STEORN DEMO LIVE & STREAM in Dublin, December 15th, 10 AM

[exnihiloest]

I even posted the news about the jury on this thread many days ago, where it was disclosed that the jury tested nothing. Steorn failed to show up with a working model within the contract time...

Well I missed that, sorry.
The situation is worst than I imagined. Steorn announced their discovery in 2006, and they pretend they have not even been able to give matter to a physicists team in 2009?! Well I feel a new hoax story.

If you say they are neglecting the energy consumed in spin alignment, then why is there a trace at all at the input? If they are subtracting (I^2)*(RR')/(R+R') from the input, the trace should come out a flat 0.
Instead we see many peaks, with each peak falling to a higher value, which means some energy is being lost inspite of subtracting heat losses. (I hope you must have seen that huge spike in dE/dt also). Where is this energy getting lost ? My best guess is, in the spin alignment and hysteresis loss. So they are not neglecting it.

Steorn estimates the wasted power from R*I^2 but R*I^2 is more than the losses.
The wasted power is R*I"^2 where I"=I-I' = I-U/R', R' being the equivalent resistance of the process providing energy to align the electron spin.
It follows that at each step, a power R'*I'^2 is added in the trace and this explains why we see each peak falling to a higher value.

It is the same absurd method as measuring current/tension powering a filament light bulb, then substracting R*I^2 from U*I (R being the resistance of the lighted filament of the bulb) to get the useful energy, then observing that no useful energy is consumed and finally concluding that it is OU because when we remove the losses we get light for nothing!!!
Then followers, misunderstanding that useful energy has been wrongly accounted for losses, tempt to remove the losses of the light bulb, still expecting for light  .

(I^2)*(RR')/(R+R') is only a part of the equation of the circuit. There is still the classical term -L di/dt (and probably others of second order due to non-linearities of saturation). All phenomenon influencing the current/voltage are not known and some are antagonist, for example the flux change due to di/dt probably acts against the flux change due to dL/dt, leaving us with flat U/I.

Of course the trace will prove nothing. But if we can show that the trace makes sense, steorn gets sorta plus points and it becomes less probable that they are scamming people for money. They are trying to show something, although in a very bad manner.

I agree but this work is for Steorn, not ours.

[exnihiloest]

Hmm,
this could be easily tested with a well done calorimetric test.
...

I agree, Stefan, but such a test is not so easy as you think.
One would have to distinguish the heat of the coil resistance from the heat of its ferrite core (because re-orienting the magnetic domains has also losses and provides heat).
And the useful mechanical energy that the motor can actually provide is very weak. Not sure one could measure a significant torque. This means that almost all energy from the pulse generator, changing in whatsoever form, finally ends in heating.

2. I am still wondering, that no one of the replicators is using the idea
of extracting the inductive BackEMF as Ben (user K4ZEP) said in his video, where he showed the blinking LED.
As we also gain electrical energy by the delta muR factor during the switching...

I don't agree this viewpoint. In classical physics you gain electrical energy by increasing a self-inductance L for example by tightening coils carrying current because you change mechanical energy into electrical energy. But in Steorn's motor, it is the coil current that works to change L. It is the reason why I think a part of this energy is wrongly accounted for losses.
BackEMF is not a problem. Conventional electronics allows us to easily recover it. Nevertheless we must remember that back emf is never more than the energy we put in the coil and that when coil current is switched off, L is low, thus not much energy can be expected.

[Omega_0]

It follows that at each step, a power R'*I'^2 is added in the trace and this explains why we see each peak falling to a higher value.

It is the same absurd method as measuring current/tension powering a filament light bulb, then substracting R*I^2 from U*I (R being the resistance of the lighted filament of the bulb) to get the useful energy, then observing that no useful energy is consumed and finally concluding that it is OU because when we remove the losses we get light for nothing!!!

Trace falling to a higher point means a loss of energy at the input. If you subtract this loss also, the trace will fall back to original level, which means no energy was consumed and OU is infinite. This is not possible for a real life setup.

Your light bulb analogy is problematic because light is energy and its about 2-3% of total energy consumed, so even if you subtract all the heat loss, you will find 2-3% extra consumption , which is given off as light.

http://en.wikipedia.org/wiki/Incandescent_light_bulb#Efficiency_comparisons

There is little point in arguing about this thing.

[lumen]

I agree, Stefan, but such a test is not so easy as you think.
One would have to distinguish the heat of the coil resistance from the heat of its ferrite core (because re-orienting the magnetic domains has also losses and provides heat).
And the useful mechanical energy that the motor can actually provide is very weak. Not sure one could measure a significant torque. This means that almost all energy from the pulse generator, changing in whatsoever form, finally ends in heating.

I don't agree this viewpoint. In classical physics you gain electrical energy by increasing a self-inductance L for example by tightening coils carrying current because you change mechanical energy into electrical energy. But in Steorn's motor, it is the coil current that works to change L. It is the reason why I think a part of this energy is wrongly accounted for losses.
BackEMF is not a problem. Conventional electronics allows us to easily recover it. Nevertheless we must remember that back emf is never more than the energy we put in the coil and that when coil current is switched off, L is low, thus not much energy can be expected.

@exnihiloest

I must agree with Omega, all the input power has been accounted for and shown several times on the scope.

All the input power was shown to be converted to heat AND was shown to be the SAME with a magnet present or not.
This leaves the rotation of the rotor as additional energy that can not be accounted for.
Watch the video where they show the rotor with every other magnet missing but are still providing the pulse to the coil at all locations. The waveform is indistinguishable at either location.

[teslaalset]

I received my MetGlas cores last week and did some initial research I would like to share to discuss.
Some data first:
- Metglas cores, type MP2510P4AS
- Nr. of windings: 80
- Wire: 0.5 mm diameter

Scope shots:
- top curve: 20 V/div, representing voltage at the collector of the transistor (node nr. 3)
- middle curve 1 V/div, representing the voltage at the emitter (node nr. 2)
- bottom curve 5 V/div, representing the input voltage
- time base is 50 us/div, frequency of the complete cycle is approx  3.3 KHz

Circuit:
- emitter resistor 10ohm, measuring current through the coil
- max. current throught the coil (L1) set to approx. 100 mA, allowing for firm saturation of the torroid
- input signal is a TTL level signal from a microcontroller kit (Arduino)

Experiments:
- Coil measured without external magnet, Fig 1
- Coil measured with a ring magnet directly attached to the torroid, Fig 2
- Coil measured with a ring magnet attached to the torroid with optimum spacing. (maximum coil value), Fig 3

Findings:
Figure 1 represent the measured curves without a external magnet attached.
A delayed increase of the current through the coil has been found, as well as a peak voltage over the coil that shows relevant energy freed when current is shut down.

Figure 2 represents the measured curves while the (strong) magnet is directly mounted to the torroid.
A longer delay of increased current can be seen. This means higher coil value is occuring. Also a bigger peak can be found right after current has been switched off.

Figure 3 represent the measured curves while the magnet is held at an optimized distance from the torroid, such that coil value is optimized to its maximum value. Pay attention to the large amount of energy released when switching off the current in the coil.

Thinking over these results, I find it striking that coil value is maximized when there is a specific space between the magnet and the torroid.
I have some thought of this on my own, but any thoughts of these findings are welcome to discuss.

[edit 1]
One conclusion I already would like to share is that the external magnet is essential to obtain a maximum coil value.
The findings without a magnet show the boxed B-H curve as specified by MetGlas, so including the B_R value present after switching off the current.
When applying an external magnet the H_c point of the boxed B-H curve moves to near 0 A/m values, allowing for very low B_R values and very high permeability at near zero current.
This can be compared with the B-H curves that FineMet shows, only this case has the max. permeability of metglas (1.000.000)
It's all fitting together now.

[edit 2]
My coil shows 4 wires.
- 2 for normal winding (80 in this case), used for this experiment.
- 2 for a tangential winding, not used in this occasion. I will do experiments with that winding later on, to substitude the magnet