Has anyone seen Lasersabers new motor runs on 1000uf cap

[conradelektro]

(I assume you tried to keep the rotor speed more or less the same in both cases i.e. without and with the reed switch, right?  Also what you wrote:  "Astonishing, when shortening the coils with a Reed switch the cap could be charged to a higher Voltage (2 Volt without a Reed switch, 4 Volt with a Reed switch shortening the coils."  this means the reed switch was actually in series with the coils as the schematic shows and NOT shorting the coils, right?).

Earlier I tended to accept that in LaserSaber's setup the germanium diode conducted backwards (Ge diodes indeed have a few uA reverse current) so the spike's was able to drive charging current backwards via the diode into the 10 or 1000 uF cap but now that you have used Silicon diode, this may not fully be the explanation.  Si diodes have pico or nanoAmper reverse current. 
If the scope shows a positive spike-like pulse format across the reed switch when the reed is just off, then you may try to use different capacitor values in parallel to find the best value which can give the highest DC value in the 1000 uF.

Thanks,  Gyula

I tried to always generate about the same peak rotor speed. When I went higher the rotor tilted away from the top magnet (which holds the axle in a vertical position), therefore I always tried to stay just below this mechanical limit of my motor. This "peak rpm" is visible because the axle starts to wobble.

I did the tests with and without the 100 nF cap over the Reed switch. No difference. So, this 100 nF cap is not the reason.

I did the tests with many different diodes, the best result (quickest charging of the 1000 µF cap) was with the SB3100.

The slowest charging happened with the 1N5711 which has only 20 nA reverse current. Therefore I tend to believe the explanation, that the diode reverse current goes into the coils and is "shortened" by the Reed switch, which causes a back EMF spike. This back EMF spike might cause the higher Voltage.

The circuit diagram is correct (Reed switch after the coils). I call this "shortening the coils" because the diode reverse current is "shortened" to GND. This might be a wrong terminology.

I also tried a 4 µF electrolytic capacitor (instead of the 1000 µF one) and could charge up to 7 Volt with the reed switch (and to 4 Volt without the reed switch). But the results were not consistent, I have to repeat this. 1000 µF and more have a "slowing everything down effect", which helps when doing tests and measurements.

Tomorrow I will attempt some scope shots over the coils and over the Reed switch in order to find the reason for the higher Voltage when using the Reed switch.

I also want to try this Adams Motor circuit, which you posted some days ago (is attached). But I need the "double Reed switch" for that. I have enough Reed switches, but it takes time to build a contraption holding to Reed switches so that their relative position can be adjusted.

Greetings, Conrad

[gyulasun]

Hi Conrad,

Well, the SB3100 Schottky diode has 500 uA reverse current (at 25°C ambient temp, from data sheet) so this may confirm the charging process.  This makes me pondering: what if you connect (say) a 100 kOhm resistor in parallel with the diode? (to increase greatly its "reverse" leakage...), of course this cuts back a little on the peak DC across the 1000 uF cap. (using a 1 MOhm potmeter as variable resistor may help finding the best compromise...)

Thanks, Gyula

[conradelektro]

!!!!!!! Correction !!!!!!:

Both charging circuits shown in my video http://www.youtube.com/watch?v=rFvstgIRuOA are able to charge the 1000 µF electrolytic capacitor to about the same Voltage (4 V to 4.5 V). When testing the most simple circuit (diode only) the rotor turned more slowly, therefore the charge of the cap only went up to about 2 Volt.

There is no surprising effect !!!!!!!!!

Both charging circuits published in my post http://www.overunity.com/13523/has-anyone-seen-lasersabers-new-motor-runs-on-1000uf-cap/msg362425/#msg362425 are correct, but they have about the same charging effect. The 1000 µF electrolytic capacitor can be charged up to about 4 V to 4.5 V by blowing air over the rotor.

Sorry about the stupid error. When testing the charging circuit with the Reed switch I was blowing the air over the rotor more skilfully (training effect), therefore it turned faster than in my initial tests with only a diode (no Reed switch). Today I repeated the tests more carefully and it turned out, that both charging circuits give about the same results.

Still, the diode SB3100 worked best (fastest charging of the cap with both charging circuits).

Greetings, Conrad

[conradelektro]

New cap charging attempt, see the attached circuit and photo.

This circuit allows to run the setup as a generator (no battery attached, 1000 µF electrolytic capacitor can be charged to about 1.5 V by turning the rotor with an air jet) or as a pulse motor (battery connected).

But the diode has a negative effect on the pulse motor mode. The setup consumes more power (about 600 µA instead of 500 µA) and the rotor turns a little bit more slowly.

My conclusion after my error and my tests: There is nothing surprising!

Once the rotor has very little friction one can play "cap charging games" (by speeding up the rotor with an air jet and by placing a rectifier diode in the circuit) with the expected results.

Once the rotor has very little friction one can make the pulse motor turn with very little power.

Greetings, Conrad

[gyulasun]

Hi Conrad,

The circuit diagram is correct (Reed switch after the coils). I call this "shortening the coils" because the diode reverse current is "shortened" to GND. This might be a wrong terminology.

Okay, thanks for clarifying this. I knew the diagram was ok, so the word "shorten" (to cut in length) did not cause problem. To put it simply: the reed switch discharges the 1000 uF capacitor via the coils and via the reversed biased diode (discharge current depends mainly on the reverse current characteristic of the diode and the AC impedance of the series coils).

But the diode has a negative effect on the pulse motor mode. The setup consumes more power (about 600 µA instead of 500 µA) and the rotor turns a little bit more slowly.

I think when you connect the diode in parallel with the coils, this may need a different reed switch position than when the diode is in series with the coils.  The parallel diode keeps up the current in the coils when the reed already is OFF, so the reed needs to be in a position where its own ON time is less than in the case the diode is in series with the coils. IF you readjusted the reed to the parallel diode, then please disregard what I wrote.

Thanks for the measurements. You think now the waveform across the switch is not worth checking by the scope (when the diode is in series with the coils)?

Greetings, Gyula