.
Now place all 6 magnets together and the field will increase to 1100
(2.35" from outer edge) to 3000 Gauss (outer edge). Ceramic 5 magnets
are about 3800 Gauss, but that's in a close loop.
Paul
Well that makes me feel a little better then, so the magnets I got
should do the job. Kind of good in a way as they are easier to handle
anyway.So far I have not had the opportunity to weigh the magnets, but
I hope to have that done soon, that will give us the G forces at a
specific RPM. If you feel the Lead is the best for the center, inner
disk I can start to round up some Lead and then get some aluminum and
inlay the led and aluminum together and then have it machined to a
precise size ( OD )and thickness. I guess this is where your
understanding of what is best as far as the material to use is way
above mine. Kind of looks like some people who tried to do this in the
past made a lot of mistakes? I have been beating my self in the head
trying to think of a way to change components if needed even after I
get the unit assembled. If that is not possible then so be it and I
will just try and make sure I get the best recommended parts assembled
first time.
How is the MEG coming ?
Regards
Gerry
On Feb 9, 8:02 pm, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
Hi Gerry,
I think your C5 magnets are the best for your budget, as NdFeB are
extremely expensive. I've never seen a NdFeB magnet the size of your
C5's. So you would have to buy lot of smaller NdFeB's and bond them
together. Can you imagine the cost of a NdFeB magnet that's results in
nearly 8" OD by nearly 5" wide, lol? Bruce Depalma had to use special
equipment to assemble NdFeB magnets. It's very dangerous. The only
thing you need to be concerned about is being careful not to break your
C5 magnets.
As far as voltage, I think you should get around 3 volts at 13,000
rpms. When the machine is complete, do you have a way of knowing the rpm?
Have you researched brushes? Some things to consider is what type of
metal to use for the outer conduction ring. You'll need to find the
metal that has the lowest coefficient of kinetic friction for whatever
type of brush you use. The outer conduction ring I'm referring to is
found in the second n-machine image posted -->
http://groups.google.com/group/energymover/web/n-machine2.gif
The outer red part (Aluminum) next to the gray and brown.
I'm looking at a table and it shows Cast iron (not Steel) has a low
coefficient of kinetic friction. I'm sure you can find materials with
even lower friction. Therefore, it make a huge difference if you coated
the outer conduction ring with such material. It could decrease your
brush friction by 3 to 5 times less!!
Again, you want lowest brush friction *and* lowest electrical brush
contact resistance.
Regards,
Paul Lowrance
3 Volts at 13K does not seem like a lot of usable voltage? Do we have
any idea if that is the RPM, Bruce Depalma ran his machine at?. Did
Bruce also get about the same voltage, I did understand he had a lot
of amps though?.
They have Tachometers that you just focus on a spot and it will read
out the RPM I am not sure on the cost of them, but that is what I am
looking into at this time..
I have not dug into the brushes much as I was waiting for another
person I know to get back with me as he has some knowledge on brushes.
Maybe I should just post some questions on the Internet or try and
make some phone calls to some places that handle brushes.
I like the idea of coating the outer area of the disk for the brush
contact area with a coating that is low in coefficient kinetic
friction. To me that would make the most sense, I could make the disk
out of let's say aluminum and then have the brush contact area coated
with what ever coating that would fit in the category of the lowest
coefficient of kinetic friction.
So far I have managed to estimate the weight of the magnets. Not sure
on how accurate it is, but it may start to give an idea on the total
weight of the magnets. I will try and get a more precise weight as
soon as I can. The packaging was 5 lbs and the total shipping weight
was 42lbs. So 37 total pounds for the magnets divided by 6 = 6.16 lbs.
each.
Gerry
On Feb 11, 10:30 am, "energymo...@gmail.com" <energymo...@gmail.com>
wrote:
You have tried to guide me though the steps and I really appreciate it
a great deal, but I don't feel that this project is meant for me to
do. I will keep cheering people on who have the knowledge to make
things like this work, but that is for sure not me.
Regards
Trump
With regards to the brushes, do the brushes need to contact the outer
edge of the disk for this effect to work?
What I am thinking is that you should be able to take a series of
wires from the edge to disk at say 10 degree intervals and bring this
around the casing to the axle.
Then pick up the potential via a brush contact.
See the changes I have made to Pauls diagram (n-machine4.gif).
3 volts is fine provided you have the current to go with it.
You could even build a DC-DC converter into the rotating disk to
increase the voltage to say 24V and a lot less current and that could
be better managed by the brushes.
13000rpm is a lot of rpms, I think I have some Le Mans racing motors
rated at 15000 rpm free running.
Regards
Rob