Has anyone seen Lasersabers new motor runs on 1000uf cap

[conradelektro]

Hello Conrad, I'm really happy to see you guys using the little circuits. As you know I can't take 100% credit for the idea. The simple circuit is in some of the self start circuits like SmartCreations and a few other have videos of. I agree with you on the LaserSaber motor. Very high impedance, large diameter rotor with strong magnets and almost no friction. Don't forget the the slow RPM's and pulse rate. I didn't pay a lot of attention but how did LS measure the slow frequency pulses. He couldn't have used a meter or did he? The double circuit is a very different story with some interesting output results as you might have noticed on the video. It would really make me happy to see you guys make some improvements and have some fun with something a little different than what we've been doing. The circuit is in the most basic form and needs adjustment for sure. If you notice the very interesting aspect of how it ran the model airplane motor it might show some potential. The down side of the circuit as it would pertain to a motor having some torque is that if you load the shaft and slow the motor you also loose the bias on the driver transistor and in turn the power transistor gives in to the load rather than try to power through the slow down right? Anyway guys I hope you don't mind me hanging around a little bit to see how you're doing. I've thought about making a motor like LaserSaber's just because I think it's really cool and it could be made to be a kinetic work of art. BUT, I've had my motors to the point where adding only a few RPM's charges instead of discharges the capacitor many times, it's on one of my videos. Getting the little extra you need has to come from Unobtania or Unobtainium. Ha, even if the Unobtainium is a Lenz free coil. There has to be an answer somewhere once you get that close. Good talking to you guys.
John Hav.

Hello John,

I spent hours with your videos admiring your handiwork and the machinery and tools you seem to have (both mechanical and electronic).

For me the strange motors and circuits are a form of art comparable to sculptures and paintings. The OU aspect is only a goal, like perfection, which can never be reached.

And with YouTube and various forums on the internet, ideas travel quickly around the world.

Greetings, Conrad

[conradelektro]

DadHav's circuit with a MOSFET:

I could make it work with a MOSFET, but the result is pretty strange if not stupid. See the attached circuit.

The green LED has a forward Voltage of 19. to 2.1 Volt, and just that is necessary to make it work. It does not work with a white LED and not with a red LED, but starts to work a little with a yellow LED.

I guess the answer is a 2 Volt Zener diode and some other resistor arrangement. And I admit that I do not understand this circuit.

To all the electronics wizards: any ideas?

In case the Gate of the MOSFET is bound to GND with a 100 K resistor and the 2N3906 just switches the LED (10 K resistor removed), one sees that the 2N3906 is very sensitive and switches on/off with a slight movement of the magnet near the coil. The difficulty is to connect the Collector of the 2N3906 in a way to the Gate of the MOSFET which allows switching. Many things I tried (specially the obvious ones) lead to a stuck MOSFET and a stuck 2N3906 (both are initially off, but then stay continuously on after a slight movement of the magnet near the coil).

Greetings, Conrad

[gyulasun]

Hi Conrad,

I am back from travelling. Just read through the posts from last week and this week.

Referring to your test circuit here http://www.overunity.com/13523/has-anyone-seen-lasersabers-new-motor-runs-on-1000uf-cap/msg363634/#msg363634 it is strange that current draw increases a little when the electrolytic capacitor is charging via the feedback diode 1N5711 because the reed switch is OFF when this charging is happening, right? It seems as if the AC impedance of the 100 uF capacitor appeared in parallel with the coils as an additional load, in spite of the fact that the diode represents a reverse direction to the supply voltage when the reed is ON and this reverse direction would normally isolate the loading effect of the 100 uF capacitor during the On times of  the reed switch. I do not mean your 1N5711 diode has had more leakage current than 200 nA at 50V reverse voltage from data sheet of course, I assume it was ok.  I notice though when you used a white LED as the feedback diode instead of the 1N5711 that a white LED has a reverse breakdown voltage anywhere between 6 to 8 Volts (type dependent) so an isolation in that case may have not been fully insured. But anyway, with the 1N5711 diode you found an increase in current draw and a decrease in RPM and these are strange I cannot explain. 

Referring to your MOSFET version test circuit above.  The reason that above the 2.1V magic threshold voltage both transistors are stuck (i.e. both transistors saturate) is that the DC feedback via the 100 kOhm from the drain of the MOSFET to the base of bipolar transistor is able to become too much and this excess feedback current keeps both transistors ON.
I mean when the MOSFET switches ON first (due to the induction from the magnets) and the DC voltage drop across the DC resistance of the coil(s) happens to be already enough as a bias voltage to keep the pnp transistor ON via the 100 kOhm, then there is nothing to bring back the DC bias OFF situation for the pnp transistor.  The pnp transistor will have a collactor current, this will maintain a 'perpetual' DC bias voltage level between the gate and source of the MOSFET, so drain current stays also ON.

You nicely found the narrow voltage level gap your circuit is able to operate, this starts from just below 2V (1.9V as you found) and goes up to about 2.1V. The lower limit 1.9V comes from the MOSFET's threshold gate-source voltage (Vth) below which no drain current can flow, and the upper limit 2.1V comes from the coil(s) DC resistance because at higher than 2.1V DC gate-source bias voltage the drain current seems to be causing just enough DC voltage drop across the coil(s) to keep the pnp transistor permanently ON (via the 100 kOhm), hence the DC feedback loop closes permanently and both transistors saturate, the induced AC cannot influence collector and hence drain current any more.

What could be a remedy here? Maybe use an Si diode or two in series with the 100 kOhm where it connects to the base of the pnp transistor, just unconnect the resistor and insert the series diode(s) to complete the feedback path again. This way the 0.6V bias voltage for the pnp becomes shifted up to 1.2V (or 1.8V with two series diodes) and this way you may defeat the biasing effect of the DC voltage drop of the coil(s). I believe the induced AC voltage by the magnets will still have the needed triggering effect (unfortunately the series diode(s) also reduce the AC voltages but maybe not so much as to be useless for controlling the motor).

Greetings, Gyula

[conradelektro]

1) Referring to your test circuit here http://www.overunity.com/13523/has-anyone-seen-lasersabers-new-motor-runs-on-1000uf-cap/msg363634/#msg363634 it is strange that current draw increases a little when the electrolytic capacitor is charging via the feedback diode 1N5711 because the reed switch is OFF when this charging is happening, right?

2) What could be a remedy here? Maybe use an Si diode or two in series with the 100 kOhm where it connects to the base of the pnp transistor, just unconnect the resistor and insert the series diode(s) to complete the feedback path again. This way the 0.6V bias voltage for the pnp becomes shifted up to 1.2V (or 1.8V with two series diodes) and this way you may defeat the biasing effect of the DC voltage drop of the coil(s). I believe the induced AC voltage by the magnets will still have the needed triggering effect (unfortunately the series diode(s) also reduce the AC voltages but maybe not so much as to be useless for controlling the motor).

@Gyula:
Thank you for looking at my tests. You seem to be the only one who is still interested, the crowd has moved on. May be people come back once Lasersaber posts a new 3D printed motor.

Ad 1) Whenever I did experiments with a high DC resistance drive coil in a pulse motor (which means low power draw), I observed that a feed back LED or diode over the coil caused about 10% more power draw. I think one does not see that with a 10 Ohm DC resistance coil and a power draw of several Watt. It only becomes visible (measurable) at an overall power draw of  100 mW or lower.

Ad 2) I wonder why Dad Hav's circuit works at all. See for instance the circuit at http://www.overunity.com/13523/has-anyone-seen-lasersabers-new-motor-runs-on-1000uf-cap/msg363823/#msg363823 and the 2.2 nF capacitor I needed to make it work nicely (Lidmotor's transistor choice). And I could not make it work with a BC549 and BC547. It is always strange if one needs a very particular transistor, it should work with many different transistors (e.g. general purpose transistors). I could not yet try the MPSA56 and MPSA06 but I will.

I would like to make Dad Hav's circuit work (without any additional sensor, like Reed switch or Hall sensor or trigger coil) with a MOSFET to be able to run bigger pulse motors at 20 or even 100 Watt. It is also an exercise in circuit development which I want to pass successfully. If I can do that I might understand this circuit better. And my little ring magnet spinners look really neat and clean just with one coil, nice items to show off. Further I would like to build a "ball magnet spinner" just with one coil (and then a generator coil) as is discussed in this thread http://www.overunity.com/13576/tragic-accident-with-ball-magnet-emdr-motor-generator/#.Uc8L6CnmTIU .

I can not do experiments for a week, only read and write occasionally on my cell phone. But your suggestions will be tested.

Greetings, Conrad

[SeaMonkey]

Gyula has done a good job of explaining
how the circuit locks up when initially
stimulated with a pulse.

To overcome the lockup there are a few
things which could be tried:

(1)  Re-connect the 100K resistor which
is presently attached from the base of the
PNP transistor to the drain of the MOSFET
from the base of the transistor to its emitter.
Then install a capacitor from the base of the
transistor to the drain of the MOSFET where
the 100K resistor had been previously attached.
It may take some trial and error to determine
the best value of capacitance in the feedback
loop.

(2) To enhance MOSFET switching the 10K
resistor presently in the Gate input could
be removed and replaced with a much smaller
resistor or possibly even just a wire.

(3)  The 100k resistor presently attached to
the MOSFET from Gate to Source could be
reduced to 10K, again to enhance MOSFET
switching.

(4) Use an LED in the circuit with the highest
forward voltage (or two or more LEDs in series)
such that it/they just faintly illuminate/s when
the circuit is operating to avoid depriving the
MOSFET of adequate drive signal.