Pierre's 170W in 1600W out Looped Very impressive Build continued & moderated

[listener192]

Attached is a slightly improved waveform, more symmetrical which produces a fairly good sine when filtered.
This was achieved by fine  adjustments in code delays. I achieve about 150W maximum output with this waveform but for about 450W DC input, so pretty inefficient.
A vast amount of power is being dissipated as coil heat.  You can see from the voltage and current waveforms that the 60Hz wave front is reasonable well formed.

I checked the coil currents for those in registration with the rotor and those at 90 degrees to the rotor. Interestingly, I found both to be almost identical in amplitude, not what I was expecting, as the coils not in registration should be of lower inductance due to lack of the rotor flux coupling.

This implying that a varying flux is established in the stator that is fairly uniform and that the rotor does not offer an appreciably lower reluctance path, as far as this flux is concerned.In an AC  generator with a DC energized rotor, there is a point coupling of flux onto the stator teeth in registration , it is a single flux that does not compete with any other flux (ignoring lenz for the moment).

The  flux spreads out around the stator in both directions and is at right angles to all of the stator coils, only where the rotor flux couples with the stator teeth, are those coils subjected to a changing flux  that produces induction.

Is this flux totally uniform or is there something happening in the direction of rotation along the stator?

Well yes there is and this is normally referred to as armature reaction, see attached diagram.

Note how the flux crossing point from rotor to stator is skewed in the air gap, the lines of flux trail behind the turning rotor.

In an AC generator, the effect of armature reaction depends on the power factor i.e the phase relationship between the terminal voltage and armature current. Reactive power (lagging) is the magnetic field energy, so if the generator supplies a lagging load, this implies that it is supplying magnetic energy to the load. Since this power comes from excitation of the rotor, the net reactive power gets reduced in the generator. Hence, the armature reaction is demagnetizing.

Similarly, the armature reaction has magnetizing effect when the generator supplies a leading load (as leading load takes the leading VAR) and in return gives lagging VAR (magnetic energy) to the generator.

As in this case both armature flux and rotor flux lead, induced emf E by 90^o, it can be said, rotor flux and armature flux are in the same direction. Hence, the resultant flux is simply arithmetic sum of rotor flux and armature flux.

In case of a purely resistive load, the armature reaction is cross magnetizing only.

Now, how much of the above is reproduced  in the DZ?

L192

[listener192]

Simple answer to the coil current issue.

As I am running the coils in series albeit each coil is also in parallel with an opposite coil, of course the current is shared, so the coils not in registration do not contribute to output but have limited inductance and their resistance will consume power.
Compare this to a generator with rotating flux source and you find that it is similar, with several stator coils in series forming a distributed winding (to produce a better sine). So some of these coils will be out of registration and will only have limited inductance and losses due to resistance.

L192

[listener192]

On Saturday, I discovered I had a few shorted turns on my rotor coils. I removed all wire and modified the rotor so I could wind some additional turns.
I have rewound with 0.8mm wire but still need to add some more turns to get the 230V operation I desire.

I achieve about 165V RMS with what I have wound on so far, but decided to try a load first before winding more turns on.
Increasing the load now results in step increases in output power, where before there was a maximum load achieved then further increases would result in a decrease in output power.

No magic break through however, the shorted turns were certainly impeding performance.
The 0.8mm wire is a slightly heavier gauge than the wire I removed and I see that has helped reduce the IR loss a bit. That caused me to ponder on the gauge that Pierre used, which in the first video is shown as 20AWG which is close to 0.8mm however,  Pierre's wire is quiet stiff and supports the line outlets. On that basis, I think the gauge must be closer to 1mm or 18AWG, which will make a significant difference to the current versus volt drop on the rotor windings.

For 700ft wire length the 18 AWG wire resistance would be 4.799 ohms instead of 7.109 ohms for 20AWG.


L192

[listener192]

Here is a quick check on my current build performance.

INPUT              OUTPUT          RECOVERY
117.5W Av       29W Av          32W Av@25V

224W Av          59W Av          67W Av@35V

373.5W Av       91.7W Av        110W Av@45V


Note: @50V, the max output with the same load used for these tests, is about 138W

The boost recovery circuit is working reasonably.
The ratio of input to output is constant.

A check on flux linkage using 7 + 7 coils in series/parallel energized with 50Hz sine indicates about 0.89%, (close to reasonable transformer operation), which ties up with the recovery to output ratio.
Obviously the input current is way too high, likely due to holding coils on for 7 steps.

It all comes down to generating the advancing MMF wave front in the stator with the overlapping coils.

The low turns value for each coil (35) in combination with parallel coil operation is required to get current to rise in the coil within the step period. This is at odds with what is required once a flux level has been established in the stator. The flux could be maintained by coils with a large number of turns and a lower current however, these coils would be problematic as they would present a high inductance to overcome to rise the current during the step period.

If a sine current were applied to these 35T coils to establish flux linkage  i.e 100W input (88W output), then a pulsed composite signal is established on top of the sine current, synchronized in phase.
We know that at least 80% of this pulse portion of the output could be recovered.
We also know that recovery energy is proportional to load, so the real question with this scenario, is what happens to the overall input current when it is always increasing or decreasing due to the slope of the sine i.e there is no time that current is static?

Note: this would not be using half sines as the input voltage for the switches. 

I have not figured out how to apply the current sine to the coils yet.


L192

[listener192]

Attached is a waveform close to what I am trying to achieve. I still don't have a good circuit to achieve this yet. The actual waveform would be the stepped one with the trailing pulse in the pole being sequentially turned off.
This time for each coil in the series chain that is on for more than one step, a current slope is present. The pulses generate recovery when they turn off. This recovers at least 80% of the energy stored in the coil.
The continual current slope, should provide transformer like flux linkage performance.

The second shot may be a little closer, as the pulses maintain more amplitude down to the zero cross point.

L192