Re: Here it is ! The first documented selfrunning overunity motor w/o batteries « Reply #492 on: Today at 12:27:45 AM » Reply with quoteQuote Hi Guys, Ben has done a great work on this device and here are his latest findings with pictures. Hi All, Where to start. I am going to load a series of photos in my section label starting with Fig. 1 and go up from there. It will show the progression of the motor, where it came from and how it has ended up. I will all assume the group has a basic understanding of how the SSG or Bedini Blocking Osc. motor works. My motor is built with 10 alternating N/S/N/S Neo 45 magnets on a small rotor at 36-Degree intervals. Basic specifications/parts: 10 magnets 1 Coil, Litz, wound with # 27/44 cotton-covered wire. 1 Transformer Bogen T-725 1 Transistor MPS 8099. 1 FWBR bridge 1N4007's 1 Diode mass 1N4007 X 10, part of RE internal loop. 2 pots 10K 10 Turn 1 "conditioned battery" U1 size, Wall Mart Special. Conditioned means DEAD by discharging to 1 V repeatedly. This results in a high impedance primary battery, a large capacitance and a high potential capacity secondary battery all in package or "Node". This motor absolutely started off life as a Bedini single ended device to replicate Mike's motor. I soon became disillusioned but decided to stay positive and well "geter done" myself. First as all my previous post showed, I explored the abilities of my motor, how fast it could run how slow it could and on/at what power levels. I went from 8,000 rpm where it threw a magnet to 20 rpm at 1 ma DC. The coil on my motor is very special but not impossible for others to make, as it is built out of 27 strands of #44 wire, wound in the Litz configuration. Yes, that is 27 strands of #44 wire which looks to be smaller than a human hair. Out of the 27 strands, I chose 4 to be the drive coil and 22 to be the motor coil. Hard on the eyes and back ringing out those 4. Not easy to do but it can be done. This also leaves us with a coil that is extremely close coupled, very powerful magnetically with low hystersis using the Neo magnets. I am not at all certain this is the only way to do it, I suspect the low PM field motors would work but Rpm's are a priority here as within each passage is a burst of Oscillatory action the produces the RE and more pulses=more RE. My motor at 1200 rpm produces 90,000 RE pulses/min! Yes, that is 15 pulses/magnet passage, 5 N magnet passes/rev X 1200 or 75 pulses/rev X 1200=90,000 pulses! A few pulses here and there just don't hack it in the real world. Pulses=Power! Lets start of with basic schematic. Fig 1. It's been around, works great, no problem. Well it has problems if used with NEOS but it basically is a real buzzer. Thanks to John for his insight in making this system available to us. It had switching speed problems, current problems if used with NEO's, losses in the core but still a good design. Losses in the base diode etc. An excellent place to start. Now If I mutilate any schematic, I am drawing them from memory……bear with me. Fig. 2 is the Bedini/Cole/Window motor most basic design. It is much better as it is air core, more open coupled, free running and runs at relative low rpms on relative low power if used with ceramic magnets and the correct resistance in the motor base circuit. The use of Neo magnets induces higher currents in the L2 circuit along with heating of transistor, high hystersis losses, etc. There is a trade off here in what to use. Fig. 3. This was my first step in reducing current in motor with Neo magnets. As I discovered my motor would run with multi K ohms in the base circuit of my single ended circuit, I did NOT need the diode in the base. That is with the MPS 8099. I offer no guarantees with any other transistor. 8 days of 24 hours a day run time is living proof of this. It works without it and lowers drain in the circuit. OK this got rid of excess current in the base. Fig 4. With a 6 ohm coil, a 30% duty cycle, I still had about 150- 200 ma in the collector circuit when my motor was running. I found at this time that with the correct resistance, I could adjust my motor to pulse like heck while the N magnet was passing with VERY clean 50us pulse on, 100us pulse off, up to 15 times each power magnet passage and still run. All waveforms if show later will be collector to ground. THIS REDUCED MY current down to 50-80-ma average. Still too high. At this time I had a potent RE generator but it was going everywhere and nowhere at the same time. Even though, it appeared to be 97-98% efficient like this it was still way out of the park. Fig. 4. STEP 2. I added a diode across the coil. Normally this can be added here or across the transistor but I wanted to keep all that energy IN the coil area. This is commonly used to short out BACK EMF and protect the transistor but we do NOT want to protect, we want to capture the generated back EMF and the RE produced by the multitude of pulses in the inductor/transistor interface. This DIODE LOWERED the average current down to where I could run on about 12-18 ma! at 7-800 rpm. AND IS ABSOLUTELY NECESSARY IN THIS CIRCUIT. Curiously, I could measure current in I-1 of 15 ma and current I-2 would be in the 45-ma range. There was a ratio of 3/1 circulating currents in L2/D1 node but there was no way to use them. This ratio will be found to be VERY important as we go along. Now we have 1 power node consisting of a funky battery, inductor/diode system in series with a transistor NPN junction and return to the battery. A secondary base emitter junction which I will let die a slow death here due to the high resistance, low current in it. It might have importance but have not explored it at this time. This INTERNAL node (D1/L2 )with all the power being wasted in it was the key but I was damned if I could figure out a way to get at it. I tried a third winding, bridge with poor output; I tried large car ignition coils with limited success and many times thought I was unity only to be disappointed in the end. In the thousands of questions I asked my self one answer stood out. What if I could take part of that series circulating pulsating node RE and back EMF and steer it back on the battery node? How to do it was the question. It is a basic low resistance area with the driving coil being only 6 ohms so whatever I did had to be low resistance. I tried switches, I tried open-ended coils, I tried SCR's Triacs, Diacs, SIDAC's, Voodo and cursing under my breath and nothing worked. With some gentle review from Tom Bearden in his letter, I found my self thinking about nodes. RE nodes, Curentless nodes……..What? no current? Negative electron flow, my mind got to flail away at a seeming impossible problem. That damn battery/Cap/RE regenerator node had its grip on me and I had to find a way to channel some of this RE back on itself. My motor was Soooooo close, all I needed was .1% of SOME kind of energy to kick it over the top. I knew that conventional systems with their closed loop just DID NOT WORK. Lentz law/effect saw to that…. Pure regeneration from a generator action will NEVER work. Depression, lack of sleep, mind in hyperdrive and lets try this, lets try that lets try impedance matching instead of current or voltage. Let's try a LINE matching transformer. A transformer it is but with a twist. I won't go into its theory just let's say it passes RE like a champ. Here is the final Multi Node representation of the motor in: Fig 4. Notice there are 3 basic major nodes and 1 sub node, unexplored. Loop 1-4. Loop 1 is the basic normal motor node, real voltage and real current. Loop 2 is an internal node/loop, lots of voltage, lots of current and lots of RE if my basic theory is correct. REMEMBER all this is a bunch of theories I have from observation. I do NOT know them to be facts. Then there is Loop 3, The weird one! We will first discuss what I can measure and what I can not and what effects are noticed. Then I will discuss the circuit. There is a DC potential developed in this node referenced to ground that varies from basically the battery voltage up to 150 volts depending on the loading resistance in the feedback loop. Right now I have approximately 3000 ohms in series with the output. I measure .210 V drop across that resistor which also results in a measured .00007 amp (.07 ma) which is almost 1.47 millawatts! power output or drain back into the central battery node. This definitely will NOT move mountains! BUT when I connect this node back to the central node/ battery the motor responds as if it was about a 10 ma boost in current and appears to ignore the resistor. Without the node, I was running 15 ma in the motor and 750-RPM. 45 ma in Node or Loop 2. WITH THE NODE connected, I can drop the current in Loop 1 to 5.5 ma, Loop 2 node current is now about 16 ma (ratio of 3/1), loop 3 real current .07 ma, RPM 1200! And voltage constantly rises in the central node/battery! About 1-3 mv every 15-min until some point is reached I assume, I have not reached it yet. It has risen all day. Also adjusting the resistor in Loop 3 DOES NOT EFFECT CURRENT IN LOOP 1 or two or motor speed! Motor loading does NOT effect Loop 3,2 or 1 except as it slows down the current goes DOWN. There is obviously something going on in loop 3 that does NOT show up on normal metering devices but it does show up in output of MOTOR! MOTOR RUNS FASTER WITH LOOP 3 connected but draws less current in LOOP 1 by a factor of 2/3! I assume that there is some form of power flowing there that is non-conventional. I hate that word assumes! But that's the way it is. OK, lets get down to the meat of the question. Here is the final schematic. Fig. 5. A couple things should be noted. T-1 is a Bogen T-725 matching transformer. Many more wires and combinations are available, These are simply the ones I chose for maximum output and minimum loss in the motor circuit. Notice all the diodes across L1-B and T-1. I found that as I increased the diode count, the current in the loop went up until I reached about 10 diodes then it leveled off. Anything to increase the internal loop current. I have found that the 10 UF cap across the FWBR is not needed and tends to slow down the loop control. It works just fine without it. P-1 and P-2 are 10K 10 turn pots. I think P1 is about 2.2 K but that is just a guess right now. Lower the resistance to start motor, increase as far as possible to lower current in loop 1 while still maintaining RPM's. P2 can be set anywhere from a dead short to 10 K, 2-3 K seems best for positive charging in loop. To effectively monitor the battery, you have to be able to measure 1-mv changes in the system. Without that, you are in the dark. P-2 does not seem to effect motor speed except when connected back to battery Positive then magic happens! Its really simple, the devil is in the details. All diodes are 1N4007's single transistor, what more can I say. I'm going to watch American Idol now.....I'm gone. Enjoy Ben