[LB7 (OP)]

If I use only a resistance, the power from the voltages sources is positive, it is logical. Here the power is negative with RL and when I divide by 10 or 100 the step, it changes nothing.

I will try to have a device with only the transcient power. It is possible to set ON/OFF each second but it could better to have a continuous device.

The circuit:

d95e3.png (53.56 kB . 679x672 - viewed 4020 times)

With the values:

V1 0 20 SIN(0 2V 100 0 120)
V11 21 1 SIN(0 20V 1 0 3)
Vk 20 21 PWL(0 -0.8 0.1 0 0.1) +r=0 td=0.2

R1 1 2 0.0002
L1 0 2 15mH

The device is more stable in the transcient.


defe.png (6.86 kB . 595x484 - viewed 3984 times)


And if I replace with a slower square:

V1 0 20 SIN(0 2V 100 0 20)
V11 21 1 SIN(0 20V 2 0 3)
*Vk 20 21 PWL(0 -0.8V 0.1 0V 0.1) +r=0 td=0.2
vin 20 21 pulse (-1 0 0ns 500ns 500ns 0.2s 1s)

the power is better at start:


epz3.png (23.54 kB . 1807x990 - viewed 3998 times)

[LB7 (OP)]

I think I study the bad case. In the spice simulation I have one test with positive energy and another case with a negative energy. But it is the case where R=0.02 Ohms where the energy is created. I will test with others simulators.


tg54.png (2.56 kB . 291x298 - viewed 3898 times)

The energy is low but it is something.

[LB7 (OP)]

Ahah, I made a mistake... I took the damping factor in V1 in SPICE not the phase angle... but the result is there ! If R is small the energy is created, even the current in the coil passed from 5000 A at start to 4850 A at 0.2 s around, the energy in the sources increases of 45 kW.


fr5.png (10.35 kB . 773x331 - viewed 3928 times)


gll'4.png (6.51 kB . 615x308 - viewed 3911 times)

The sum of the 3 voltages in serial gives:


frpok5.png (25.7 kB . 1795x990 - viewed 3893 times)

If I remove the pulse voltage, it doesn't work. But it works without the 20V sinus. I have the sum of the voltage:


rfpr.png (22.01 kB . 1794x986 - viewed 3899 times)

the current is:


frpsok5.png (23.35 kB . 1800x991 - viewed 3899 times)

and the power is:


frpz.png (32.18 kB . 1807x982 - viewed 3894 times)

To resume, without the sinus damping it doesn't work, without L is doesn't work, without the 20V sinus it works but the power oscillate. The inductance lost few energy some 300 J but the power oscillate at start near 10 kW

I think it is well like my mechanical device. At start, there is an inductance with a current. Like the kinetics energy of rotation of my bolt around A1. At time=0s, I add a voltage source, the extra current has no inductance, because the curernt takes time to be stabilize insde the circuit due to the inductance. When the current is establishing inside the circuit, I use the support of the inductance. Even, it is not what I would like to build, it is another case with only one inductance.

[LB7 (OP)]

But without the damping it works. The principle is easy, I discharge an inductance into a circuit that I know the voltage source is not powerful enough to annoyed the current from the inductance. It is exactly like my mechanical device. The current (positive) comes from the inductance. The voltage source is like my spring I use in my mechancal device. The inductance and the current at start it the potential kinetics energy of the bolt around A1. After, I apply the spring: the voltage source or anything else like a capacitance. The energy can be recovered directly by the voltages sources. It must be possible to recover only a part of the energy created and give it to the inducatance, like that the device could be continuous. So maybe work in two periods, the first to recover the extra energy and second to give a part of the energy recovered to the inductance, to increase its current so its potential energy. Don't forget there are 125kW inside the resistor with a current of 25000 A, I didn't count it before but it is not a so small energy. But it seems it works only with the 3 voltages in serial, so it must realized the equivalent of the angles in mechanics.

If I don't use the damping function I have something like that for the power:


frpl.png (25.56 kB . 1786x992 - viewed 3860 times)

It depends how the device is used.

The damping function must act like the angles for the spring in the mechanical device. The exponential function changes the slope of the signal.

[LB7 (OP)]

The circuit:


fe84f.png (41.33 kB . 548x485 - viewed 3830 times)

The inductance must be charged before start ! it is like the rotation of the bolt around A1.

[guest46746]

I have only read your 1st post and scanned the next two. Are you saying that an aberration in the current provides an added tension via dimensional elevation above the source current?  Very Ein Sof! lol

Do the six elevated rings create a delayed current that produces a push/pull effect?

Why 12 rings?  lol

[guest46746]

Pardon for the inquiry, but have you played with inductance, via coil spacing on the elevated ring? lol

[LB7 (OP)]

First, look at my mechanical device and maybe you can understand what I want to do with the electrical device. The mechanical device has a sum of energy not at 0, it is easy to verify, a bolt in rotation around A1 but not in rotation around itself (longitudinal axis) at start, a spring, that's all. I take a bolt because I used the helix to show how I apply the force. I tried to imagine a transformer (to build) to have the same effect, it is better to test with a simulator like ngspice than a 3d simulator. After, I thought it could be done with only the electric voltage sources and I tried a lot of values. I didn't say it is the optimum. Hey, at least try, I gave the circuit and the file.

[guest46746]

Hey, at least try, I gave the circuit and the file.

 kudos! lol

[LB7 (OP)]

Oh, I just look at all your messages ! "lol" at each phrase..., I see a happy troll...you never build a raisonnement, nor help

[guest46746]

you never build phrases, nor help

true intellectuals don't want or appreciate ...help! kudos none the less! lol

[LB7 (OP)]

About the slopes, I drew an example with sin(x) and 0.5*sin(x):


dz95z.png (41.3 kB . 575x466 - viewed 3724 times)

The exponential function (damping in Ngspice) decreases more and more the factor of the sin(x) and the slope changes, it is what I need in my mechanical device apply here in the electric.

The energy in the coil, is a real energy, I need to give before start like the mechanical device.

If the device is not continuous it is because the damping function reduces the input. But I will have the all circuit.

[LB7 (OP)]

With a RL circuit:


fr95s.png (23.63 kB . 920x647 - viewed 3731 times)



Instead of a transformer or voltages series I can use a LC circuit in serie, in resonance. The circulation of the current is like the rotation of the bolt around A1. I add a source of voltage between the inductance and the capacitor. The source of voltage is the spring. The angles is realized by the slope of the voltage. I mean, the sourve of voltage must change the frequency of the signal. Maybe I can add a positive sinus voltage from 0 to 90° and add a negative sinus from 90° to 180°, add positive voltage from 181° to 270° and add a negative from 270° to 360°.

Maybe with a capacitor too :


sz95s.png (32.27 kB . 703x543 - viewed 3734 times)

[LB7 (OP)]

So, if in my mechanical device I need to move the origin of the force, maybe here, it is the same, so I need to move an inductance. I didn't draw the core. I think like I drew the angles of the inductances between them, I  respect the mechanical device:


fd98.png (79.48 kB . 948x743 - viewed 3718 times)

I think it is better to vibrate with a piezoelectric motor. Like the current is alternative,  I think it must have the good direction for the power.

Another method, use the electrons like a mechanical weight. The weight in rotation around A1 is the electrons. I need to use electric field to accelerate and slow down the electrons. I use an alternative current and a piezoelectric motor in vibration. I don't know if the efficiency could be good with that method. Maybe it is what I reach to do but with a capacitor or an inductance. Maybe the only way is to use the electrons like the weight in the mechanical device.

Or maybe like that, the source of current I2 is dephased of 30° relatively to the current inside L1:


fr95.png (64.01 kB . 1169x542 - viewed 3589 times)

or easier:


fr9e8.png (48.91 kB . 1058x439 - viewed 3506 times)

or maybe I need to have an helix for the small radius of the inductance L1 because I don't know how to realize the difference of length I have in the mechanical device:


t8rr8.png (62.56 kB . 1071x587 - viewed 3579 times)

Or L2 the helix:


t88rr.png (59.08 kB . 1065x599 - viewed 3592 times)

Maybe I need to think with a continuous voltage V1 and like that I see how I recover the energy and how I can give it easily. Move a coil inside a continuous magnetic field is an alternative voltage. So if V1 is continuous then V2 is alternative:


f985.png (39.05 kB . 985x389 - viewed 3566 times)

And if I want V1 alternative, V2 is modulation of amplitude AM signal.

[LB7 (OP)]

I think the analogy with the mechanical device needs to resolve the question of the displacement of the motor, the velocity of that speed.

1/ Analogy with angular velocity <=> electric current:
One turn is realized by the current, very fast, the electrons run at 0.5*c, so it will be very difficult to move so quickly a big distance. Of the small diameter of L1 must be very small, and maybe a piezo can do the job, or another method.

2/ Analogy with angular velocity <=> electric voltage
Here, I can move at the velocity I want. If I have an alternative voltage for the capacitors, I can move in relation with the frequency of the voltage source.

3/ I take the analogy angular velocity <=> inductance
For example, a solenoide is L=k*N² with k a constance and N the number of turns of the solenoide. The energy is 1/2*L*I²=kN²I². If I consider the current like contant inside the solenoide (I build a device to realized that), the energy is E=k'N². I can change easily the solenoide with a saturated core: I add a DC (constant) current inside the core, to saturated it less or more like I want. With a solenoide the energy is big, I work with an alternative signal, I have low losses if the resistance of the inductance is low. There is no mechanical movement except the source of force. I will think about that.

[LB7 (OP)]

If I use the capacitor the geometry implies a very low capacitance of maybe with semiconductor it is efficient, like it is possible to work at high frequency and like the signal are sinusoidal the losses are reduced.

If I use the inductor, maybe I can use the flux, I'm not sure about what I need to take in reference, I mean, in the mechanical device I must try to accelerate the bolt from something in rotation around A1. But here, it is the current itself ? or the effect of the current like the magnetic flux ?

But I need to move faster than the equivalent of the mechanical inertia. If it is the current, it is difficult but possible in a lab and for the matter, it is not a problem. If it is the voltage, the voltage can be fixed, so it seems possible but the capacitor must move something, and I think it is the voltage not the geometry and it is better.

[RD]

Maybe with a capacitor too

Work is done to charge a capacitor : You don't get back all the electrical-energy you put in.

[LB7 (OP)]

Yes, you're right and it is a big problem ! But when I discharge a capacitor, the sum of energy is constant, half part is in heating the other part in electric. Here I recover part of electric, a part of heating but the sum of energy is not constant (if my mechanical device works). But it could be the explanation why nobody reached to create the energy with a device because even with a mechanical device the potential kinetic energy is so big, the losses are bigger than the extra energy. With the capacitor, half is lost in heating. With an inductor, you must be faster than the current or it is not the speed of the current but the value of the current, in that case it could be easier.

But here, maybe it is possible to recover the energy from the capacitor whithout discharge it. Maybe a LC resonant circuit and L is the primary of a transformer, in the secondary I recover the energy. It is just an idea like that.

Maybe for the inductance, I can have L1 in aluminium and L3 in copper like that the current is faster.

[RD]

With an inductor ...

Analogous to a capacitor : inevitably energy is lost when energizing an inductor,
e.g. see ... https://en.wikipedia.org/wiki/Transformer#Energy_losses

... aluminium ... copper ...

All room-temperature metals/materials have electrical-resistance : than means inevitably some electrical-energy is lost via heat.
You cannot get-out all the electrical-energy you put-in.

... if my mechanical device works...

All perpetual-motion, (and "over unity"), machines are impossible ... https://en.wikipedia.org/wiki/Perpetual_motion

[LB7 (OP)]

I'm agree with you. And the transformer lost at least 1% or more of energy in heating. In my mechanical device I supposed not friction just to simplify the calculations. I know, there is always a heating because there is friction but heating is an energy and even the device gives a heating the sum of energy included heating is constant. Maybe my mechanical device is wrong but I don't find my mistake. It is here : https://www.thenakedscientists.com/forum/index.php?topic=74864.0 look at the last drawings. If I'm wrong in my mechanical device it is not useful to try to find an electric device. I simulated with Comsol without friction (it is easy with a software) I have the same result I calculated (for a small angle of rotation from start).

Edit: for the solution with the capacitor, I don't need to dischaged all the capacitor, just enough to recover the initial condition. So the energy lost is 50% of the energy in addition and it is exploitable.