Self accelerating reed switch magnet spinner.

[conradelektro]

@Conradelectro,

Quote from you:

"We will soon all eat the same gene manipulated and poisoned food out of gigantic factories".

Consider moving to Costa Rica; Pura vida!

There must be some very nice tropical areas in Puerto Rico and growing food should be easy with may be two harvests per year. Agriculture should still be more natural than in central Europe. The travel agencies call Costa Rica a tropical paradise and recommend hiking and trekking tours.

As an old man I would probably not support the climate very well and getting used to the situation there would require more strength and adaptability than I have left. Never uproot old people, they tend to die.

Greetings, Conrad

[synchro1]

       "The small tropical nation of Costa Rica celebrated 65 years without a permanent standing army this past December, and now, they're moving toward another important milestone: A legal framework for becoming a 100% GMO-free country".

[conradelektro]

I am still playing with the LM311 because more modern Comparators and OpAmps which I ordered did not arrive yet.

After learning about offset and hysteresis I changed the circuit and could reduce oscillations. But there still are are few ups and downs when the drive coil is switched off (see the attached scope shots).

After much testing I came to the conclusion that the back EMF spike of the drive coil causes these few remaining oscillations of the LM311 although I cut off the back EMF spike with a reverse LED. Just the presence of the drive coil in the circuit (between positive rail and collector of the internal transistor of the LM311) makes it prone to self oscillations.

It is a bad idea to switch the drive coil with the LM311 (a separate robust NPN transistor should be used) but I wanted to test offset (pins 5 and 6, BALANCE) and hysteresis (positive feedback) in a minimal circuit.

I also think that one should create an external offset between positive and negative input of the OpAmp (as TK is doing it in his TL082 circuit) instead of using an internal BALANCE, it gives more flexibility.

The 100 pF cap between positive and negative input of the LM311 is very necessary to stop oscillations when driving a coil (cap can be between 100 pF and 1 nF). Offset (BALANCE) has to be set in a position which keeps the internal transistor in the off state in case there is no signal from the trigger coil.

If I switch only a LED with a 2 K resistor (no drive coil, no reverse LED) as indicated in the circuit diagram I get absolute clean switching and the 100 pF cap can be removed. By help of BALANCE (offset) I can set whether the LED is on or off in case there is no signal from the trigger coil.

I hope this makes sense, for me it was an exercise in OpAmp circuit design.

Greetings, Conrad

[TinselKoala]

I'm amazed at the distance you are getting. The op-amp sense system can be made very sensitive, but the drive coil coupling to the magnet rotor is what is really amazing. That is some respectable distance there.

Meanwhile, here's another version of the same kind of circuit, using a Cd-S photoresistor and ambient light, to detect the shadow of "something"  and trigger a high-current electromagnet. The 7809 voltage regulator keeps the op-amp sense circuit stable when the mosfet fires, and the "boost" power supply allows the electromagnet to use a much higher voltage than the sense/control circuit needs. The mosfet is a 55 volt, 110 amp part in a small TO-220 package. The mosfet runs cool, no heatsink seems necessary at the moment. The DC resistance of the electromagnet is about 29 Ohms, with wire scrounged from a TV deflection yoke ferrite wound around a u-shaped core of mild steel rod.

[conradelektro]

I'm amazed at the distance you are getting. The op-amp sense system can be made very sensitive, but the drive coil coupling to the magnet rotor is what is really amazing. That is some respectable distance there.

Meanwhile, here's another version of the same kind of circuit, using a Cd-S photoresistor and ambient light, to detect the shadow of "something"  and trigger a high-current electromagnet. The 7809 voltage regulator keeps the op-amp sense circuit stable when the mosfet fires, and the "boost" power supply allows the electromagnet to use a much higher voltage than the sense/control circuit needs. The mosfet is a 55 volt, 110 amp part in a small TO-220 package. The mosfet runs cool, no heatsink seems necessary at the moment. The DC resistance of the electromagnet is about 29 Ohms, with wire scrounged from a TV deflection yoke ferrite wound around a u-shaped core of mild steel rod.

@TK:

I am still playing with OpAmps which open up some interesting "trigger" possibilities, as your latest circuit shows.

From the capacitors and the 7809 voltage regulator in your latest circuit I deduct that you have problems with self oscillations. What ever I try, there are self oscillations which have to be worked out carefully.

My latest idea is that a trigger coil has to be AC-coupled. At the moment I have little time, but I will post the circuit soon.

What interests me most is the very low power demand of the ring magnet spinner I show further up in this thread http://www.overunity.com/13852/self-accelerating-reed-switch-magnet-spinner/msg376292/#msg376292 .

The LM311 consumes about 7 mA at 10 Volt, which leaves only 3 mA for driving the ring magnet spinner at about 2000 rpm. I ordered the MAX931 comparator which should only consume 4 µA. Together with magnet bearings I should be able to get a very low power draw for a ring magnet spinner. The MAX931's unique output stage continuously sources as much as 40mA which is enough to drive my spinner without any other component.

Once I achieve the very low power draw (I hope for less than a 1 mW) I want to try various ideas for recuperating some energy into a drive capacitor.

Greetings, Conrad