Storing Cookies (See : http://ec.europa.eu/ipg/basics/legal/cookies/index_en.htm ) help us to bring you our services at overunity.com . If you use this website and our services you declare yourself okay with using cookies .More Infos here:
https://overunity.com/5553/privacy-policy/
If you do not agree with storing cookies, please LEAVE this website now. From the 25th of May 2018, every existing user has to accept the GDPR agreement at first login. If a user is unwilling to accept the GDPR, he should email us and request to erase his account. Many thanks for your understanding

User Menu

Custom Search

Author Topic: Empirical equations predicting Newman Motor performance  (Read 7210 times)

kmarinas86

  • Full Member
  • ***
  • Posts: 156
    • YouTube - kmarinas86's Channel
Empirical equations predicting Newman Motor performance
« on: December 14, 2009, 08:33:54 AM »
On my YouTube channel page, I have long ago renamed the videos so that the data appear in the titles:

http://www.youtube.com/user/kmarinas86

I have made comparisons between several of the motors. Specifically, I looked at six videos meant to achieve the following:

1) To compare the change in performance between one and two coils.
2) To compare the change in performance between two and three coils.
3) To compare the change in performance between three and four coils.

From this I deduced that when minimal other modifications are done:

For each coil added (4000 ft of 28 AWG wire):

For a given voltage, the RPM drops to 90% of the original value.
For a given voltage, The Current / Volt drops to 50% of the original value.
For a given voltage, the power output drops by (90%)^3 or 72.9% of the original value.

From there, I have developed an empirical method to predict what will happen once I get around to adding more coils:

n= newly added coils (past the current 4 coils)
v= voltage increase multiplier (1, or 100%, means the same voltage)

Note that for conventional technology, (n+4) at this power level is VERY small. It seems, however, that the value of (n+4) is only valid so as long as the wire gauge is the same. It may turn out that there is much more to it than just the number of coils when we change the gauge of the wire. Remember that empirical formulas are only estimates of what is going on. There is not much understanding right now as to the emergence of the relationships.

________________________

Adding a coil

(.9)^3/.5 = change in power out / change in power in
1.458 = change in power out / change in power in

Adding n coils:

[(.9)^3/.5]^n = change in power out / change in power in
1.458^n = change in power out / change in power in

________________________

Cooling effect

n*(backspike current)^2

________________________

Output Power

[(.9)^3]^n*v^3
[(.9)^n]^3*v^3
[v*(.9)^n]^3

If the machine is to operate at the same power output as before:

v^3 = [(10/9)^3]^n
v = [(10/9)^3]^(n/3)

If n=2, v=1.23

________________________

Input power:

(.5)^n*v^2

If the machine is to operate at the same power input as before:

v^2 = 2^n
v = 2^(n/2)

If n=2, v=2

________________________

Efficiency:

[2*(.9)^3]^n*v

[1.458]^n*v

If the machine is to operate at the same efficiency as before:

v = [1/1.458]^n

If n=2, v=0.4704

________________________

Cooling effect / Output power

[n*(backspike current)^2] / [[(.9)^3]^n*v^3]

[n/[(.9)^3]^n]]*[(backspike current)^2/v^3]
[n/[(.9)^3n]]*[(backspike current)^2/v^3]
[n/[(10/9)^(-1)^3n]]*[(backspike current)^2/v^3]
[n/[(10/9)^(-3n)]]*[(backspike current)^2/v^3]
[n*[(10/9)^3n]]*[(backspike current)^2/v^3]

For large n, n*[(10/9)^3n] rises faster than v^3 even if n is proportional to v.

________________________

Input current density (without cooling effect) (wire gauge unchanged):

(.5)^n*v

Restore to original input current density (awg rating unchanged):

2^n = v

If n=2, v=4

________________________

Output power density:

[(.9)^3]^n*v^3 / [n+4]

4 = original coil number

________________________

Output power density / Input current density ::

[[(.9)^3]^n*v^3 / [n+4]]/[(.5)^n*v]

[[2*(.9)^3]^n*v^2] / [n+4]

[1.458]^n*v^2 / [n+4]

If input current density is kept the same, the following would be the projected power density of the device:

[1.458]^n*[2^n]^2 / [n+4]
[1.458]^n*[2^(2n)] / [n+4]
[1.458]^n*[4^n] / [n+4]
[5.832]^n / [n+4]

Here it is graphed:

-3   0.0050413570150648400   0.02
-2   0.0147005970559291000   0.06
-1   0.0571559213534522000   0.23
0   0.2500000000000000000   1
1   1.1664000000000000000   4.67
2   5.6687040000000000000   22.67

1st column = n
2nd column = [5.832]^n / [n+4]
3rd column = 2nd column "normalized" (n=0)

rows 1 to 6: from 1 coil to 6 coils

jadaro2600

  • Hero Member
  • *****
  • Posts: 1257
Re: Empirical equations predicting Newman Motor performance
« Reply #1 on: December 14, 2009, 02:13:15 PM »
I don't know Stefan hasn't enabled the LaTeX plugin for SMF..  there are, after all, maths being used here. :)

See if you can get flow to do something like this:

---->
<-^--

Yay ascii art! ...  this translates, in my periodic mind, to voltage flowing backwards over current!  Think about clouds!

Meanwhile, I'm going to try to square the circle.

guruji

  • Sr. Member
  • ****
  • Posts: 470
    • http://andyborg.tripod.com
Re: Empirical equations predicting Newman Motor performance
« Reply #2 on: December 19, 2009, 07:56:04 PM »
Hi guys I am building a Newman Motor and need a little help regarding the rotor. I am doing a circular round wood rotor and want to install eight magnets on it. What polarity should the magnets be on the rotor four N and four S?
Any help please?
Thanks

kmarinas86

  • Full Member
  • ***
  • Posts: 156
    • YouTube - kmarinas86's Channel
Re: Empirical equations predicting Newman Motor performance
« Reply #3 on: December 19, 2009, 11:41:55 PM »
Hi guys I am building a Newman Motor and need a little help regarding the rotor. I am doing a circular round wood rotor and want to install eight magnets on it. What polarity should the magnets be on the rotor four N and four S?
Any help please?
Thanks

Make a giant magnet out of those eight you got.

guruji

  • Sr. Member
  • ****
  • Posts: 470
    • http://andyborg.tripod.com
Re: Empirical equations predicting Newman Motor performance
« Reply #4 on: December 20, 2009, 09:34:41 PM »
Make a giant magnet out of those eight you got.
Hi Kmarina thanks for response. Do you mean I should do four on one side and four on the other side?
Can I place them evenly on a round rotor with space between NNNN SSSS or it does not work that way?
Thanks

jadaro2600

  • Hero Member
  • *****
  • Posts: 1257
Re: Empirical equations predicting Newman Motor performance
« Reply #5 on: December 20, 2009, 11:33:03 PM »
You can regionalize the magnetic field ..four north on one side and 4 south on the other, this will create lines of force between them in an more even fashion.

kmarinas86

  • Full Member
  • ***
  • Posts: 156
    • YouTube - kmarinas86's Channel
Re: Empirical equations predicting Newman Motor performance
« Reply #6 on: December 21, 2009, 02:35:10 AM »
Hi Kmarina thanks for response. Do you mean I should do four on one side and four on the other side?
Can I place them evenly on a round rotor with space between NNNN SSSS or it does not work that way?
Thanks

It will be weaker that way, but it will still work. If you want to do this, you might as well focus on the strength of the coil to make up for it. Start with a square coil, unlike what I did. Then you could use a cylindrical column. It will look like a window motor:

http://www.youtube.com/results?search_query=window+motor&search_type=&aq=f

Keep in mind that if you use a weak magnet, your motor efficiency will be compromised. Also, if your magnets are tiny compared to your coil, you can expect your motor to run a long time, but mechanical output power will be small.

The commutator is the most important bit. You want to generate the cold spikes. The generation of cold spikes is enabled by several things:

1) Large coil thickness (important for obtaining higher inductance/resistance which will increase the impedance (i.e. voltage/current) of the flyback)
2) Higher voltage
3) Spark gap
4) Flyback current (do not confuse this with back-emf)

If any of the above items is compromised you will have a harder time achieving Newman's observations.

guruji

  • Sr. Member
  • ****
  • Posts: 470
    • http://andyborg.tripod.com
Re: Empirical equations predicting Newman Motor performance
« Reply #7 on: December 21, 2009, 12:51:49 PM »
Thanks Jadaro and Kmarina for response. Maybe if one do not include wood in the rotor and do it with metal would be better for magnetisim.
Regarding the circuit I am going to use that of Peter Lindemann.
Thanks guys Happy Christmas.

kmarinas86

  • Full Member
  • ***
  • Posts: 156
    • YouTube - kmarinas86's Channel
Re: Empirical equations predicting Newman Motor performance
« Reply #8 on: December 22, 2009, 11:21:26 PM »
http://www.youtube.com/watch?v=3KtqdUOzh2g

FAN:
Flex-a-lite Heavy Duty Division
32" 6600 Series - .625" pilot - no bolt holes

POWER:
242 volts
0.035 amps
8.5 watts

RPM:
120 RPM

Fan tip speed:
11.4 mph

kmarinas86

  • Full Member
  • ***
  • Posts: 156
    • YouTube - kmarinas86's Channel
Re: Empirical equations predicting Newman Motor performance
« Reply #9 on: December 23, 2009, 12:01:49 AM »
Also, another thing I am realizing is that the cooling ability has a non-linear relationship with the back-spikes.

I have observed "measurements" of large negative spikes over 0.7 "DC" amps on my digital Extech 411 multimeter. With this was often a lack of proportionality in the cooling effect. As we should all know from science, all current, no matter what direction, produce heat.

The energy of the flyback has basically two ways to be used up through the circuit:

1) As heat.
2) Output power through the fan.

It turns out the anomalous mechanical power output is not highest when there are large negative spikes, but rather, it happens when the current going either direction, is kept low relative to the voltage.

As I have said before, the impedance relation (i.e. voltage/current) of the flyback should be greater than that of the input.

Theory of the electric field in motor windings:

If the electric field changes rapidly, it will create a residual magnetic field that is circular around the wire. A small amount of coulombic energy, dumped for a short amount of time, is the result of a large voltage slew rate. The electric field strength is the ratio of the voltage divided by the distance over which is it diluted.

The energy density of the electric field is:

(1/2) * (efield)^2 * (electric permittivity)

Expressed in another way, this is:

(1/2) * (volts/length)^2 * (electric permittivity)

The fluctuation of the energy in the electromagnetic field is converted between electric magnetic and magnetic fields. Thus, over time:

(1/2) * (volts/length)^2 * (electric permittivity)

is proportional to the magnetic field potential energy density:

(1/2) * (amperes/length)^2 / (magnetic permeability)

Thus the field that can be generated is a function of:

(1/2) * (volts/length)^2 * (electric permittivity) :: (1/2) * (amperes/length)^2 / (magnetic permeability)
(volts/length)^2 * (electric permittivity) * (magnetic permeability) :: (amperes/length)^2
(volts/length) * ((electric permittivity) * (magnetic permeability)) :: amperes/length

Where "::" is symbol representing proportionality. In mathematics, its lowercase alpha "α" (which doesn't show up well in Tahoma font).

The volume of a wire can be related to its resistance by the following formula:

resistance = resistivity * length / cross-sectional area
resistance = resistivity * length^2 / (cross-sectional area*length)
resistance = resistivity * length^2 / volume
volume = resistivity * length^2 / resistance

Thus the energy in the electric field in a thin wire is equal to:

[(1/2) * (volts/length)^2 * (electric permittivity)] * [resistivity * length^2 / resistance]
[(1/2) * volts^2 * (electric permittivity)] * [resistivity/resistance]
[(1/2) * resistivity * (electric permittivity)] * [volts^2/resistance]

So the energy stored is determined by the three following factors:

1) resistivity
2) electric permittivity
3) volts^2/resistance

"volts^2/resistance" in an inductive circuit is not itself a measure of power, but rather is it factor of proportionality determining how much power may actually be involved, and this is relative to the resistivity of circuit and its electrical permittivity.

guruji

  • Sr. Member
  • ****
  • Posts: 470
    • http://andyborg.tripod.com
Re: Empirical equations predicting Newman Motor performance
« Reply #10 on: December 23, 2009, 09:54:25 PM »
Hi Kmarina today I tried to put five magnets together to make them all five N facing upwards. I could not do it cause they're repeling each other??.
I only have to put them one on each other to have a bigger magnet?
I was going to use a soft alloy rotor to attach the magnets to instead of the wood is it ok this?
Thanks

kmarinas86

  • Full Member
  • ***
  • Posts: 156
    • YouTube - kmarinas86's Channel
Re: Empirical equations predicting Newman Motor performance
« Reply #11 on: December 24, 2009, 04:18:10 AM »
Hi Kmarina today I tried to put five magnets together to make them all five N facing upwards. I could not do it cause they're repeling each other??.
I only have to put them one on each other to have a bigger magnet?
I was going to use a soft alloy rotor to attach the magnets to instead of the wood is it ok this?
Thanks

Yes and yes.

I use a steel shaft and square bolts and attach my magnets to that. So it is possible to use the soft alloy rotor to hold it. It might actually help with the concentration of the field lines. Wood is not a good choice in my opinion.

kmarinas86

  • Full Member
  • ***
  • Posts: 156
    • YouTube - kmarinas86's Channel
Re: Empirical equations predicting Newman Motor performance
« Reply #12 on: December 31, 2009, 02:02:43 AM »
I had developed the empirical equations above from a limited data set.

Now I have compiled much more data into excel format. As you can see, this is far beyond what I was doing just two and a half weeks ago. See attachment below (uploaded as .pdf to save space).
« Last Edit: December 31, 2009, 03:09:44 AM by kmarinas86 »