[LB7 (OP)]

I take in reference the mechanical device : https://www.thenakedscientists.com/forum/index.php?topic=74864.60

I explained and proved the sum of energy is not constant. The main problem is the energy recovered compared to the potential kinetics energy inside the device, I can create something like 1% of the potential energy, but the potential energy is a mass in rotation and the problem comes from the losses: friction from air and rolling-element bearing. Even the device create the energy the goal is to recover a mechanical energy or at least a heating without give any energy (the potential energy is gave at start and it can be recovered at final).

I drew the small circular rings with an angle of 30° relatively to the horizontal but the angle like I show in the thread about the mechanical device can be any others angles. The angle change the efficiency.

So, I do the electric analog device. It is the first time I do that, so maybe my model is not perfect. But I think it is easy to verify with a simulator or in reality. I took the impedance analogy but it is possible to take the admittance.

The device is like that:


d9s6q.png (95.9 kB . 1207x566 - viewed 7626 times)


df6s5.png (54.39 kB . 716x405 - viewed 7601 times)

I drew the arrow to show the direction of the current in a half period, the direction is opposite in the second period.

I imagined the device alimented with sinus tension but it is possible to use any type of signal : square, triangular, or even pulse, etc. Maybe it is better to increase the current when the diameter of the ring is bigger.

To give the force to the small ring, I can use a voltage transformer that I add in each small wire (serial). Like that it is the equivalent of the force I give in mechanical device.

I have very small losses from the electric copper or I can use supraconductor materials. And nothing needs to turn like the mechanical device. The frequency could be very high, so the power density (or power mass) could be very high.


se99d.png (38.73 kB . 940x572 - viewed 7603 times)


ze9s5.png (57.91 kB . 468x404 - viewed 7542 times)


The device is like :


s9e6d.png (116.03 kB . 1184x572 - viewed 7514 times)

I use an electronic device to recover the extra energy, I don't want the sources give an energy to the electronic device directly, it is only the energy created. I drew only one ring to recover the flux but it is possible to have several blue rings.

I didn't drew irons nor ferrite cores but it is possible to add them to increase the inductance of the device, in the center:


sz95s.png (118.96 kB . 1168x572 - viewed 7483 times)

And maybe between the copper wires too, to increase the inductance.

If I don't recover the extra energy in my mechanical device and if there is no friction, the angular velocity of the device increases more and more until infinite (unless the device is a the basic particle it will break itself before). Here, it is the same, the flux will increase more and more and the copper will fuse. So, it is necessary to recover and use the extra energy. The device is unstable.

I drew circles for the copper in the side view, but it is possible to have any shape: rectangles, etc. I drew only one small ring but it is possible to have more.

[LB7 (OP)]

The image shows the device that gives the equivalent forces in mechanical device:


d98s6.png (200.62 kB . 1025x774 - viewed 7360 times)


dee23.png (36.32 kB . 977x432 - viewed 6815 times)

The violet ring (A) is alimented in tension with a generator alternative for example. I drew arrow to show the direction of the current in a half period.

I updated the device:


s96d.png (182.65 kB . 1505x761 - viewed 6834 times)

I drew a wire of copper (or any conductor) but it is possible to have 100 rounds for one ring.

It is possible to use more than one violet ring to increase the efficiency. For example, one like I drew, the others separate by 90° on the side view.

[LB7 (OP)]

Maybe I can do that:


s958d.png (192.01 kB . 1519x804 - viewed 6759 times)

The secondary generator is ON at start, and I change the phase angle and the voltage during the time. I think the secondary generator recovers an energy at start, and need to recovers more and more energy with time. I recover energy too from the electronic device, but just enough to keep constant the electric field from the generator 1.

I will test this device.

[RD]

The image shows the device that gives the equivalent forces in mechanical device:


d98s6.png (200.62 kB . 1025x774 - viewed 7360 times)

That looks like a transformer, which can increase a voltage but cannot increase (create) energy.
It is impossible to create energy. 

I drew a wire of copper (or any conductor) but it is possible to have 100 rounds for one ring.

Copper has electrical resistance, so you will lose energy via heat.
Also heating via magnetostriction ... https://en.wikipedia.org/wiki/Magnetostriction
Also loss of energy from your system via RF.

[LB7 (OP)]

That looks like a transformer, which can increase a voltage but cannot increase (create) energy.
It is impossible to create energy. 

Yes, it is a transformer. And the goal is to increase the current like the mechanical device does and increases the angular velocity. The design is more:


ed6d3.png (106.24 kB . 460x456 - viewed 6809 times)

Like that I have the analogy. So, the secondary generator recovers an energy I think.

It is not a device like that, it is based on my mechanical device. Maybe the analogy is not perfect but my mechanical device creates the energy. It is easy to verify with maths the mechanical device but calculate the electromagnetic device is not easy. In the contrary, it is easy to build in reality the electromagnetic device but it is difficult to built a real mechanical device to test, maybe in a lab. I done the calculations for the mechanical device and now I will build the electromagnetic device.

Copper has electrical resistance, so you will lose energy via heat. Also heating via magnetostriction ... https://en.wikipedia.org/wiki/Magnetostriction
Also loss of energy from your system via RF. 

Sure, it is very complex to calculate like that. There are losses and I don't know the pourcentage of the energy I can create. I will try to simulate it today but if someone want to help, he's welcome. I hope the electromagnetic device is more efficient than the mechanical because it is lightweigth and it is possible to work at a big frequency.

I think the electronic device is not necessary in the test because the power will increases more and more and the copper will fuse even the power of the generators is not enough to fuse the copper. The core (iron or ferrite) will create high frequencies so more losses, so I think it is better to test without any core. And the signal is better in sinus, the measures will be better at low frequency like maybe 50 Hz, and the RF losses difficult to calculate or estimate will e lower. Maybe the phase angle of the second generator is necessary, I'm not sure.

The following image shows the 3d model:


sx6.png (41.56 kB . 792x669 - viewed 6782 times)

I think I need to take the energy for the secondary coil from the primary coil:


fr32e3.png (188.58 kB . 1325x778 - viewed 6758 times)

The upload disk is full so I post here my device:

https://postimg.cc/qz0QLn3k

Like that the device is well the mechanical device at least for the moment of inertia. Without C2 the device has an inductance L1+L2 and it is the inertia of the bolt around the fixed axis more the inertia of the bolt around itself. If I active C2, then the inertia is L1 and I need no energy for C2, and it is like only the rotation of the bolt around the fixed axis, the rotation of the bolt around itself is L2 and L2=0 because the currents are in the contrary directions.

Like that it is better:

https://postimg.cc/G4mWX6yh

And another :

https://postimg.cc/dDfSjD93

The design for one turn of each coil:

https://postimg.cc/VrDF7s0K

With 2 real wires:

https://postimg.cc/gallery/16m5tl5hg/

[LB7 (OP)]

I think the solution is easier like that:

https://postimg.cc/2VdGKfLR

G1 gives less energy with C1 like I drew compared to full turns of the red copper. G2 gives an energy to C2. C2 gives an energy to C1.

I didn't draw the device to recover the energy (a coil below C1).

Like that the initial condition of the mechanics is realized:


f9d6.png (78.61 kB . 1216x869 - viewed 6537 times)

The current of C1 turns clockwise but the red triangle gives a counterclockwise rotation of the red current. So it is like the rotation of the mass around A1 and  the bolt that don't turning around itself is C2: cancel the flux by himself.

Note: I took 2 examples for the angles, maybe it is not the best angles. And maybe C2 must be like C1 but the current reversed:


d6s5.png (90.34 kB . 1253x855 - viewed 6421 times)


I drew only one wire for the "triangles" red and violet but it is possible to have severals triangles distant from each others.

I don't know if the red copper in the plance (x,y) must be under the violet or the red triangle. Maybe the work from the violet triangle is better.

The generator G1 gives the power in parallel to each red round (C1) but it is possible to build a spiral and set the wire in serial, again I don't know what is the best maybe it depend of the position of the violet triangle above C1.

I think I need to replace G2 by a resistor, like that the energy is created not destroyed:


s8d5.png (90.27 kB . 1284x863 - viewed 6440 times)

[LB7 (OP)]

I corrected the direction of the current. One device create the energy the other destroy it.


d9ss5.png (92.47 kB . 1226x857 - viewed 6458 times)


d6s33f.png (91.9 kB . 1253x862 - viewed 6435 times)

I drew one wire (red line or violet line) but it is a coil, it is several wires inside each line, like that the value are greater.

The best could be to add a magnetic core composed of amorphous metal.


d5t8.png (63.37 kB . 804x452 - viewed 6383 times)

And I think I need to take the power from C1 only during the first half period because the coil C1 increases its power like the bolt could increases its velocity around A1 (mechanical device).


d9e6s3.png (115.54 kB . 1288x870 - viewed 6365 times)

I think I need to cut the circuit of C2 (only the half period must give power to R), because imagine I use a push spring in the mechanical device and I accelerate and decelerate the bolt around A1 like the sinusoide does inside C1, in that case at the end of the acceleration+deceleration the length of the spring is at the same length than at start, and I lost and I recover the same energy from the torque around A1. So, the device can create and destroy the energy it depend when I take the power from C2.

[LB7 (OP)]

I have some problems due to the inductance stored inside C2. In the mechanics device I didn't study a continuous device, just a device to prove the energy is not conserved. But here, if I want to measure, I need a continuous power because the energy is small and if it is in a transcient study the results will be not enough appropriate because the transcient analyse will create the harmonics and the rmultimeters or oscolloscope have a limited widthband.

So, I need a continuous measurement without transcient if possible. For that, I need to change the type of attraction in my mechanical device or change the spring... but here the inductance is there at final of one half period. I need to change the force on the support around A1, so here I need to change the direction of the flux because reverse the rotation in mechanical is change the voltage. For that, I need to change the direction of the current. And for that, I need a motor that can rotate C2, parrallel to the axis X. It is not very fast but maybe I can rotate the motor with a continuous angular velocity, the efficiency is lower but I can rotates very fast. I will find a static solution.




d6s5z.png (94.17 kB . 1315x868 - viewed 6290 times)


And I found:


d8f5e.png (59.6 kB . 1051x614 - viewed 6255 times)

With interuptors:


gf8r5.png (71.82 kB . 1062x666 - viewed 6258 times)

And I can have a triangular shape:


s8d5z.png (57.85 kB . 945x479 - viewed 6279 times)

I think I need to move the flux outside C1, if I use C2 without a core I think the sum of the flux is 0 from it. So I use a core to move outside C1 the flux. Like that, inside C1 I have what I want from C2:


d9r6f.png (94.54 kB . 1359x873 - viewed 6302 times)

And I can recover the energy from C2 from another coil in the core 1 (I don't need 2 cores I think):


d8fr9.png (94.65 kB . 1314x877 - viewed 6297 times)

The core 1 don't pass around C1. Like that I can recover the energy from C2 without C1. I can open C1 and recover then energy with another coil like a standard transformer.

Like that:


gf89d.png (118.86 kB . 1347x873 - viewed 6284 times)

No, the best is to open C2 and recover the energy from C1. More difficult to do in practice, need electronic. And I need cores ! Like I drew at start, it is not good for harmonics but I need to add a flux from C2 to C1 !


s9s6.png (188.53 kB . 1129x621 - viewed 6240 times)

Open C1 and give the energy to the core is like brake the bolt on the disk. I think the cores are needed to give the path of the flux like I want. But the device need an electronic circuit.


g9d5.png (112.4 kB . 1255x862 - viewed 6212 times)

[LB7 (OP)]

With a core like that:


fg95g.png (122.68 kB . 1223x856 - viewed 6132 times)

The flux is in the direction I want from C2. I think I can't place the core in the center of C1.

[LB7 (OP)]

In fact there are a lot of solution to convert the principle of the mechanics motor to the electrics transformer. Maybe this one is better:


g6fd.png (202.22 kB . 1496x708 - viewed 6175 times)

Analogy mechanics <=> electric

1/ The generator 1 is the motor that allows to have the bolt in rotation around A1. The advantage of the electric is the inertia is reactive power, it cost near nothing so the losses must be low. Another advantage is the possibility to use the alternative power, it is like I rotate in one direction the disk around A1 and just after in another direction, in the mechanical device if I do that, even the energy is created, I can't see it because the losses will be very high.

2/ Inertia of the angular velocity of the disk around A1 without the bolt around itself <=> inductance of C1, note it is easier here to have a bigger inertia of the bolt around itself compared around A1. The bigger radius of C1 is the rotation around A1 and the smaller radius of C1 is the rotation of the bolt around itself. Like I drew it is not necessary to have a plate coil. It is better like that. Look at the direction of the current in C1 (I drew only the current in C1) because a part of the current turns clockwise the other part, in the small radius, rotates counterclockwise.

3/ The inertia of the bolt around itself is C3, maybe it is not perfect like I drew, I'm not sure.

4/ The generator 2 is the spring I used in the mechanical device. The spring gives a force to the bolt to rotate around itself not to rotate around A1 (the othen end of the spring is discuste in 5/), here it is the same I give only the tension to C3.

5/ In the mechanics, the spring has its reference to the rotation around A1, here I do the same, the generator 2 is in serial with C2 and C2 has the same shape than C1 I mean when I rotate the bolt around A1 I don't rotate it around itself, so here it is the same.

6/ To recover the energy in the mechanics device I recover a torque and a rotation, here it is the current and the voltage from C5.

7/ The angle I have for the bolt is represented by the angle relatively to the horizontal I have for the small radius of C1.

I updated the device:


e9d6.png (203.69 kB . 1536x865 - viewed 6165 times)

C1 with the iron2:


ed6s6q.png (296.28 kB . 686x478 - viewed 6536 times)

[LB7 (OP)]

With that design, the area is better:


r8f5d.png (220.98 kB . 1538x884 - viewed 6112 times)

At least, it allow to understand where I want to go. I don't need C3. I can directly put the flux like I drew. The only problem is to minimize the flux from C1 to the core2 and I can take a lower reluctance.

μr for the iron 1 = 2000
μr for the iron 2 = 200

If the flux from C1 increases the inducance of C1 because I added the core2, this will ways the bolt rotates at start in the mechanical device. To create the energy I need to have the angular velocity of the bolt lower than the angular velocity around A1. Even a small difference is enough but for increase the power-mass and the volume-mass it is better to have a big difference.

The coil C1 alone:


dr8f5.png (36.58 kB . 1052x462 - viewed 6085 times)

C4 and C5 are standard (circle shape)

The cores alone:


d98e5d.png (31.09 kB . 1234x597 - viewed 6089 times)

To test with a core more easy to build (less efficiency):


d8s5d.png (63.63 kB . 1115x539 - viewed 6046 times)

II drew 2 circles for C1 in the top view, the bigger circle with a radius R1 is well the circle but the smaller with the radius R2 is an ellipse. So if I want a circle for the smaller radius R2, the image is more:


f8r5.png (36.02 kB . 1001x417 - viewed 6024 times)

[LB7 (OP)]

Maybe I can have that geometry:


d8c5.png (140.43 kB . 1056x724 - viewed 5995 times)

If I don't need C2 the cost is lower.

Maybe I need to have another coil in the core2 to put voltage (force in mechanical device) in the part of C1:


d8e5d.png (139.29 kB . 1070x755 - viewed 6000 times)

Or I need to have a different slope for C2 :


d85f.png (79.65 kB . 1380x731 - viewed 5994 times)


I need to move in translation in mechanics, so I need to move here, and the analogy seems to be the charge:

dz9d58.png (135.24 kB . 1386x845 - viewed 5991 times)

Or like that:


d9s5d.png (148.66 kB . 1355x863 - viewed 5930 times)

The device like that is easier:


ed95d.png (78.5 kB . 1181x695 - viewed 5908 times)

The diplacement is the charge, the spring is G2 so I think I need to place the capacitor in serie with C6:


des9d6.png (58.23 kB . 1215x610 - viewed 5901 times)

But maybe there is a phase angle between G1 and G2.


f5d6.png (118.2 kB . 1192x728 - viewed 5892 times)

[LB7 (OP)]

Maybe the capacitor is between C1 and C6 because the displacement in the mechanical device is between the rotation around A1 and the bolt around itself:


d98fe.png (103.32 kB . 1051x671 - viewed 5828 times)

In the mechanical device there is no movement relatively to the axis A1, only relatively to the bolt, so maybe the capacitor is inside only in relation with the coil 6, one end attached to C6, the other end too:


d8es65.png (102.69 kB . 1041x695 - viewed 5814 times)


I'm not sure I can replace the capacitor by a phase angle transformer, I use 2 devices. Instead to use a capacitor, I can use an inductor. Because the reason to exist of the capacitor is to create a phase angle I think, and it is possible to do the same with 2 identical devices (in opposition), I'm not sure I can create an opposition with the coil C1 because it is used. So the device could be:


f5gt8.png (65.1 kB . 809x758 - viewed 5825 times)

The schema is:


dc6d.png (36.24 kB . 415x426 - viewed 5737 times)


It is easier to build inductor than capacitor especially in high power.

I'm not sure of the angle of the green coil... in the mechanical device I have a different angle for between the perpendicular of A1 and the force of the spring and I have also a difference between the force of the spring and the axis of the bolt (axis around itself). And I didn't find the best couple of angles, so I think there are several solutions. Liek that device, electric device, I think there are several solutions, it is possible to work with capacitances not coil, it is better for small embedded devices, like maybe a watch.

[LB7 (OP)]

I think I can place the coils in the same plane. The capacitors must realized the function "angle' I used in the mechanics:


f5fd6.png (111.2 kB . 1118x598 - viewed 5769 times)

Maybe I need to use coil of Frager:


d9qz5d.png (128.51 kB . 1204x613 - viewed 5786 times)


d9e6d.png (134.98 kB . 1223x698 - viewed 5752 times)

Maybe the flux must be dephased from the green relatively to the red with spires of Frager:


DE95F.png (159.59 kB . 1238x799 - viewed 5734 times)


dzz9.png (161.17 kB . 1242x815 - viewed 5712 times)

Maybe I don't need the capacitors:


de95de.png (191.44 kB . 1318x826 - viewed 5709 times)

With notations:

 [ Invalid Attachment ]


dzz9.png (161.17 kB . 1242x815 - viewed 5712 times)

I corrected some details:


de95de.png (191.44 kB . 1318x826 - viewed 5709 times)

[LB7 (OP)]

I corrected the sign of current, but I don't think I need to have a phase angle between G1 and G2 and it is not necessary to have a variation of the voltage of G1 and G2. I have only a doubt about the displacement of the spring. I think I need to have like the mechanical device an helix, but what is the equivalent of the helix in electric ? I think the device need to change the inductance or the capacitor, but if it is a capacitor, it must change its capacity ! but it is more complicated to change the capacitor than the inductance. Or maybe with a piezoelectric device: 2 conductors separated by a isolant piezo material and change the thickness. Whay not. I don't know if these sort of capacitor exist or not. The value must change very few if the frequency is high, but if the power is high, the changement of the inductance or the capacitance must be higher. Electronic can easily with no mechanical movement change the frequency.

Or with only a modulation of the voltage on G1:


Or the frequency, because changes the kinetics energy is change the velocity in rotation, and it is change the inductance, and change easily the inductance is easy with the frequency:



With a variable frequency on G1 not G2:


I can use a varicap ! like that it is easy to change the capacitance. I need to find where to place it. Like the generator 2 and the greometry of the iron is the spring, and like the spring moves, I need to find where to place it:

 [ Invalid Attachment ]

With the voltage of control Uc:


t89h9d.png (187.8 kB . 1363x833 - viewed 5666 times)

Several solution to change the inductance:


de98fe.png (191.81 kB . 1342x821 - viewed 5670 times)

I need to modify, I forgot the return of the flux from C1...:


tg5g9d.png (217.7 kB . 1365x768 - viewed 5624 times)

[LB7 (OP)]

I think I drew the solution with the loop created by the parts of C1 with a radius R2 and R3. It is a loop exactly like a spire of Frager, and the flux must be dephased. I need just to control the current in the loop, it is easier than variable frequency or a limited varicap:


fr95f.png (228.08 kB . 1366x788 - viewed 5656 times)

The current can be limited by an electronic circuit or even by a variable transformer.

Maybe with a ballast:


f9e6d.png (231.03 kB . 1353x804 - viewed 5574 times)

Even a controlled electronics ballast can limit the current inside the frager spire. The Frager spire is the part of C1 with the surface A2.

I drew the true section of the irons:


cd9s.png (36.7 kB . 576x374 - viewed 5563 times)

[LB7 (OP)]

I simulate the device, it is not like the reality but it is a start. I can have a negative energy without count the energy inside the coils. I use one of the methods I explained before. It could be better to have the equations from the mechanical device and build the electrical analogy. I will try.

Here, I just need something to store the energy in the transcient, and repeat again and again the cycle. It is like I have only one shot of the mechanical device.


gf9f.png (6.5 kB . 823x534 - viewed 5519 times)

With another values and at start:


r8fs.png (7.07 kB . 649x444 - viewed 5526 times)

But maybe my circuit is not good:


de95ds.png (42.1 kB . 441x693 - viewed 5550 times)

I tried a lot of values to optimize the device but it miss something just to integrate the work from start. Even V2=0V the sum of energy is negative at start. I hope the couplage of the inductance is correct.

With different value I can have that:

fr98.png (17.46 kB . 1723x849 - viewed 5513 times)

V1 0 20 SIN(0 20V 100 0 0)
V11 20 1 SIN(0 200V 210 0 5)
V2 0 11 SIN(0 0V 100 0 0)

V5 12 8 SIN(0 0.00001V 100)

R1 1 2 0.0001
R2 3 4 0.0001
R3 3 5 0.0001
R4 7 8 0.0001

R7 7 11 0.0001

L1 3 2 10mH
L2 0 4 5mH
L3 0 5 100mH
L4 0 12 2mH

I added a voltage square in serial with V1 and the result are bigger with a non 0 at final:


cc95c.png (20.28 kB . 1778x965 - viewed 5527 times)

zoom in:

fd8s9.png (3.35 kB . 451x395 - viewed 5507 times)

And with a small voltage square, I don't need to give an energy:


eds98.png (19.85 kB . 1800x989 - viewed 5507 times)

With:

Code: [Select]

V1 0 20 SIN(0 20V 100 0 0)
V11 21 1 SIN(0 20V 1 0 5)
V2 0 11 SIN(0 0V 100 0 0)
Vk 20 21 PWL(0 -1 0.1 0 0.1) +r=0 td=0.01
V5 12 8 SIN(0 0.00001V 100)

R1 1 2 0.0001
R2 3 4 0.0001
R3 3 5 0.0001
R4 7 8 0.0001

R7 7 11 0.0001

L1 3 2 10mH
L2 0 4 5mH
L3 0 5 100mH
L4 0 12 2mH

K12 L1 L2 -1
K13 L1 L3 -0.8
K14 L1 L4 0.4
K42 L4 L2 0.6
K43 L4 L3 0
K23 L2 L3 0.8
The power is:


ser6.png (24.11 kB . 1786x963 - viewed 5497 times)

In fact, if I set all K=0 and decrease a lot L3 and L4, I test two cases Rx=0.0001 ohm and Rx=0.01, in the first case the sum of energy is negative a lot (kUnits) and in the second case the energy is positive:

Rx=0.0001:

Code: [Select]

*** transformer ***


*Fit-Range: 0-30V


* transfomer


V1 0 20 SIN(0 20V 100 0 0)
V11 21 1 SIN(0 20V 1 0 5)
V2 0 11 SIN(0 0V 100 0 0)
Vk 20 21 PWL(0 -1 0.1 0 0.1) +r=0 td=0.01
*V5 12 8 SIN(0 0.00001V 100)

R1 1 2 0.0001
R2 3 4 0.0001

R7 7 11 0.0001

L1 3 2 10mH
L2 0 4 5mH

K12 L1 L2 0


.tran 10us 2.5s
.control
save all @r1.r2.r7[p]
save all @v1.v11.v2.v3.vk[p]
save all @v1.v11.v2.v3.vk[i]
run


plot avg(v(1)*i(v1))

.endc

.end


de843.png (2.52 kB . 295x297 - viewed 5518 times)

Rx=0.01:


sd654.png (2.65 kB . 290x301 - viewed 5467 times)

[LB7 (OP)]

Code: [Select]

*** transformer ***


*Fit-Range: 0-30V


* transfomer


V1 0 20 SIN(0 20V 100 0 0)
V11 21 1 SIN(0 20V 1 0 5)
V2 0 11 SIN(0 0V 100 0 0)
Vk 20 21 PWL(0 -1 0.1 0 0.1) +r=0 td=0.01
*V5 12 8 SIN(0 0.00001V 100)

R1 1 2 0.01
R2 3 4 0.01
R7 7 11 0.01

L1 3 2 10mH
L2 0 4 5mH

K12 L1 L2 0


.tran 10us 2.5s
.control
save all @r1.r2.r7[p]
save all @v1.v11.v2.v3.vk[p]
save all @v1.v11.v2.v3.vk[i]
run


plot avg(v(1)*i(v1))

.endc

.end


ds3s.png (2.49 kB . 287x293 - viewed 5372 times)

[LB7 (OP)]

So, nothing to do with my transformer but it is interesting. I mean, there no core and only one coil. But I think I replaced the severals coils by severals voltage sources and it is better to test with simulator. I think it is th principle I have in the mechanical device but apply to the electric model with sources of voltage.

So the device is only 3 sources of voltage in serial:

20 Volts sinus at 100 Hz : V1 0 20 SIN(0 20V 100 0 0)
20 Volts sinus at 1 Hz dephased of 5° : V11 21 1 SIN(0 20V 1 0 5)
Square signal -1V/0V, at 5Hz :  Vk 20 21 PWL(0 -1 0.1 0 0.1) +r=0 td=0.2

In serial with a resistance of 0.02 Ω or 0.0002 Ω

In serial with an inductance of 15 mH

Code: [Select]

*** transformer ***


V1 0 20 SIN(0 20V 100 0 0)
V11 21 1 SIN(0 20V 1 0 5)
Vk 20 21 PWL(0 -1 0.1 0 0.1) +r=0 td=0.2

* R1 = 0.02 => power destroyed
* R1 = 0.0002 => power created
R1 1 2 0.0002
L1 0 2 15mH

.tran 1us 200s
.control
save all @r1[p]
save all @v1.v11.vk[p]
save all @v1.v11.vk[i]
run

plot avg(v(1)*i(v1))
plot i(v1)
.end


r=0.0002

d87s.png (2.48 kB . 288x283 - viewed 5467 times)

r=0.02

s9d5.png (2.57 kB . 291x296 - viewed 5344 times)

If I divide by 10 the step, I have always the same. I can understand the energy from the generators is positive because the coi have an energy inside. But the contrary... It is based on Spice3, maybe there is a bug ?

With r=0.0002 the current is :


defz5.png (2.69 kB . 282x292 - viewed 5460 times)


de97.png (26.62 kB . 1785x976 - viewed 5458 times)

It is a lot, the real generator could have a big problem to generate it but why spice is wrong ?

[LB7 (OP)]

And at final the energy return at 0 but slowly:


y9r8.png (18.23 kB . 1802x977 - viewed 5469 times)



r98t.png (14.91 kB . 1797x983 - viewed 5432 times)


gt6e.png (15.71 kB . 1802x969 - viewed 5423 times)