|
Hi Smudge,
Having thought about this overnight I can see where this is coming from. The problem is that our electrical power sources are all voltage ones, supplying voltage from a low impedance source. When used to drive an electromagnet, the steady state condition after the inductor is charged is determined by the resistance of the coil and that is all losses.
--Yes, this is true, but I am interested in reducing thermal losses within the time of the first time constant.
The magnetic field energy that we want to do any work is supplied during the charge-up period. It is clearly advantageous to keep the continuous losses as low as possible by minimising R and then you need smaller voltage to sustain the current. If you then need to minimise the losses during charge-up you want this to be fast, and one way of getting faster charge rate is to initially supply a much greater voltage than that needed to sustain the current, then gradually reduce that voltage during the charging period down to the sustain level. I think this is what you were getting at.
--Yes, quite similar. The upshot is that we can control thermal losses in the capacitor or inductor within the first time constant. I think by analogy, and this can get me in trouble sometimes, but in terms of a cap, you can reduce thermal losses to near zero within the TC by using a voltage variable source, and controlling the V so it is just above that of the cap it is charging, so current equals Vcharge-Vcap/R, and is kept low. The dual situation in an inductor is to control the through current while keeping the V low.
There is nothing in there that suggests the V should be high at any time during this process. The magnetic field will be the same at 1 TC in the controlled as in the uncontrolled situation. The issue is how fast the magnetic energy gets to that level. "fast" is a relative term, based on the resonance frequencies, or 'internal clock' of the particular usage. I've found that the perception of adiabatic systems being slow is a vague usage.
There is a bit of a design compromise between absolute efficiency and the speed of response of the system, but since 'slow' adiabatic circuits run at Mhz, and power circuits run at Hz (even an electric motor is 'just' a couple of Khz) there is no problem with using adiabatic inductive energizing-- and de-energizing-- of our power coils, stators, etc. If my analogy is correct, the magnetic field can be delivered to the inductor in one time constant or less, with very low thermal losses, by using a current ramped, and voltage suppressed circuit. This is not to say that the presence of an electric or mechanical load will not affect this process, but the absence of high voltage at the start does not automatically prevent us from using the field that results.
I grant that such circuits are 'different' from the usual inductor energizers, and our global system is not adapted to their use, but at least they should be used in our overunity machines, even if they are never adapted by the general technical public.
As an example of the need to set the internal clock of the system in order to limit losses, take a look at the attached paper. The authors show by theory and experiment that putting a sine wave through a CR circuit that has a TC much shorter than the period of the signal TC can reduces the thermal losses, as measured by an accounting of the energy contained in the capacitor. Pg. 2 has the circuit, and the thermal losses at different frequencies relative to the CR time constant are shown at top of pg. 3. The amount that dissipation can be reduced has been a matter of controversy, but this experiment demonstrates that the dissipation reduction is not limited.
orthofield
If we had current sources delivering power the problem would not arise as the high impedance then creates the shortest possible L/R charging time. The voltage would automatically start high then reduce down to the sustain level. But such a form of power is not compatible with our infrastructure, we would need switches where the off condition is a short circuit.
Smudge
[/quote]
|
Author
Topic: Ultra-efficiency through adiabatic design (Read 14409 times)
