Confirming the Delayed Lenz Effect

[PiCéd]

@PiCéd: A coil which rotates needs two brushes to make the connections. Mechanically this would be a challenging task.


A new spin and power consumption measurement with the set up and drive circuit disclosed in my Reply #952:

0.2 Watt (13.5 V and 15 mA on average) cause the rotor to spin with about 6000 rpm (100 Hz).

I made better supports for the two Hall sensors, but the transistor H-bridge will have to wait for P-Channel MOSFETs which should arrive next week.

For the brushes I know that is the only way for this configuration.

For acceleration, a test into a short circuit and a non short circuit is not sufficient, a test with and without the coil with core is much better, with of course approximately the same input energy.

[gyulasun]

0.2 Watt (13.5 V and 15 mA on average) cause the rotor to spin with about 6000 rpm (100 Hz).

Hi Conrad,

Nice progress in rpm with less input power to your pnp bipolar transistor test setup wrt your reply #952 and #955. It occurs to me a test from RomeroUK or Doug Konzen to place a permanent magnet behind the coil core and see with the distance how rpm and current draw change.  EDIT: I think this does not work with alternating magnet poles, sorry.

Re on the horse shoe core suggestion: now that you have found a higher distance of the coil from the ring magnet can give better result the size of a possible C or U shape ferromagnetic core may increase even higher than earlier I imagined from your setup sizes. Higher sizes C or U cores from ferrite (size of 13-15cm in length between the U prongs or maybe higher) cost a lot unfortunately. So I cannot suggest any good solution at the moment, perhaps at ebay there are some offers on flat ferrite cores with rectangular cross section like the one used for ferrite antenna in portable radios in the past, from such shapes somehow an rectangular U shape could be built perhaps.

One more thing: perhaps it would give you a more precise Hall sensor positioning possibility if you built a small (OD=3 or 4cm) 'control' disk with small cylinder magnets embedded, this disk would rotate together with the ring magnet of course on the same shaft. In one of Bedini videos you can see such control disk on which the small magnets control a reed switch but your Halls can also be controlled like that of course.  see here http://www.youtube.com/watch?v=4TICXxP1jI4  By using 3 or 4mm OD cylinder magnets on the small disk the control could be more precise than you have now.

As long as your input current is in the some ten mA range (say not higher than 20-30mA) the I*I*R_DC coil loss may not be high, especially if you manage to use coil(s) with well under 100  Ohm winding resistance. (20mA on 100 Ohm gives 40mW loss). And it is a good idea to use separate generator coils which may have thicker wires and less number of windings wrt the relay motor coils.  Naudin or Ossie have used air core coils manufactured for audio cross over filters, these are pricey, unfortunately, if you cannot wind them.

rgds, Gyula

[synchro1]

The two battery system for switching polarity in the horseshoe core coil of Igor's motor is fiendishly simple. A DPDT reed switch could run the entire motor placed near the rotor magnets. The core volume increases lag time. The Tesla bifilar reduces resistance to change in current direction in the coil to zero.

A large industrial DPDT reed switch could be wired in series with powerfull batteries, and turn a large size inertial rotor. The Horseshoe core could be a back yard throwing type.

[conradelektro]

@Gyula:

Thank you for the many suggestions, I collect them and will try to incorporate the suggestions in Version 2. The current version and your comments willy teach me how to do it right eventually.

At the moment I am playing with my set up and the attached circuit (because the P-Channel MOSFETs which I ordered have not arrived yet).

From the Bedini Motor circuits I took the idea to use a small Neon bulb in order to cut off the "spikes at the coils switch off times" (which you also mentioned in connection with choosing a MOSFET). It seems to be a good idea, because I measured spikes up to 140 Volt (with a scope).

Then I tried to cut off the spikes with a LED or a diode over the coil (alternatively to the Neon, see the circuit diagram). The spikes are nicely suppressed but the rotor slows down by 30%.

The Neon is visible on the photo (on the breed board).

Concerning the planned transistor H-bridge for commutation, see my Reply #933 on April 08 (note, the pull up resistors are missing): I might need some logic circuitry to avoid the simultaneous switching of the two Hall sensors (because that would cause a short circuit in the transistor H-bridge). Tests with the two Hall sensors are under way. It looks like I can place them just right to avoid simultaneous on conditions (but it is potentially a bad situation).

The integrated motor driver H-bridges all have internal circuitry to avoid accidental shorts in the bridge.

Greetings, Conrad

[gyulasun]

Hi Conrad,

The use of a Neon bulb across the collector-emitter came from the fact that the suggested switching transistor V_CE rating is only 60-70V (for 2N3055 or similar types) and a simple means had to be found to defend the transistor. In case of a HV transistor type the Neon bulb would not be needed at all, however a HV transistor costs more and is not readily available everywhere. Your 2N6111 transistor is a 40V device, it is fortunate it has not got ruined from the higher voltage spikes.
The explanation why you did not find RPM reduction whenever the Neon was ON is that it does not influence the transistor ON time in any way, it is always an open circuit whenever the voltage difference (either AC or DC) is below its trigger voltage (70 to 120V, depending on Neon types). And in the schematics you or many experimenters use for such pulse motors the peak pulse voltage can only go up into that voltage range when the switch-off time is quick enough to interrupt the current flowing in the coil during the transistor ON time (you surely know induced voltage at switch-off is L*(dI/dt) where L is the coil inductance, I is the electric current change and t is the switch-off time.)

Regarding your LED in parallel with the coil, it is the reverse voltage breakdown rating for the LED which clamps i.e. limits the battery voltage when the transistor is ON. LEDs normally have 5V to 7V reverse voltage rating, so higher than 5 or max 7V across a LED in reverse direction cause a reverse current draw like in a Zener diode (and it may ruin the LED) so the (say) 11V supply voltage cannot reach the coil, this is why your rotor slows down. (If you connect LEDs in series, then their breakdown voltages add up of course, just like their forward voltages do.)
IF you use a normal Si diode (say 1N4007) instead of the LED in parallel with the coil, then albeit there would be no problem with the reverse breakdown voltage/current, the diode would change (delay) the switch-off time by letting the spike voltage drive current further on in the coil, hence magnetic field would be present for a longer time. To compensate for this, you would need to reposition a little bit the Hall sensor wrt to the best position for the no diode case.

Regarding your worry about the simultanous ON condition for the H bridge, yes this is a real possibility and if happens then any of the two MOSFEts under each other can conduct and short the supply voltage, i.e. both MOSFETs may get ruined. If your power supply has no built-in adjustable current limiting feature, then it is a good idea to use a resistor in series with the supply output to limit the current below the max drain current rating of any one MOSFET during the adjustments and later to short or omit the resistor. (Simply use Ohm's law to know the resistor value, considering data sheet continuous current value for the MOSFET and the max supply voltage you may use for sensor adjustments.)
I agree that the  planned circuit for your H bridge is a potencially hazardous situation for the MOSFETs, although with great care (and maybe with the series resistor) the MOSFET damage could be avoided.

rgds, Gyula