Confirming the Delayed Lenz Effect

[conradelektro]

@Gyula: thank you very much for looking up some Schottky Diodes, I ordered them. The scope shots over the drive coil with a diode and an electrolytic puffer capacitor look more or less like the ones from my Reply #1181 on: April 21, 2013, 10:48:15 PM. I will make more scope shots with my new setup which is almost finished.

I could put the two drive coils in series, then the power draw would drop. But it is more realistic to put them in parallel.

In this pulse motor http://www.youtube.com/watch?v=MEEjlYvZ5OM I put all drive coils in parallel.

In the attached pulse motor I had the toroid drive coils in series plus a resistor to adjust power draw (resistor values 5 Ohm to 50 Ohm). The toroid drive coils are of course the Steorn design. Strange enough, the toroids worked best in the depicted arrangement. But ordinary drive coils worked better. I could adjust the pulse width with two opto-gates, but it turned out that the pulse width was not very critical (neither for power draw nor for torque), therefore I switched to a Hall sensor (which is much simpler than opto-gates). I was very proud of the drive circuit (see the attached PDF-File).

I still have to try a sensor coil for my recent setup. I am a big fan of sensor coils, pulse width can be adjusted by distance from the passing magnets. Pulse timing can be adjusted by shifting the sensor coil in relation to the theoretically good position.

I built some pulse motors with Reed switches, but for me it looked like the Reed switches were difficult to control and only reliable at very low frequencies and very low power draw. One can build small things that turn slowly with very little power, but for more demanding situations Reed switches were not useful for me. For very low power applications I would now use the transistor 2SK170 or ALD110800 / ALD110900 and a sensor coil instead of a Reed switch, much cleaner.

@Synchro: I am aware of Mopozco's TROS motors, but I did not test them yet. I will try with my new setup (two drive coils). My main concern is not component count but low power draw and clean switching (for long duration tests, nothing should get hot or strained).

Greetings, Conrad

P.S.: my first setup will of course stay as it is for comparisons.

[TinselKoala]

@Conrad: That's a very nice pulse motor of the core effect kind. You have taken advantage of both poles of the rotor magnets, something Steorn didn't manage to do with their simplistic design, and you've also relieved a lot of bearing problems by making the "bump" caused by magnet-core interactions happen symmetrically aiding and in the plane of rotation, rather than radially and in opposition. Your physical layout has several advantages over the "Orbo" configuration. When I made "Orbette 2.0" I was interested in doing a precise replication of Steorn's motor performance, so my core-rotor geometry was set up like theirs, with some additional adjustability. The motor worked well with Hall sensor timing but I needed optical timing for various reasons, and wound up using 2 stacked USDigital optical encoders, one with large segments for pulse timing and the other with lots of tiny lines for data output-- rotor RPM and acceleration data.

Seeing your design is inspirational. I wish I had my lathe and mill here; I'd make one on your plan, to add to my collection. May do so anyway if I can figure out how to make the parts with the necessary precision.

It would be neat to compare the timing and dwell positions that worked best for your layout, comparing the regular coils and the toroidal core coils.
The regular core coils can be made to work in either attraction or repulsion, or in your design even both, depending on pulse polarity vs. rotor mag polarity, but the toroids should only work by core effect, and the polarity of the pulse to the toroids shouldn't matter.

ETA: I meant to also say that the optical commutator can be modified for dwell (pulse width) control very easily. Simply make another disk with the flags (or slots, or holes) , and stack and overlap them. Then your dwell can be controlled by the degree of overlap of the two discs.

[conradelektro]

@Conrad: That's a very nice pulse motor of the core effect kind. You have taken advantage of both poles of the rotor magnets, something Steorn didn't manage to do with their simplistic design, and you've also relieved a lot of bearing problems by making the "bump" caused by magnet-core interactions happen symmetrically aiding and in the plane of rotation, rather than radially and in opposition. Your physical layout has several advantages over the "Orbo" configuration. When I made "Orbette 2.0" I was interested in doing a precise replication of Steorn's motor performance, so my core-rotor geometry was set up like theirs, with some additional adjustability. The motor worked well with Hall sensor timing but I needed optical timing for various reasons, and wound up using 2 stacked USDigital optical encoders, one with large segments for pulse timing and the other with lots of tiny lines for data output-- rotor RPM and acceleration data.

Seeing your design is inspirational. I wish I had my lathe and mill here; I'd make one on your plan, to add to my collection. May do so anyway if I can figure out how to make the parts with the necessary precision.

It would be neat to compare the timing and dwell positions that worked best for your layout, comparing the regular coils and the toroidal core coils.
The regular core coils can be made to work in either attraction or repulsion, or in your design even both, depending on pulse polarity vs. rotor mag polarity, but the toroids should only work by core effect, and the polarity of the pulse to the toroids shouldn't matter.

ETA: I meant to also say that the optical commutator can be modified for dwell (pulse width) control very easily. Simply make another disk with the flags (or slots, or holes) , and stack and overlap them. Then your dwell can be controlled by the degree of overlap of the two discs.

@TinselKoala:

I built that Steorn replication thing in 2010 during the Steorn Hype. Yes, a magnet pulls itself towards the cores of the two Toroids and would get stuck between the two Toroids. A current pulse into the windings of the Toroids (no matter what polarity) frees the magnet, so that it can travel on. The torque depends on how strongly the magnet pulls itself toward the Toroid cores (strength of the magnet, size and material of the Toroid). Once the current pulse is strong enough to completely "mask" the Toroid core (so that the magnet is not slowed down when leaving the position in between the Toroids), the torque will not increase any more.

In contrast, with ordinary coils one can increase the torque by increasing the current theoretically ad infinitum, because the coil will pull and push the magnet depending on Watts fed into the coil (if the timing is right).

But I never did systematic comparisons. Once I saw the limitation of torque by the core effect of the Toroids I wandered on to other projects.

And I had mechanical problems with the ball bearings. I find it difficult to set ball bearings correctly. When turning the axle to fit into the centre of a ball bearing I either get the diameter too large or too small. If it is too large I ruin the ball bearing when hammering in the axle and when too small the axle rattles.

I found that the motor of a hard disk drive (the bigger the better) makes a good bearing for a disk which then carries magnets for a pulse motor. But balancing is a problem. A disk is a big thing which can produce a lot of force when tumbling.

Now I fiddle again with ball bearings for my latest attempt to make a fast "ring magnet spinner". The first version which I showed e.g. in my Reply #1194 on: April 22, 2013, 06:40:53 PM, has a rattling axle and it spins up to 9000 rpm with a maximum power draw of about 0.5 Watt. It spins at 2000 rpm with 0.1 Watt.

The thing I am building now is about the same but has two coils (the idea is derived from my Steorn replication, that is why I showed it although it is old news). See the attached photo, the ring magnet and the Hall sensor are not yet mounted. The drive circuit is shown in my Reply #1202 on: April 23, 2013, 01:34:13 PM. This time I got the axle right but the ball bearings I used seem to have some heavy grease in them, so they resist turning quite strongly. May be I can get the grease out by heating them up a bit. I will report how it goes. This latest design allows me to change ball bearings more easily, so I hope to succeed eventually. I have to look for these model helicopter bearings. I have a collection of ball bearings, but they are low quality and some are decades old.

The white coils I am using for this latest design are from 12 V relays and have 90 Ohm DC resistance, but I can also mount the 260 Ohm DC resistance coils from 24 V relays (as in the first spinner, see my Reply #1194).

The goal of this "ring magnet spinner" is to test various "Lenz free coils" (if they indeed exist, which I doubt). Success would be a fast "ring magnet spinner" that uses very little power. Torque should not be an issue with a "Lenz free generator coil".

Greetings, Conard

[profitis]

hey tinselkoala,kondralektro,ive often wondered what would happen if we used a steorn setup,using normal inductor coils instead of toroids,and simply short-circuit the coil as the magnet passes by,would this counter the magnetic attraction?any ideas?

[ALVARO_CS]

Hola conradelektro

I use to submerge the bearings some five minutes in acetone or in thinner to clean the grease.
After clean I use any light mineral oil eg. the one used to lubricate sewing machines.
About rotors, I prefer to use the ones that come in old video players, they act also as flywheels.

cheers