Thread: Eric Dollard
View Single Post
Old 07-17-2012, 09:21 PM
GSM's Avatar
GSM GSM is offline
Silver Member
Join Date: Jan 2011
Posts: 688
Originally Posted by garrettm4 View Post

In general, in a reactive circuit, if there is no resistance, there is a phase angle of exactly 90 between the vectors of voltage and current, this represents Circulatory Energy and not necessarily Power as we usually understand it. To describe this specific situation, we use the term called "Reactive Power".

With instrumentation, this is visually seen as a distortion of one or both of the vectors of current and voltage. This, as a displacement of the node maxima of the affected quantity. Curiously, the zero crossings of the respective vector quantities remain in quadrature relation but the displaced node maxima can either return more energy than stored or alternately, store a portion of energy that won't ever be returned, where this energy goes is unknown. These actions cause the wave of the electric field (power) to be distorted and asymmetric, thus denoting the excess return or excess consumption taken. Here, energy is either converging into the circuit from an outside source or is diverging into an unknown sink, presumably not as thermal radiation.

Arguably, this action is seen as a negative resistance in the circuit because, reactive elements can only store energy and cannot impart energy to themselves. If they are merely storage elements, the energy source that filled the time-variant storage medium cannot be the storage medium itself, and cannot be the original source due to the energy increase happening on the falling quarter period, or return of stored energy. Furthermore, despite the quadrature relation of the zero-crossings, the node maxima are displaced causing the constituent waves to align asymmetrically, either with opposing signs (negative power factor) or same (positive power factor).

If a positive resistance was seen as a voltage drop and a divergence of energy from the circuit into the ambient, seen as a load. By the law of duality, a negative resistance would have to be seen as a voltage rise and a convergence of energy into the circuit, seen as a source. If we return more energy than was stored, a negative resistance would be seen as a negative power factor. Whereby the source becomes a sink and the load becomes a source for a select portion of time.

Garrett M
Oh Garrett - how pleased I was to read these words of yours.

This is something I studied 20 years ago in relation to loudspeaker crossover systems and audio power amplifier distortion.

Yes, exactly as you write, audio transients arising within music time, but coincidental with reactive loudspeaker system generated back-EMFs generated in loudspeaker time totally unrelated with signal input can cause the (amplifier output) 'source to become a sink'.
When the NFB loop phase is not coherent with input and output (as is so often the case with SS power amplifiers in order to guarantee amplifier stability, often with phase turnover wrt output commencing from 100Hz upwards), whereupon the output voltage can momentarily leads input prior to current correction, the crossover artifacts which 'expert' designers claim to be completely non-existent due to NFB, become clearly audible due to NFB phase delay.
Only amplifiers where output terminal sensed NFB error potential is controlled fully in phase with input at all audible frequencies do not generate crossover distortion, and there are precious few of them in existence.

The dynamically induced return of reactively stored and released energy is something which is simply ignored and not tested for on the audio workbench, and this explains why amplifiers supposedly having 0.001% distortion specs still do not sound good beside low or non feedback designs.

I should however like to comment upon your words -
"can either return more energy than stored or alternately, , where this energy goes is unknown."

The return of reactively stored energy is always displaced in time with respect to source/ input.
That return from the reactive circuits, relates to how many phase changing circuits coincidentally become simultaneously energised before there is reflection/ return event. The greater the number of phase changes prior to reflection, the greater the reactive return boost.

At LF in loudspeakers there can be air spring motion as well as L-C crossover reactances, potentially causing dynamic impedance dips as low as 2 ohms on 8ohm LS systems, and yet the true loading of these dips will not (cannot) be revealed by steady sine testing - this exactly as you suggest Garrett, with your complex dynamic current/ voltage phase shift waveform observations.

However "store a portion of energy that won't ever be returned" is not something I have observed to be correct, for that energy is always returned/ dissipated the moment that the energising source input is either switched off or disconnected, this in the form of a reactive settlement tail which can be recorded with respect to 't'-off at the input.

Surely it is only photonically/ electromagnetically radiated energy, or the electric/ magnetic field transduced portions doing work, which won't ever be returned ?

Cheers ........... Graham.
Reply With Quote