Video of Bedini SSG-esqe alternate magent arrangement rotor

[min2oly]

Hi Patrick,

The switch is in the wrong place, when does the primary cap ever discharge the switch is only between it and the battery, the primary cap needs to vent its energy out to another low voltage battery or be shorted or what not. So the cap dump switch needs to be between the + and - of the primary cap not where it is now. It won't work with the set-up from the diagram. I'll be curious to know how it works with a battery instead of pick-up coil charging the cap, I suppose that could be a problem, you will be shorting the bat at the same time as the cap but I think you will see the effect if you move the switch.

If you take a look at the two vids I posted on page 4 Radiant from Cap 1 and 2, they detail my set-up. Good luck.

I'm still lost, I look forward to the vid or even some scribble/doodle on some paper.
Thanks for sharing your ideas.
KR - Patrick

[ZPDM]

Hi Patrick,

Here is a video detailing the radiant from a cap using batteries instead of a pick-up coil to charge the cap. The audio occasionally goes out, don't know why, maybe my thumb.

It is pretty straightforward so I haven't bothered to try and diagram anything at this point. I have some more video which if I ever get around to parsing it down and uploading it may try to share, but to not leave things hanging from the end of the first video the other point made which would be made in subsequent video is that if you discharge the cap into a second cap (with the second cap connected to a battery) you still preserve some of the appearance of the radiant.

To take this to a further level one would want to have a pulse train (555/arduino/etc) discharging the cap. Like many things this looked easy until I tried it. My first attempt was a relay driver circuit. However, any relay has an inductor in it. There is then a spike off the collpasing inductor when a relay switchs, Peter Lindemann I believe did some work with this with another person on the Energetic forum. Either the radiant from the relay coil is much larger than that from the cap or somehow the two are interacting but I kept using smaller and smaller caps until I found the best result with no cap at all. So scratch that one. I then tried to do the same with transitors. Again I ran into problems, The first attempt was just to short a cap (and the charging battery) through a transistor, same as I was doing by hand in the video. For some reason and I have no idea why the radiant disappears when I do this. However, if I charge a cap, disconnect the battery, then discharge the cap through a transistor, the radiant is back same as before. So it looks like what needs to be done is a transistor circuit which first breaks the connection to the charging battery then makes a connection to discharge the cap, i.e a transistor equivalent of a SPDT relay.

I am having trouble doing this and appeal to the board for help. This board has an incredible number of smart people on it, I still recall in trying to build my mag ramp motor how I spent two weeks trying to figure out how such a circuit would work finally asked and the next day someone gave me the Bedini Bipolar communtator circuit. I have been reading and trying to solve this myself for a couple days and I think I either need some combination of a NPN/PNP or I need to build a NOT gate inverter, however, I feel certain very many people on this board could at least nudge me in the right direction if not just show me a circuit. Again the idea is to charge a cap then disconnect the charge and discharge the cap so a transistor based SPDT where in the off state switch one is connected (charging cap) switch two is disonnected(cap not discharging), in the on state switch one is disconnected(cap not charging) switch two is connected(cap discharging).

So that's it, I think you should see this effect now Patrick, at least at lower volts it is probably pretty subtle as compared to the radiant off a collpasing inductor but it is there. Any help/guidance from the many people on this board that know much more about building circuits than I do would be most appreciated.

[Tom C]

I kinda understand what you want...... how about a block diagram. although it seems like a tesla switch variant would do what you want. a pair of mjl21194 driven by an opto on either channel of a 3524.... Niteyesh should probably chime in here.

Tom C

[ZPDM]

Hi Tom,

Excluding the rest of the circuit, here is a crude block diagram. I haven't had a go at the Tesla switch, but yea if I understand correctly the switching issue is the exact same, though likely more simple. I plan to drive this with an auduino so I can look at PWM, but I don't know how to make the pulse from the arduino turn one transistor on and another off at the same time. Thought this would be easy and a basic circuit but from poking around on the web maybe not so much. However, obviously if the switching issue is the same as in the Tesla switch ... well I'm going to look at what Bedini did there and see if I can understand it, but this may be even a simpler circuit.

[ZPDM]

Just a quick update, I've found a work around. If I discharge the cap not through one transistor but through a darlington pair it behaves similarly to if I short the cap by hand, i.e. the radiant is back. I'm still shorting the battery as well, but I should be able to pulse this with arduino and perhaps I can get away with the battery shorting. Will have to see. The other issue might still be solved to improve this but I am good to go, thx, Ciao. Paul

[Tom C]

p.m. Bobzilla he is really good with the arduino stuff... he should be able to do a short script for you fairly easily. what board do you have ?



Tom C

[ZPDM]

Hey Tom,

Just been experimenting for the past couple hours. Will try to get out a video in the next few days which often generally translates to next few weeks. I have the Arduino Uno board and did have success with pulse width modulation, just took the "blink' example and plugged in new delay times. However, that is good to know that Bobzilla has worked with Arduino because I realized at one point today, the arduino board has more then one output, you could just use it to solve the problem of how do you use one "clock tick" to make one connection and break another though it would require a bit more programming. However, just to be sure here, and I suspect I sometimes go between electronics 101 and electronics 404 questions w/o being certain myself which I am asking, but that is what you are doing with transistors in the Tesla switch right? You are making the connections to charge in parallel, breaking those, then making the connections to discharge in series?

Have had some success with the arduino pulse width modulation approach. A limiting reagent appears to be the transistors, had no success with a 3055. Did have success with two 3055s in Darlington pair. Also occurred to me along the way to put a 1 kOhm resistor between the charging batteries and capacitor, i.e. stops the batteries from shorting when the cap discharges. Fiddling with pulse rates/duty cycle/resistor value about the best I've seen so far from an 18V input is 8-8.5V radiant output with cap shorting to itself. With cap shorting to a 1.5V battery, radiant down to about 3-4 volts. Amperage isn't great from radiant but not bad either, enough to charge another 1.5V over I don't know maybe two, three hours. Did run a set-up like this for about half an hour.

Tried a MOSFET, on paper it looks great, fast switching time, can transfer lots of current, however, and to be honest this was the second time I tried and first time I got a MOSFET to work in a circuit, it wasn't very good maybe 0.5V radiant off 18V input. I don't know what the problem is, I am going to pick up a second one from Radio shack tomorrow and try a Mosfet Darlington Pair, heck it worked with BJTs. Not fully optimistic about it, I think somehow, even though the turn on voltage is supposed to be 2-4V I am not getting full saturation. So if the "Darlington MOSFET" doesn't work will try other Mosfets and see if I get better results. I don't know on paper MOSFETS look like they might improve things further, but so far no success and maybe it is no dice unknown reason.

Still, with the BJT Darlington, will probably set-up a spreadsheet to look at voltages, etc, then look to figure in the switching costs from the Arduino and go from thence, etc. etch and so on and so forth. This thing is very like the Tesla Switch with the differences that a) The abruptness of the discharge is intended to be determined by the cap rather then a fast electronic switch, though as noted transistors seem to be a limiting reagent for me b) the radiant isn't entirely reverberated around the circuit but shunted off elsewhere. Thought I had a c) but it slipped my mind, couldn't have been important, but really for the briefest fraction of a second that discharging cap is putting out loads of watts, anything that slows that discharge slows the change in V, which diminishes the appearance of R (at least as I see it) so, as I said I think this thing may be decent as it is, any help with switching, i,e MOSFETS and it may really sing.

edit: remembered c) would be you "rotate" at least some of the batteries, by hand, every 20 minutes or so, not automatically every cycle as with the Tesla Switch.

[ZPDM]

Am making some progress. As noted I have a circuit that demostrates this approach and am now waiting on parts. Likely won't make a video of this first circuit unless anyone really wants to see it as it is more a gamma than a beta version, the circuit doesn't plain stink but it is like three day old socks that need darning it smells bad and is full of holes. The goal is simple, use a battery to charge a thin film cap, shut off the supply from the battery, disharge cap, repeat. My first attempt involved an opto which directly switched the power supply to the cap, then a separate signal from arduino to discharge the cap through a darlington. The two (arduino signalling and cap discharge sides) weren't fully isolated, the transistor side of the opto I believe is not intended to handle any appreciable current, (at one point when I when to 18V I elected a new Pope, i.e white puff of smoke as opto failed) and the darlington is likely not able to switch very fast. I think what I need to do is use the opto to switch a power Mosfet. It is incredible what you can find on the internet and entering even something as obscure as "opto driving Power Mosfet" will often bring up the circuits, dicussion about, a video or two on it and maybe an instructable page.

My first attempts at using a Mosfet were not successful in this area but I don't think I had it set-up right and I think I have an idea how to do this now and am 90 -99.9% certain it is the best approach. If I am not mistaken this approach is what John Bedini uses in his solid state Tesla switch, if it works with discharging high voltage batteries into lower voltage batteries quickly I really think it will work with capacitive discharge as I saw with the Darlington. As an aside, I put off watching Part 2 of the Tesla Switch series when I first got involved in this (from ideas generated by part I) because I didn't want to get pulled off in yet another direction by yet some other odd and incredible thing John might mention. Now I want to watch it to perhaps learn more about switching and can't because I have this Lubuntu laptop It may just drive me back to Windows. So I went to radioshack to pick up some Mosfets and they are all closed what with Radioshack going out of business. Ordered some 4n35 optos and two different kinds of power Mosfets from Digikey, used to order from Mouser but then they put "Cheech and Chong" from "Myth Preservers" on their splash page and now want nothing to do with them. Was about to go to sleep when I wondered has anyone made a discrete component that combines an opto with a Mosfet, yes they have, so I ordered a few FOD3180 opto Gate drivers https://www.fairchildsemi.com/datasheets/FO/FOD3180.pdf I may still need to go from there to a Power Mosfet but don't know. If I can get the radiant from the cap right from the 3180 that will be great.

I will say if I had thought I was going to have some measure of success in this area I likely wouldn't have introduced myself to John and the board as the arrogant Jackass Court Jester guy, well I was drunk, so I likely would have found some way to be a Jackass, but perhaps braying differently. If you read this John, you needn't bother to reply, but I am sorry about my foolish introduction.

As I am waiting for parts and again have had a few beers I will regal the board with theory and math before closing.

The first part is this, whenever there is a change in voltage there is a change or appearance of radiant energy, I would guess simultaneous phenomena. This is common knowledge to experimenters here and I know this solely because John showed that this negative or radiant energy can be rectified with a diode. I know of no other "sensor" for it though to imagine its magnitude in high voltage spark plugs or for that matter lightening strikes is something. As the radiant can be rectified and put in a cap with a diode we can consider the "power" from the radiant and what parameters affect this. The following is somewhat of a guess, I have done nothing formally, though it should be done, it is a guess based from looking at collapsing inductors and tap discharging caps by hand, and a few days with a poorly designed circuit. Regarding the parameters that affect the radiant.

Three parameters affect the magnitude of the radiant, four if you count pulses per unit time. 1) abrupness of discharge 2) current 3) voltage and 4) pulse per unit time.

1) There is a linear, 1:1, relationship between the abruptness of discharge and the magnitude of the radiant. I.e if you have 50V -> to 0 and it does so in ten milliesconds of time you get say 1 unit of radiant, if 50V goes to 0 in 5 msec you get 2 units of radiant. This part might follow a square law I don't know, I am guessing linear, to test it put resistors of various values off a collapsing inductor or discharging cap and measure the power into another cap.

2) Current: Current also follows a 1:1 linear relationship, I go back to the huge currents coming online in Tesla's Niagra Falls thick power lines which led to radiants leaping out of the line, I don't know the voltage of those lines, but current definately plays some role. It may follow an inverse square law, but it does something and in my brief work with cap discharge I have to guess 1:1

3) Voltage: As John Wayne said to the space alien, "I don't know what planet you're from partner but round here we like to follow square laws and inverse square laws". Terrible joke perhaps, but no denying a lot of phenomena around these parts follow square laws. The magnitude/power of the radiant follows a square law dependent on voltage. So lets look at some possible examples of this. Lets say you have 1 watt of power from 1V at 1 amp, it always discharges in exactly the same amount of time and the discharge of this 1 volt yields 1 unit of radiant. If you increase the voltage to 2V and remain at 1 amp you now have (2*2) 4 x 1 amp = 4 units of radiant for 2 watts of input. Three volts then 3*3 =9 times 1 amp = 9 units of radiant for 3 watts input, 4:16, 5:25 etc.

Lets now say you keep the power input the exact same yet double the voltage, so to stay at 1W you now have 2V at 1/2 amp so (2*2) 4* 1/2 amp = 2 units of radiant. The input power has remained exactly the same yet the power from the radiant has doubled. Likewise if you retain 1W and increase voltage to 4 V you have (4*4) = 16*1/4 amp = 4 units of radiant from one watt power input. The above is what I think is going on.

4) Pulse rate, not much to say if the voltage change happens more frequently, more radiant per unit time, i.e more power. Concerning power in versus radiant out, there would be more to that with inductors, less so if you are dealing with capacitors, where you could just fill it then turn off the power supply and discharge.

That's it, I feel pretty darn certain about 3. I think I am right about 2 though it may be an inverse square law, might be right about 1, though it may be a square law. We know how to rectify the radiant with a Bedini diode to a cap, these suppositions can be expermentally verified or corrected.

[ZPDM]

Been from Azerbaijan to Zimbabwe with capacitive discharge radiant and other things. First then, I'm not real good with circuits whether that was the issue or not I don't know but trying half a dozen ways to Sunday I couldn't get more than 3-5% power out from the radiant capacitive as compared to the input power to the cap, still not bad, but I feel like it wasn't really what I was seeing when I was tap discharging the cap by hand. So all what I was saying in the previous post maybe a pile of it, maybe not I really don't know. Then I figured what have I had fun with, so I built a solid state Arduino Bedini (pulsed coil) radiant charger. About the only wrinkle was I powered it at six volts but with three six volt batteries utilizing the TS common ground set-up, i.e. 12V to 6V common ground to load. This was really worthwhile and taught me a lot being able to so easily modify pulse width and frequency, an Arduino costs about ten bucks. Measuring power into a cap it was only maybe 60-80% recovery. However, when I took two new 6V sitting at 6.34 and charged 3 dead 6V SLABs all under 4 volts (one on the TS side, two being radiantly charged) they all charged up to 6.05 - 6.1 volts seemingly OU, i.e. the new bats were maybe at 6.2 now. After this it all fell apart the charge batteries wouldn't budge. I got frustrated so I disconnected everything and first hit them with a heck of a conventional charge then from a DC power supply an absurd radiant charge. I kept my fingers on them to see if they were heating and they weren't then one squeaked and I remembered something about cold boiling, I'm a lucky idiot. But short story they still settled 6.1 ->6.15. So having set for years at under 4V thats all these batteries could do. Pretty interesting, I thought to really formally repeat it all but figured heck it will take a week or few and I want to move on.

So I looked at optocouplers, and finally figured out how to trigger a MOSFET from an opto, also realized I want MOSFETS with more than 60V breakdown between source and drain as otherwise that portion of a radiant spike is likely lost. Figured out you do not! want to use a BR with a Bedini SSG type set-up, no idea why, but compare it with single diode. Was able to back-pop a cap from MOSFETs to the source. Didn't run it long with batteries, with supercaps was getting about 60-70% maybe 80 energy recovery. First attempt with backpopping using BJTs failed but in retrospect I think this is because really just for space on the board reasons I like to run the load on BJTs on the low side (i.e. emitter to ground) the only way I could make the MOSFETs go was running the load high side (VCC -> collector) so I suspect if I set-up the BJTs that way I might also have success. MOSFETs are quirky as heck to me, but I have a set-up that seems to work.

So now at long last I can return to the magnetic ramp motor. I will make a few design changes then for this Mark One version. First I will make it all N facing out on all ramps and I will only have two ramps each partially covering 180 degrees. You could make it N then S as I had in my initial attempt. They are reasons to want to do this, the bigger would be if you were using a C shaped coil you could bring both sides of an electromagnet to bear. As I'm not doing that at the moment and so many things go different in experiment then on paper I will avoid the added complexity of hall effects and the Bedini Bipolar commutator circuit at this time. Second, I'm darn sure now I need a separate timing wheel. if you try to time it off the ramp you are going to almost certainly oversaturate the coil. I am going to leave some space between the two ramps, at least initially as 1) I have no idea what effect magnets have on the radiant and don't want the coil discharge going into a magnetic field 2) as it will be monopole eventually you don't want the coil repulsing away from one ramp and repulsing (more weakly) away from the oncoming ramp.

The easiest way to visualize this is just a Bedini SSG with racing stripes, however the sensor coil is removed from the drive coil to provide a pulse only at the sticking point of the mag ramp. It also looks a heck of a lot like a Lindemann attraction motor again except for racing stripes, finally aside from the fact that you capture the inductive spike and the movable rotor is where the ramp is, it looks like the Wankle magnetic motor. Now that I have at least scratched the surface of how the radiant spike behaves, a big question is how well does the magnetic ramp play with a pick-up coil or coils.

[ZPDM]

Been making some slow steady progress almost certainly nothing worth a video update, but wanted to mention some to me strange and notable things before moving on. If you have a magnetic ramp against an iron core, say N N N N, you get good pull, if you go N S N S, again against an iron core, you get almost no pull, in fact it will sometimes latch between the first NS. That makes no conventional sense. You could say that it is something with the induced field in the electromagnet, but it comes to a dead stop in the middle of the ramp, there is nothing going on then in the coil. You will say, which is true, that the magnetic fields have now changed, but what does the iron core care, both N and S attract it right? I don't know what to make or do with it but an iron (this isn't magnetic shielding of another magnet) core, is not attracted by alternating NS even at a dead stop.

I tried angling the, all N facing out, magnets to try and get some Howard Johnson stuff going on. It may have been a bit better, certainly could bring the electromagnet closer to the rotor, but probably close to the same as having the magnets flush outward. Did notice something quite interesting though, the previous 50-90 volt radiant from the collapsing drive coil now set right at 10-12V. Don't know if power output was less, most likely, but the rotating magnets certainly play a large role in any Bedini SSG set-up.

Currently running the blessed thing in attraction mode as this is the set-up where I saw RPMs increase with a load attached to spiked power coil. I may want to reverse that but am happy to continue on this line further and suspect it may be the way to go. Actually, as a monopole ramp was built the power from the collapsing coil decreased, but so did the power needed to drive the machine at a set rpm and they both decreased proportionally.

So with say a 120 mA input I've gone from maybe 5 to 15 mA from the pick-up coil with a 4-5 magnet ramp, with, if anything, decreasing input. After 5 magnets things get worse but I am pretty certain this involves the reed switch timing when I am in attraction mode. I.e I looked at this months ago in repulsion mode and basically saw improvements until the magnets couldn't see the core 10-15 magnets. So there are trade offs.

One maybe interesting question, when the drive coil collapses on an SSG in attraction mode does it push off the magnet? I.e. you are getting energy out of opposite polarity, does the magnet see only a decreasing electromagnet, or does it see a reversal?

The idea is to paint the perimeter of the rotor with as many magnets as close to the perimeter as possible without losing rpms. It may be as with the Wankle the best way to do this is a very, very slight slope to the mag ramp. Alternately, while only pulsing one ramp, one could have two or three separate ramps for each pulse. One thing I guess, but don't have the background to know, it is the rotating field cutting 180 degrees that produces power in the pick-up coil correct, not the change in flux when the magnet rotates away? That is to say if you had magnets placed perfectly entirely around the perimeter of a rotor and spun it, would it produce any power in a pick-up coil, or does the flux need to change in the direction facing the center of the rotor?

Anyhow, I'm hoping to eventually set-up one ramp N one ramp S, as I could bring a C coil to bear without, I now realize, any need for a Bedini Bipolar commutator circuit, still a ways off. If I am correct, this overall approach should work, basically because it is derivative of two approaches, Bedini SSG and Wankle, that work. Whether I can get it to work outside of a Bedini/Menlo Park type place is another question, but if I start to see what's important the playing field, outside the intellect part, is being greatly leveled by 3d printing and the like.