Has anyone seen Lasersabers new motor runs on 1000uf cap

[gyulasun]

Hi Conrad,

Re 1) 
I would be curious to know whether you have an explanation on the increase on power draw and on the simultaneous rpm decrease when the feedback diode is used (at the some ten mW input power levels or less) to utilize the energy of the collapsing field.  I also thought of a possible 'masking' effect of the low impedance coils at higher power levels to notice any increase in power draw but my common sense tended to rule out this possibility because the presence of the diode and its associated circuit components must be 'invisible' power-draw-wise to the input supply source.  I also thought that with high impedance coils, especially when they are in series to have an even higher impedance (both AC and DC wise), any circuit component placed in parallel with them may reduce their high impedance and "predestinate" the drag but again my common sense says that the switch is OFF (i.e. input power is cut to the coils) when the collapsing field is utilized so no increase in input power draw ought to be sensed. The decrease in rpm may involve a different explanation: the just interrupted current in the coil(s) is maintained by the diode (and by its associated components) till the collapsing field fully diminishes and this makes sense to me to be able to influence the original rpm of the rotor.   

Re 2)
How the DadHav circuit works (you wonder why it works at all)?  Well, let's consider first the pnp - npn bipolar transistor circuit with MPS-A56, -A06 types). These both types have a minimum DC current gain (h_FE) of 100 and not higher than 150 (from data sheet). This means their current gain is relatively well defined so it is easier to replicate most circuits with these types.

Basically the DadHav circuit is a DC coupled switching circuit but it can only operate with AC pulses injected either via the coils by induction or by injecting pulses into the base of the pnp transistor (via say a coupling capacitor).
Notice that without any AC input, there can be no DC current flowing via the transistors when you apply supply voltage because none of the transistors can receive any DC forward bias at its base (putting this otherwise: both transistors are in Class C mode of operation without any DC bias).

The moment an AC pulse is induced in the coil(s) with a polarity that is able to bias the pnp transistor into conduction via the 100 kOhm,  the npn transistor will also conduct because a bias current is injected into its base from the supply voltage via the pnp transistor and the other 100 kOhm, hence the coils can operate as electromagnets. When the inducing magnet passes the coil(s), the circuit should finish operation (if it has correct component values of course, see below) because in the lack of any AC input pulse, the lack of any DC bias prevails again.

Obviously there can be "issues" with the operation as follows: if the induction somehow can induce a prolonged AC bias into the pnp base (say the magnets are too strong or too close to the coils etc) then the electromagnets' ON time may become spuriously controlled, this is what I think you solved with using the 2.2 nF capacitor, effectively reducing the induced AC amplitude by forming a (lowpass) integrating filter with the series 100 kOhm base resistor.
The next "issue" with this circuit is the DC resistance of the coils if the resistances are relatively high: if the npn transistor is able to drive a certain current via the coils which current causes a high enough DC voltage drop across the coils, then this voltage drop may be able to insure a continuous DC bias to the pnp transistor via the 100 kOhm, meaning the pnp cannot switch off: so both transistors remain in ON state, regardless of any further magnet inductions.  To put this process simpler: the moment the npn transistor saturates it biases the pnp transistor via the 100 kOhm and this state remains because the pnp will keep the npn in ON state too: a locked situation. Consider this in you MOSFET circuit version, the same problem may occur when the MOSFET is fully ON.   

I believe that these "issues" could be remedied by using a variable resistor across the base-emitter of the pnp transistor, say a 100 kOhm potmeter first and see its effect when adjusting its value. I believe even the 2.2 nF could be removed and still get a correct ON pulse. A second 100 kOhm potmeter placed between the base and emitter of the npn transistor may also help for fine adjusting or when the first 100 kOhm potmeter at the pnp may need a too low kOhm setting and this would attenuate also the AC trigger pulse to the base of the pnp.

Now you can understand that in case the transistors have a high DC current gain, the chance for collector-emitter saturation is higher: thus the operation of the circuit may become also problematic as you found with the BC547-549 types, these may have a h_FE of 300 to 700 so the 100 kOhm resistors may have to be increased to some hundred kOhm or even to 1 - 4.7 MegaOhm values to get a better chance for correct operation.

These latter transistor types (you also refer to in your above post) recall your test with the Ossi motor 1 reed 2 transistor circuit:
http://www.overunity.com/13523/has-anyone-seen-lasersabers-new-motor-runs-on-1000uf-cap/msg363434/#msg363434  and I mention this because both transistor types are npn and none of them is pnp. If you simply mistyped the names in the schematic or in your above post that is okay but if you really used two npn types then it may explain why the circuit was willing to operate from 12V supply voltage and at higher voltages only: one of the npn transistors got a reverse bias and behaved as a 12V Zener to block any input current below 12V?  (base-emitter junctions behave as Zener diodes in reverse biased condition) Sorry for mentioning this, I do not assume you used two npn types but only mistyped the names. (originally BC238-BC558 npn-pnp types were used in the video.)

Sorry for this long post,  I hope it serves you and others fully understand the operation of DadHav's circuit.

Greetings, Gyula

PS  Re  SeaMonkey's suggestions I can agree with his 1st suggestion to reconnect the 100 kOhm resistor from the drain of the MOSFET to the emitter of the pnp i.e. if I understand well: connect the 100 kOhm between the base and the emitter of the pnp and use a suitable capacitor to couple the AC induction from the drain of the MOSFET to the pnp base, this way the DC coupling is eliminated to the base of the pnp from the drain, preventing the possibility of the locks up situation. Regarding his other suggestions the result is doubtful because you (Conrad) described your findings with tinkering with those resistors and found the best values empirically for this particular circuit. (unless you make the change mentioned in the 1st point by SeaMonkey.)

[conradelektro]

@SeaMonkey: Thank you for the suggestions I will try them.

@Gyula: Thank you for the explanations, that helps a lot.

You are right  I made a very stupid error, I thought the BC549 is a PNP transistor. What would be a good complementary pnp type for the BC547 or the BC 549?

It is very akward to use my cell phone for the Internet, will be back next week.

Greetings, Conrad

[gyulasun]

Hi Conrad,

To err is human.  At least one puzzle is solved (the needed higher than 12V supply voltage).
The complementary pnp types for the npn BC547-549 are the BC557-559, data sheet is here:
http://www.play.com.br/datasheet/BC557.pdf   

Good pnp types are also the BC327 or BC328,  these are also selected by h_FE but are not labeled with letters A, B or C but with numbers in the suffix:   -16    -25   or   -40  like  BC327-25  and see data sheet how these groups include the h_FE ranges:
http://www.play.com.br/datasheet/BC327.pdf    These pnp types have their complement npn types as BC337, BC338 with the same number suffix classifications in h_FE.

rgds,  Gyula

[Magluvin]

Had to clear the bench and some of my apartment because of tenting for termites at the other end of the building, so im just getting things all back to normal. What a pain in the butt it all was.

At work a parts and hardware guy comes by once a week. While looking for supplies in his truck, these clips for car door panels caught my eye. Pic below. The other pic is of my wire stand for coil winding and my rotor Im going to use. Its not a needle bearing but these bearings and base are the ones I used graphite in the bearings after cleaning out all the grease. Over 14min rundown from 1200 rpms. Its on YT. 2 vids because they only gave us 10 min back then.

The winding length is short and the outer diameter for winding is the size of a penny. There will be 24 coils for this build. Will be winding the coils using the laser rpm meter in count mode to count the turns of the drill to count the turns. The wire is so fine. I saw a pic of its cross section compared to the cross section of an average hair. The wire is smaller. Not quite half. Building the stator base ring tomorrow.

Working on a rotor and base for the tiny bobbins I had shown earlier, and going ahead with this build first.

Mags

[Magluvin]

Almost forgot.  These will be wound bifi. Imagine each coil has 2 wires wound bifi and 1 wire is labeled at the ends 'A' and the other wire is labeled 'B'. 
All 24  'A' wires will be wired successively in series and all the 'B' wires are wired in series, then the 'A' series circuit will be connected in series with the 'B' series circuit.

There is a reason to wire them this way that I think has a pretty cool outcome.

If I wire each bobbins bifi windings in series, where we only have 2 wires for connection with the bobbins coil, and then just connect it in series with the next series bifi coil, and then the next, we will be dividing the input voltage amongst the coils/bobbins as a whole. So we have 24 bifi coils in series and we end up with 2 wires for input(or output  ). Lets make it simple.  We will provide 24v to the motor. Well we have only 1v across each series bifi on each bobbin.    Not good. Thats only .5v between adjacent windings!

But, if we do as I suggest above, by putting all the 'A' wires in series and all the 'B' wires in series, then series 'A' and 'B' we will then have 12v between all adjacent windings throughout, not just .5v .   This is a 24 times increase in voltage in the capacitance of adjacent windings. Not a bad gain in the potential of the coils capacitive effects. Just by a trick move in connecting the ends of each bifi.     Think about it.

If some people out there have used series bifi in series with other series bifi coils, they may not have gotten to see the full potential of those bifi coils because of voltage division of each series bifi in series with the rest. But 'A' and 'B' doesnt have that problem. I think this 'A' and 'B' way of winding motor coils might be a good thing. Will see. 

And the circuit of coils as a whole will be the same resistance either way you do it. But that voltage difference between turns is not minuscule nor marginal. There should be some kind of noticeable difference.

Mags