Donald Smith Devices too good to be true

[ricards]

Excellent Ricards !! Now consider the possibility of "Never Discharging" the energy in the caps by altering the balance and allowing a natural re-balance. This does amount to a capacitive voltage divider but the current increase is extraordinarily impressive.

https://www.youtube.com/watch?v=XTVbTQnRmKM

Take note of the watt meter - these are 50 watt bulbs in parallel moving an average of 15 amps through them. The primary storage caps are never discharged in this demonstration.

1/2 of the original energy comes from the battery, the other half comes from the cap storage - all the energy is then stored in the opposing cap. The re-balancing is free which restores the original energy in the both caps. The loss is the 1/2 used from the battery to initiate an imbalance. Approximately 1/4 of the total energy exerted on the output is depleted from the battery.

I didn't quite get that... haha can you explain further...
I was thinking If I don't deplete the capacitor's fully to charge it from 0, the charge from that cap would limit the current from my battery, this will make the output in between caps low as well..

The best output I got was to deplete the charged capacitor's fully.. so I can charge it from 0, this will give greater output In between capacitors, and to discharge it rapidly (which is already the nature of the capacitors..), just need to make sure my transformer's primary impedance/resistance is very low.

I can think of two ways to do this.. Either Increase my source voltage or discharge through a spark (negative resistance), or both.. like tesla did.. (now I know why he did what he did).

The shorter the Charge/Discharge time (Impulse) the greater the output in both sides of the capacitor.

I think its too early to conclude what is going on in between capacitor's is by environment input. but it is also hard to say that it is because of the source.. well it was influenced by it, But I literally did not PUT a charge on that side.. yet there is something there that can power loads.. you get what I mean?..

[ricards]

Throwing the Idea Out

after finally understanding Electrical resonance...

I've re-read Don's PDF again and realized some things. in the section "Air Core Induction Builders Guide".

Don smith was very specific about the construction of L2 Coil (Output). to the point that it has to be Full, 1/2, 1/4 Wavelength, not specifying why... he then throws.. you have to divide by 247 for 1/4 wavelength, by the frequency you chose.. etc... and then match this and match that.. its an incomplete guide.. but yes there are hints...

I will literally throw my Idea Here..

First the Idea why don was suggesting to divide chosen frequency by 247 for 1/4 wavelength to get wave length was because EM wave travel at the speed of light if it travels along the wire and "Stationary Waves" occur the Idea is to Tap at those Node and Anti-Node Points.
In His example 247 / 24.7 Mhz is 10 ft..
The 1/4 Wavelength of EM wave traveling at speed of light (983,317,760
ft/s) is 9.95 ft... pretty close!

The Primary Coil was suggested to be 10 turns again for no reason.. If we are trying to make a "Resonance Energy Device" oscillate at "Resonance" we should also consider this... why?... in order to get a Standard value Capacitor... its more practical to vary the Inductance by adding spacer on coils rather than trouble ourselves on variable capacitors on nf-uf values..

Suggested Power Input was Neon sign Transformers.. In the Videos it was stated to be 30khz again... why?.. aren't we supposed to pulse at 24.7 Mhz?
is there a power supply like that?.. maybe a signal generator.. to pulse a FET?.. I remember seeing a Replication of "salty citrus" without spark gaps, but with electronic circuitry..
I never saw the table top device run in any of the videos... maybe I missed it.. or maybe... It was never meant to run... (speculations )..
There is a phenomenon called "Harmonic" if you double the frequency you will have more overtones and the wavelength of Standing waves decreases..
Let us try that in Reverse.. let us decrease the frequency and keep dividing it by 2... in an Excel Spreadsheet I've included here one lower "Harmonic" is 24 khz which is close to 30 khz.. you can actually Engineer the whole design so that the Coils are to match the capacitors (at standard Values) and still use the 30khz neon sign, or you can do the other way around, build a resonant coil L1 and L2 and have a variable frequency input.. I've Included in the spreadsheet attached my calculations..
anyone can use and Check if any of my calculations are wrong ...
so basically to cut the long story short... we have to pulse at lower Harmonic Frequency BUT we have to make sure our coils and capacitors both L1C1 tank and L2C2 tank are designed to oscillate at this lower frequency..

so to reiterate things..
Design L2 Coil by EM Wavelength to Desired Frequency at MHZ range, use speed of light..
Find Lower Harmonic frequency of that Mhz at Khz Range.
Design your L2 Coil so that you will arrive at Standard Valued Capacitor..
Design your L1 Coil also to have a Standard Value Caps, but make sure Its in a Step-up configuration..
L1C1 Tank and L2C2 Tank must resonate at the same frequency.
Pulse L1 at this Specific Frequency.

Attached is an Excel Calculator to help with the overall design...

well.. that's it!.. my wild Idea of Don's Tabletop Device...
all thats left is to build and see If I'm learning or Hallucinating things..

[tswift]

well.. that's it!.. my wild Idea of Don's Tabletop Device...
all thats left is to build and see If I'm learning or Hallucinating things..

Yes, building and trying is pretty much the only path to progress. It's possible to formulate any number of interesting-sounding theories only to discover that they don't work in practice. I think you're understanding the resonance in the Don Smith device, configurations like the tabletop device are essentially dual-resonant Tesla coils. In order to get a much smaller size than a traditional Tesla coil, Don uses a cap on L2 because its natural 1/4 wave resonance is much higher. With a configuration like this, when L1 is spark excited you can actually see both the LC resonance (at perhaps a few hundred kilohertz) and the 1/4 wave resonance (at perhaps a few megahertz). Just putting the scope probe within a few inches of L2 is sufficient due to capacitive coupling through the air. If you do the arithmetic on the tabletop device pictures, using the presented cap values and making a reasonable estimate for the inductance of L2, you arrive at a resonance of something like 200 KHz. I did it once but now I don't remember the exact number. The drive frequency of the NST/HVM really doesn't seem to matter, it's just a spark excited Tesla coil.

However, as can be seen from some of Don's other devices, it's not the spark excitation itself that has anything important to do with the radiant effect. It's the oscillating dielectric field around the dipole (L2) that makes things around it radiant as they try to restore ambient equilibrium. So with that in mind I've been designing and building a better configuration that is a whole lot quieter than my spark-excited Tesla coil. It's a true SSTC suitable for power levels up to 100W or so. The drive circuit is an SG3525 IC powering the gates of two 13N06 MOSFETs at 320 KHz. The drive IC is powered from a separate supply than the +V going to the MOSFETs. I have a new better bench supply that can go up to 32V and 5A, you can see it in the pictures. Current draw is modest if the tuning is good, the MOSFET heatsink gets warm but not unreasonably hot. I have experimented with both antenna feedback and current transformer feedback and so far haven't gotten good results for self-tuning. Right now I just manually adjust the resistor on the SG3525 for resonance. This is the same chip used in the PVM12 unit, the drive circuit is very nearly the same. The chip can only go up to 400 KHz, so I wound a new L2 secondary with a lower resonant frequency for this application. For SSTC design you usually want a high coupling ratio between primary and secondary, so a shorter wider coil becomes desirable. At these power levels, operating in CW mode, the coil doesn't throw off sparks. A tiny bit of corona is evident at the tip of the secondary if you view it in complete darkness.

The output circuit is Don's, essentially the same as the plasma ball device. A two-turn current pickup loop feeds a diode bridge (four 1N5822 Schottky fast diodes), which charges the 12V 83F supercap bank. The voltage on the whole pickup/output assembly is floating, it is not grounded and thus it doesn't affect coil tuning very much. This does make taking measurements a bit hard since you can't just attach a scope probe or voltmeter to it while operating, I measured almost 3KV p-p voltage on the whole assembly using my HV scope probe. Thankfully I have some of these cheap analog panel meters kicking around and they work just fine if you don't need high accuracy. This arrangement worked better than I expected and is fairly efficient at delivering charging current to the supercaps. It takes about 18V of drive input to the MOSFET half-bridge to reach 12V on the supercaps. I also scavenged and reused my previously built shunt regulator (basically just a beefier zener diode) and attached it as a way to keep the cap bank from overcharging. This also provides a direct way to measure the charging current, because once the cap bank is full and the shunt regulator kicks in you can measure the current through it, as seen on the ammeter. With 28V drive, the charging current is right about half an amp. If I up the voltage to 32V (max for this P/S) I can get almost an amp.

The remaining details to finish the circuit are the design and construction of the step-down converter. I plan on building another SG3525-based driven inverter stage that will feed a 12V-to-12V isolation transformer. This transformer will have to have very good insulation between primary and secondary since it could see 3KV or more of voltage difference. This is also where the power gain should occur, since the charge in the cap bank is radiant. When this is used to drive the step-down inverter stage, it should run cold and show a power gain due to Lenz's law reduction. Then all that's necessary is use a beefy full bridge of high-speed diodes to get 12V at high current which can then be used to feed any off-the-shelf inverter for grid-compatible power. The 12V drive can be used to run the L1 side inverter, through a boost converter if necessary to get the 12V up to 24V or so depending on the current requirements of the step-down converter. So far in my experiments with SG3525-based designs I can get the no-load current requirements down to about 150 mA at 12V.

[med.3012]

Yes, building and trying is pretty much the only path to progress. It's possible to formulate any number of interesting-sounding theories only to discover that they don't work in practice.

did you tried all the interesting-sounding theories? so you are sure they don't work !!

[ricards]

so far so good,

I have built a Crude Coils and some standard capacitor and used my variable frequency oscillator.. run things at 12 Volts input pulsed DC.

at loose Transformer coupling If Pulsed at a Non-Resonant Frequency I get NOTHING! ..

BUT upon tuning the Frequency of my oscillator. I'm getting 0.2 V. (still CRAP) but hey think on the bright side.. I'm getting something because of resonance in comparison to non-resonant circuits in which I get nothing.
(getting 2.5V without the current limiting resistor)

I had been experimenting with High voltage Pulses and had the Impression that even at very loose Transformer Coupling one could get output from the secondary imagine if its in resonance..

I'm trying to stop myself from arriving at conclusions as I want to see things first hand, but I think Its worth to post this to whomever still experimenting at Don Smith related stuff or resonance related stuff...

so far so good, now what I would like to find out what will happen at my output If I double (or triple) my Input Voltage... is it going to be like what don smith claimed... squaring of flux line..

@tswift,
based on my experiments I think The Drive Frequency of the NST matters A LOT.

in resonant circuit pulsing I notice the Input amperage to be drastically reduced.

[tswift]

Don Smith build complete... but no power.

at loose Transformer coupling If Pulsed at a Non-Resonant Frequency I get NOTHING! ..

Yes, you've noticed one of the biggest problems with Don-style circuits, which is the power transfer. It's hard to get enough watts throughput to run anything useful at the secondary. Basically there are two choices: loose coupling and resonance, or tighter coupling. Tighter coupling gets more problematic at higher and higher voltages. I have been fighting this problem for years now, trying to build a configuration with enough power to run an inverter circuit for the step-down/isolation transformer. Now I have it.

This circuit is essentially the same as in Don's plasma ball device, where a low-turn induction coil charges a cap bank through a diode bridge. This is the non-resonant but tight coupling method. To get enough power to run anything useful, I built a solid-state Tesla coil to drive the circuit instead of the plasma globe. This really helped, now that I can ramp the input up to almost 100 watts it can actually run something at the output.

I have finally completed the rest of the circuit. The step-down converter that feeds the isolation transformer is another circuit basically identical to the SSTC drive circuit. It's an SG3525 chip driving two 13N06 MOSFET's at about 40 KHz. The isolation transformer itself is a 3" powdered iron toroid with a 44 turn bifilar primary (of 22 gauge zip cord speaker wire) and a 48 turn secondary (of 16 gauge zip cord speaker wire). There are two layers of yellow dielectric tape between the primary and secondary to provide several kilovolts worth of isolation capability.

The secondary feeds a 400A schottky diode half-bridge array (two diodes back to back in one package) that's mounted on a beefy heat sink. Theoretically, when the power shows up it could be necessary to handle this kind of current and dissipation, so I designed everything to handle it.

However, testing is showing no measurable power gain. The circuit itself works pretty well: the SSTC resonates strongly, it will light up a fluorescent tube anywhere close. Power levels are just below where the top of the coil will start breaking out a spark. Running in CW mode like this is totally silent and the MOSFET heat sink gets warm but not excessively hot. Likewise, the radiant capture part works pretty well. The supercap bank charges up nicely even from fully discharged (although the coil tuning shifts slightly as the bank charges). When charged to about 12.8V, the pickup coil can deliver about 1A of current in the cap bank. This is plenty to run the step-down inverter stage (kind of a misnomer since it's really a 12V to 12V converter), which draws about 150 mA with no load. There is also a shunt regulator circuit which limits the cap bank voltage by just dissipating excess current as necessary, that's the 3055 transistor with a heat sink.

If I load the output from the isolation transformer with a 16 ohm power resistor (that aluminum rectangle is a 1000W 16 ohm resistor), it draws just under 1A of current (as you would expect at 12V nominal voltage). The discouraging part is that the Lenz effect on the isolation transformer primary can be clearly seen on the ammeter. It takes essentially 1A in to get 1A out. The conversion efficiency of the step-down inverter is pretty good: the MOSFET's don't even need a heat sink, nor do they get perceptibly warm. Transformer core losses are very low, I deliberately used about 3 times the normal number of turns for this core in order to limit flux density and keep hysteresis losses very low.

Loading the output even further confirms this result. With a 1 ohm resistor the voltage at the secondary collapses to about 2V, while the current on the ammeter climbs to about 2A.

Conclusion? Obviously, something still isn't right somewhere! This is what I meant before about it being easy to formulate nice-sounding theories that don't work out in practice. Ultimately only experimentation can prove what is real and what is not. This is the best Don Smith-style configuration I have ever built, and if my understanding was complete and correct, it should have worked and shown overunity gain in the isolation transformer. With this size unit, it should be possible to loop the 12V output from the transformer back to run the input, and at the same time attach an inverter of 1000W or more, thus the need for that insanely beefy rectifier. However, even without more detailed measurements it's quite obvious no magic is showing up.

I do think that even with this negative result that I am still on the right track. I think that whatever is lacking has to be some small detail that doesn't seem important, and Don omitted it either accidentally or intentionally. Even people like Zilano who claimed to have a working unit, might have done something unintentionally that makes it work and thus didn't even think to mention it. So far the leading candidates for additional investigation are the grounding (two grounds? one ground? tunable ground somehow?) and the winding direction of the transformer. Those are the only good ideas I have for now as to what other steps to take. Given how much work I put into building this configuration, I intend to leave it assembled on the workbench and continue to tweak it to see what else I can come up with.

[bistander]

... Transformer core losses are very low, I deliberately used about 3 times the normal number of turns for this core in order to limit flux density and keep hysteresis losses very low.
...

Hi tswift,

I'm curious why you think the number of turns limits flux density in the core. Can you explain that?

Thanks,

bi

[tswift]

I'm curious why you think the number of turns limits flux density in the core. Can you explain that?

Sure, it's standard electrical engineering, volt-seconds per turn directly relates to the peak flux density in the core, which in turn directly relates to core losses. See this page for details:

Micrometals - Iron Powder Cores

In this case I have a Micrometals T300-52 core, so A is 1.68 cm^2. Using the formula for square wave waveform, this is 12V (average AC voltage per half-cycle) divided by 4 * A * N * F where A is 1.68 cm^2 (or .000168 m^2 for the formula units) and N is the 44 turns of the primary and F is 40,000 Hz. So this gives 0.01T or 100 gauss. Looking at the nomograph of core loss for the 52 material at 100 gauss and 40 Khz (ref. Micrometals - Iron Powder Cores) gives about 10 mW per cubic centimeter of core material volume. The T300 core has 33.4 cm^3 volume, so this gives total core loss of about 330 mW.

You can do the arithmetic again with only 12 turns and see the result: the core loss is much higher. Depending on the design this might be perfectly acceptable but I wanted very low idle current requirements so this necessitated more turns to get the core loss down. In practice I didn't do all this to begin with, I just wound 12 turns on it and noticed that it worked but the idle current was more than I wanted. Then I wound some of the 22 gauge wire I had on hand and one layer made 44 turns. I hooked it up and it was much better so I went with that.

[dragon]

Start with nothing...

If you start with nothing you can only go up from there... right? Or, I suppose, you could end up with nothing.... either way you have nothing to loose and maybe something to gain.

Just for fun, look at the 3 diagrams... The only power on those lines comes from a charge separation. Between the antenna and ground you have a basic "make n break". Fig 2 a coil is added, Fig 3 a tuned coil is added. The only power required is that needed to operate the "make n break" which can come from the movement of energy once its operational.

Can you figure out a way to use a charged cap in conjunction with the make n break without discharging the cap?

Do you suppose you could do something similar with a current source.... like maybe 2 earth grounds with a make n break? yea, probably wouldn't work...

Maybe with a little creative thought it might be possible to store some of the energy and reuse it to enhance the outcome... hhmmmm... maybe a charge pump of some sort...

A Challenge for all those interested... use the diagrams to create a simple circuit that can light a single LED... you can start with a simple manual make n break circuit - learn and understand it - then work your skills to a self sustaining LED circuit. Most of this can be done without spending any money using coils collecting dust on the shelf so the only investment is time along with some creative thinking.

[ricards]

If you start with nothing you can only go up from there... right? Or, I suppose, you could end up with nothing.... either way you have nothing to loose and maybe something to gain.

Just for fun, look at the 3 diagrams... The only power on those lines comes from a charge separation. Between the antenna and ground you have a basic "make n break". Fig 2 a coil is added, Fig 3 a tuned coil is added. The only power required is that needed to operate the "make n break" which can come from the movement of energy once its operational.

Can you figure out a way to use a charged cap in conjunction with the make n break without discharging the cap?

Do you suppose you could do something similar with a current source.... like maybe 2 earth grounds with a make n break? yea, probably wouldn't work...

Maybe with a little creative thought it might be possible to store some of the energy and reuse it to enhance the outcome... hhmmmm... maybe a charge pump of some sort... nah, probably not...

A Challenge for all those interested... use the diagrams to create a simple circuit that can light a single LED... you can start with a simple manual make n break circuit - learn and understand it - then work your skills to a self sustaining LED circuit. Most of this can be done without spending any money using coils collecting dust on the shelf so the only investment is time along with some creative thinking.

I like the Idea of going back to the basics, and do some simple testing and observing.. try and achieve simple but meaningful goals.. I'm interested and I could picture something in my mind that I might do later....

Well.. it will be a while to redo my setup. with regards with don' related stuff.
it all popped and smoked.
ahh resonant rise...
the good thing though is I make all my circuitry from scratch, so I wouldn't go **** when they go like that.

There are some take away to that popping and smoking though.. I could get some good current flow from a Loose Coupling...

yes really something is not right, from my own testing, I could not get an accurate Voltage Reading on the output, I would have higher Voltage reading when I ground the common lead (Black) rather than measuring the two output leads.. my step-up transformer configuration somehow is acting to a step down.. (maybe because of the very loose coupling).. and when If I double my input voltage, I should be expecting double in current as well, but as my testing shows,
@ 12v-1.9A
@ 24v-2.0A
@ 36v-2.1A
@ 48v- popped

oh yeah I forgot to attach my spreadsheet calc, if anyone is interested... hope you can comment as well.. if you see errors..
Resonance_Calc.zip

[ricards]

A Circuit

Can you figure out a way to use a charged cap in conjunction with the make n break without discharging the cap?

Hi dragon,

here is the circuit I was thinking,
Back_N_forth.PNG

I think this concept relates to some other topic as well and most probably has been suggested before by other people in this forum but was little considered, but am posting it here anyway.

should be easy to do with Transistors and Relays. to try and see what it does.

It will be hard to do resonance with this though, it will produce a weird waveform..

maybe I should put another transformer at SW3?..

[dragon]

Good thinking Ricards, although, it seems you may have recreated something closer to the bedini scalar charger or very similar to the 3 battery system that Dave and Matt are working with, which actually works quite well.

Here is a short version pdf which might help in the understanding of the slight variations in what I'm calling a charge pump. Very similar to the above mentioned but with some interesting changes that I found while working with similar arrangements. My concern isn't with the 15-20% losses in C1 as these can easily be replaced - for me personally it was more about the current generated in the process of moving energy through a load while limiting losses.

Although, combined in some of my research, the "start with nothing" is a separate system, one of many pieces to create the whole. If your using a forced system ( an oscillator circuit driven with a battery or other power source ) how would you know if the system is drawing from an outside source? How would you measure the difference if there was any? How could you measure success? Like looking for a black rock in a totally dark room... By combining efforts to preserve charge and drawing from an outside source to offset losses it becomes more likely we can come close to unity while driving a load. Unity in the sense that we are recycling energy as well as drawing from a small source.

For instance if you grossly limit input from a battery and focus on ways to increase output by harvesting from an outside source and/or recycling the energy you are more likely to succeed over simply pouring energy into it hoping to see something different. I've attached a schematic of a system I built many years ago that isolates and limits the battery. The center coil harvests energy to drive the system. I could never remove the battery completely but it was an excellent learning tool in finding ways to do more with less. It would drive 30 LED's in parallel with a draw on the battery of less than 1ma.

[ricards]

I tested the circuit out manually.. the amp meter jack itself backwards every switch at SW3 it does what I thought It would do.

yes I did get the Idea here on this forum but I only get to literally work on the Idea this few days (too busy chasing other butterflies)..
the reason is I'm trying to be open as possible and give anyone who's presenting something the benefit of the doubt. and try to analyze and test.. experiment.. if I can't "thought it out". build things to verify if what I see in my mind is the same with what going to happen list things that are real and not and build a conception of "Energy" in my mind.

well that experiment of yours was an added proof that energy was just an abstract concept, its not consumed. you can use it over and over. as long as there is a potential difference. Its not worth trying to literally get THE "Energy" (because it does not exist) but to study how this nature "Balance" itself, If our circuits is in the path of this "Balancing Process" then we will have "Energy". not that I'm trying to lecture.. simply just trying to express my thoughts.. so far that is my conception of "Energy" I'm about 80% sure that I'm right.

I think you should not try to remove the battery anymore . I mean its part of the circuit. a body would not function without the heart..
Don said his suitcase has a battery in it, most of his device also have one.
John Bedini have lots of batteries.
Tesla always had a generator.
its an essential part. A "dipole" an Imbalance
1ma for 30leds is enough to prove the point.

How to measure success?..
IMO you have already succeeded and I thank you for sharing that.
you just have not looped it

[Solarlab]

tswift; from your post, quote:

... "However, testing is showing no measurable power gain. The circuit itself works pretty well: the SSTC resonates strongly, it will light up a fluorescent tube anywhere close. Power levels are just below where the top of the coil will start breaking out a spark. Running in CW mode like this is totally silent and the MOSFET heat sink gets warm but not excessively hot. Likewise, the radiant capture part works pretty well. The supercap bank charges up nicely even from fully discharged (although the coil tuning shifts slightly as the bank charges). When charged to about 12.8V, the pickup coil can deliver about 1A of current in the cap bank. This is plenty to run the step-down inverter stage (kind of a misnomer since it's really a 12V to 12V converter), which draws about 150 mA with no load. There is also a shunt regulator circuit which limits the cap bank voltage by just dissipating excess current as necessary, that's the 3055 transistor with a heat sink.

If I load the output from the isolation transformer with a 16 ohm power resistor (that aluminum rectangle is a 1000W 16 ohm resistor), it draws just under 1A of current (as you would expect at 12V nominal voltage). The discouraging part is that the Lenz effect on the isolation transformer primary can be clearly seen on the ammeter. It takes essentially 1A in to get 1A out. The conversion efficiency of the step-down inverter is pretty good: the MOSFET's don't even need a heat sink, nor do they get perceptibly warm. Transformer core losses are very low, I deliberately used about 3 times the normal number of turns for this core in order to limit flux density and keep hysteresis losses very low.

Loading the output even further confirms this result. With a 1 ohm resistor the voltage at the secondary collapses to about 2V, while the current on the ammeter climbs to about 2A.

Conclusion? "

F.Y.I. (as a quick comment/observation)

It appears this circuit is not yet "asymmetrical" - this is one of the prime
requirements!

- input must be isolated from the output - load can not affect input; no BEMF (OEDS) or
any disturbance of the system's resonance (neither series nor parallel resonance shifts)...

Although I am not familiar with Don Smith's nor your circuit replication; maybe a review of
implementation while considering the asymmetry requirement might shed some light on
the situation; or at worst, provide a (sometimes) helpful break... !!!

Sorry but I do not recall what method Don Smith used to isolate/eliminate BEMF - probably
his coil winding/management arrangement; since, I believe, his pulse/frequency was set at a fixed rate.

Also, just as a note here - forget considering loose or close coupling within the system per se,
to avoid possible "orthodox" thinking re: "transformer action" [think - energy conversion???].

A variety of sources discuss the asymmetry requirement in detail including Vasiliev, Gorchilin,
AAbramovich, Utkin, and so forth ...

Good work, Good Luck...

FIN

[SkyWatcher]

Hi all, Hi dragon, thanks for sharing the pdf.
I'm making some experiments, i first tried with a 12 volt tractor battery in the C1 position and used 20 volt-1.5 farad car audio capacitors in the C2 and C3 positions.
Then also tried placing a 12 volt tractor battery in the C2 position, or the in series with C1 position.
Using a 10 ohm-10 watt resistor in load position.
The 12 volt battery in C2 position seems to be recovering the energy very well.
Using an amp meter to measure the amperage going into C3 and the amperage being recovered into the battery in C2 position, looks equal to me, around 1.23 amps at peak and tapers down.
I placed the capacitors in series to handle the voltage in the dual battery experiment.
The capacitor in C2 and C3 position may be more efficient though, since it probably has less charging losses.
peace love light