Language:
To browser these website, it's necessary to store cookies on your computer.
The cookies contain no personal information, they are required for program control.
the storage of cookies while browsing this website, on Login and Register.
 Storing Cookies (See : http://ec.europa.eu/ipg/basics/legal/cookies/index_en.htm ) help us to bring you our services at overunity.com . If you use this website and our services you declare yourself okay with using cookies .More Infos here: https://overunity.com/5553/privacy-policy/If you do not agree with storing cookies, please LEAVE this website now. From the 25th of May 2018, every existing user has to accept the GDPR agreement at first login. If a user is unwilling to accept the GDPR, he should email us and request to erase his account. Many thanks for your understanding

Custom Search

### Author Topic: LTspice IV Computes That Over Unity AC Circuit Works!  (Read 26170 times)

#### vasik041

• Full Member
• Posts: 209
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #15 on: April 30, 2009, 05:36:02 PM »
I would like copy/paste here fragment of the post from LTspice group by John Popelish
(http://tech.groups.yahoo.com/group/LTspice/message/28962)

-- snip --

Your screen capture shows a great many operation cycles of a
system that stores energy in an inductance over some period
of time, and then releases that energy over a much shorter
period of time. So, low power going in for longer time,
higher power coming out for a shorter period of time. But
the average power going in is still less than or equal to
the average power coming out over a whole cycle of this
process. Energy storage elements like capacitors and
inductors allow for circuits that charge these devices
slowly (low peak power) but dump them quickly (high peak
power) but the average power charging is always equal to or
lower than the average discharge power. Conservation of
energy requires this and LTspice models energy conserving
inductors and capacitors, as their mathematical definitions
require.

-- snip --

#### D.R.Jackson

• Jr. Member
• Posts: 58
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #16 on: May 03, 2009, 05:45:40 AM »
@ D. R. Jackson, I have a simple question.

In your circuit shown above, can I assume that the emitter of Q1, and one leg of the input signal leads are connected directly to a common negative?

This is an important aspect of the circuit to me that needs clarifying before I can offer any pertinent comments on the circuit operation.
I grew up in the analogue era, and designed many audio (sine wave) amplifiers in my youth.

Cheers

Yea that is a common ground, the softwares uses a arrow icon for the common ground and thats the only ground symbol they have.  But you can draw a conduct for ground like I do when I hand draw a diagram.

I had a break through I will post in a reply to another post below.

#### D.R.Jackson

• Jr. Member
• Posts: 58
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #17 on: May 03, 2009, 05:50:08 AM »
I would like copy/paste here fragment of the post from LTspice group by John Popelish
(http://tech.groups.yahoo.com/group/LTspice/message/28962)

-- snip --

~ input versus output power wave cycle period ~

Well those comments challenged me and set me on some research and a whole lot of circuit models and analysis as quick as I could to answer you with what I found out.

First of all the comments made me examine the circuits from every possible angle.  And I had to filter out the periodic energy of L1 from reflecting an AC current into the power supply V1.  And so, obtaining more of a DC power in V1 like it should be, I could study the effects of the circuits power and losses.  As well as work on keeping the output power wavecycles working on the fundamental drive frequency input into Q1.   So that there is no periodic change of the wavecycles.

The model in this reply that you can download below is going to blow your mind....  Like I said a few post back ~ it gets stranger as we go.

My Original Concept

In the circuit diagram below this reply you will see that I have heavily filtered out all periodic energy from the power supply V1 so that the perodic energy of the rest of the circuit, in the equation of things will not effect the energy of the power supply.  Though a small wave cycle energy still exist due to the capacitor and the coils of the filter.

Given that I could make a circuit that was well energy balanced, and was able to replentish a few of its losses.  I would attempt to allow the circuit to replentish some of its current it draws through L1 back onto a capacitor where it would take off current on the negative half cycle and replace it on the positive.  And I would use a capacitor voltage divider for the voltage I need.  And so I would attempt to not allow this current to flow through the power supply.

In the model I uploaded in this post. I have done that.  I do need a small amount of DC collector Current in the uA range for the intial start up solution for Q1 so I have two resistors across the capacitors of the capacitor voltage divider.  And these restrict the DC current I need down to that for Q1 when it is in the cutoff cycle and just begins to conduct.

I was not sure I could ever get a model of this concept to work in LTspice.  But given the balance of the circuit and its concepts.  I was hoping that the concept of resonant energy would help along with the concept of replentishing a seperate current into the circuit in the virtual power supply feature of Dx.  And Dx only has to add a little collector current into the initial operating solution and it is 180 degree phased from I(L2).

Well the circuit fired up in LTspice and ran.

You can download and run it and check the input power of V1 with the output power across C6 in this model.  Check you nodes so you know which ones to enter in, for you wattage plots.

My theory is this, all we need is a voltage potential at input and this can come from a mere capacitor voltage divider.  Though LTspice reveals that we do need a little DC current for our Class C no signal cutoff condition.

If we can induce a current in the circuit with the voltage across the capacitive voltage divider then we might be able to recirculate it around and around in the circuit.

The circuit appears to break down in analysis when I try to add loads to the circuit to take of the power.  My original theory says that the power is now a recirculation of current around in the circuit just like in a DC of AC generator that recirculates the same electrons around in its closed system of the generator windings and the load.

SO now in my theory the circuit should be my generator.  And no current should be flowing in V1 except the small collector current we need at a certain point in the phase cycle.

LTspice thens to still want to take the wattage away from the power supply in the analysis of adding a load.  Of course the load has to be placed in the ideal place of the system to balance the equations.  The load should only upset the balance of the circuit.  And no current or power should be required from the power supply in this circuit except that small amount previously mentioned.

So either I can not find the proper place to load the circuit to extract some power, or else LTspice breaks down in analysis beyond this point.

Anyways, this circuit here was the original idea and the whole matter of investigation by me to see if I can make the concept run is software.

If this can lead to something in the future, it will take many minds looking over it and analyzing it all.

Technically, following conventional views and thinking this circuit should not work at all.

But LTspice seems to support my field force theory of which the capacitor voltage divider was the original concept objective.

I am sure that the folk at the LTspice group now will not be able to understand this circuit in a long time to come.

Run it for as many seconds as you want.  And according to conventional theory there should not be any circulating power of any kind in this circuit and certainally at the levels that LTspice computes.

If you have trouble seeing in the graphic below the thin line of the plotted power supply input power of this model.  Being its not a real good graphics card I have, just run the Spice file download below and see for yourself. (V(n001)*I(V1)

So, here is an anomoly.

Like I said earlier in these post, it get stranger and stranger as we go along here.

This one model here is one I am sure many people around the world will ponder for a long time to come and can not be explained.  I can however explain it in my original theories.  And I just had to see if I could model my theories.  I did not think that this original idea would work in the end after I tweaked up the models to run good with a conventional power supply. But I did save the best for last.

And we need many more minds here, and so, my models from this point are open source and please enjoy them.

Please note that you can change the two capacitors of the voltage divider to as low as 10000uF each and not have to modelt the circuit with two 1F caps.  The small AC current in V1 will be of shorter wavecycles but of about the same level.

I sure hope that there is someone in this world who can find a solution for taking off the power of this circuit!  So good luck.  How do we load it?  We need an engineer with some advanced math analysis to do this.

D.J.

(The capacitor voltage divider supplies +50V, and the two divider resistors are for a little collector current.  L2 and DX provide more DC collector current by the self induction of the circuit.  The power across C5 and C6 are extraordinary when you analyze this circuit that is not even suppose to work in conventional views.)

You want to plot:

V(n001)*I(V1)

V(Vc)*I(C6)

It will be hard to explain this model, and how it is designed to work.  But all we need is the field force of 50V and the circuit is designed to circulate around the current that it needs.

#### rensseak

• Sr. Member
• Posts: 330
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #18 on: May 03, 2009, 12:02:07 PM »

#### D.R.Jackson

• Jr. Member
• Posts: 58
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #19 on: May 05, 2009, 12:12:58 AM »
Ok I have done some analysis of Pure DC input power versus AC output power wave cycles and their period and average of the peaks.  And have a model I am working on to tweak up where the period of the wave cycle is the same as that of the input signal, but the postive going power cycle peak is of longer duration (longr than 180 degrees) than that of the input siganl making the powering up cycle longer per wave cycle and the average of the peaks come out to over the input.  I will upload those models as soon as I am done working on them with a text as well as reference graphics.  Both the duration of the power on the postive peak is longer and the average of the peaks are greater than the pure DC input power when multipied by 2 for peak to peak AC.

Neadless to say it shocks me.  I will upload those as soon as I can for you to run and prove to yourself.

I am just passing through here quickly to add some comments to my post and must get back to work here.  So when I have free time to read up on all of the replies to my post here I will answer them.

I have had some very important analysis observations and break throughs in the past days here.

But first of all I must comment that, I have to watch about forgetting to make power supply adjustments in my models when I replace one transistor type in the circuit with another since I can end up with too much transistor collector voltage on the model.  The remedy for you is to turn down the power supply voltage if I forgot and left it high.  The results is lower output watts, but all things being relative the circuit performs the same ratio wise.

I did not like the way that my capacitor voltage divider was performing over time in my novelty version of my sonic resonator circuit and so scrapped that idea for the moment.  And went back to the power supply to power the circuit.  But I will return back to look at this idea again later.

For the moment I have made some extra energy conservation improvements to the circuit and have made it work better.

What I have been doing now to measure the performance is comparing the input power in terms of DC peak power versus the average output power wave cycle peaks across a resistor R1 between Vc and ground.

First of all, contrary to someones measurements of the period of the wave cycles, the period of the wave cycle is exactly of the same period as the input signal source.  However, the positive powering half cycle is converted to a longer duration, making the output power on the half cycle longer per cycle and thus powering the output longer in joules per time terms than the original signal sinewave input.  The positive hald cycle peak of the output power wave cycle is thus longer than that of the input power and equates to longer powering time per cycle which alone is an energy enhancement.

Next of all, by adjustment of the parts with regards to the concepts of the principles of the circuit and analyzing the peaks of the output per positve half cycle and then averaging them as the folk over at the LTspice forum claimed that they did.  The averaged power output peak to peak terms with regards to the input power in DC peak terms is over twice the DC input peak power by 1 watt in some models.  However,since the positive half cycle is also of a longer duration than 180 degrees, this equates to the output power in watts being higher than the average of the peaks, in terms of joules per time.  Or duration of the power cycle.

At time certain things did not add up right like the theory says.  However what I found out is that I have to work on adjustment of my partts in the circuit to recover and conserve energy losses and these proved to me that with the above comments that it is possible to make the model realize and demonstrate that the comments and critque of the good folk over at the LTspice forum were wrong.

I realize that there are many good folk over at the forum and they have helped me with things for my radio circuit design projects before.  But I realize also that sometimes someone out of habit might look at a plot and formulate a conclusion without actually measuring things in the plots.  Which, well that happens.

What I am going to do for you here who are interested in running some models that will stand up to the analysis.  Is to upload those models I am working on measuring now, along with the input power wavecycle plots and the analysis of those things mentioned above.  And you can run those models in LTspice to examine whether not yourself if what I am saying and have found out is true.

There in one thing though about the circuits that make it kind of hard to use the power and that is the way it wants to be loaded at out put that will not effect the resonance (timing) of the circuit and degrade its performance.  And that is the fact that it wants to see a capacitor or a resitor across the output.  And so, the circuit seems for the moment to only be usefull for power enhancements to residential lighting (florescent or incandesent) as well as for resistive loads such as electrical heaters.

Knowing that the circuit did not want to work into anyother thing that is useful to us other than resistances such as electric lights and electric heaters.  I decided that if this is all that the concept of the circuit could ever be upscaled and used for.  Then that was what I have to focus on.  And so, it might not be something that will revolutionize the world of electrical energy as we might have had hopes for.  But we do need something to enhance our electric light and electric heating technologies and so, I guess we have to go with what the circuit demonstrates that it wants to see as an output.

I will give as many analyusis notes and observations as I can in notes insid emy Spice file upload that I will upload here when I am done.  And LTspice plot graphics also of my analysis I done and refer to.

And once everyone is then satisfied that the models running on a power supply (and not via a capacitor voltage divider such as I uploaded as a novel model concept) compute in their LTspice simulations to an energy enhancement.  I will then return back to some former ideas and see if I can better realize them with my newly teaked up models.

True, there are days when it all looks good and then bad.  But my discovery that my original ideas for tweaking the circuit to effect it wave cycle period and its averaged out power by analysis of the waveform peaks averaged out.  Made me look up again.  And so, if I can tame the capacitor voltage divider concept.  Which is a simulation that runs way to long on myslowPCand would have to run a few hours to test my theories for solving the discharge and allowing the circuit to balance it out in time.  I will return back to that and see if I can come up with something that will make it all better even more.

In theory the lower capacitor of the voltage divider should charge down the voltage and then stop somewheres in time and the circuit should charge it back up and thus over time have a slow running wave cycle of positive and neagative peaks.  But that remains yet to be seen.  But if I can accomplish that, then we can have the kind of power expression that the current simulation of that model demonstrates.  And I need a faster PC to analyze that slow running simulation.

Well I try to stay on top of energy conservation and loss recovery views in my circuit concepts, that follow the laws of physics and do not violate any of them.  Which is the whole matter of the design ~ energy conservation and recover (or to say recycling of losses).  Abd so, as long as the views are not violating physical laws, then LTspice with all of its programmed in math and physical laws will run the models.  If the laws violate those inside LTspice, then the model will not run or not demonstrate anything useful.  So all things must be accounted for in terms of energy loss recovery and conservation, and then the idea is to allow the electromechanical resonance of the circuit to add in, if the theories of all the Tesla buffs are true, an amount of latent energy that resonance provides.  It is such that the idea of electromechanical resonance is considered to be an energy plus rather than a loss.  And so, this is the idea of the circuit, to test those assumptions long held by Tesla buffs.

And who knows, what I might not be able to do with my circuit.  Others in universities with advanced degrees and insight as well as advanced math analysis in teams scattered out across the world.  May come up with a synopsis of the circuit and an equation from which they can convert the circuit into another form one day that will realize things that I can not at the moment.  But I do have to present software analysis of the principle in a beyond a shadow of a doubt way that anyone in this world can run in the software and analyze for themselves. So thats my next step. Those software analysis of proofs for feasibity reasons.

So I will be here when I am done to give you those models I am working with now for you to run and analyze.

What I will provide then as an LTspice model for you to run is the one that I can design with the best expression of power versus power with all the notes on how to effect the performance of the circuit. And then, advanced engineering students and researchers can analyze everything mathematically.

All I ask of anyone who uses these models for analsysis and study is that when you make public mention of them ~ you always give me credit by name. And that if you want to market the technology you must compensate me with something good in return, you know, licensed by me.  And accredited by me.  For all of my many years of work in conceiving of the concepts and principles as the father of the invention.  And so, these devices are provide to all in an Open Source development manner.  Also, if you come up with a unique model you must provide it to me for me to look at and analyze.  And it must always be in the form of an LTspice file and not Tina Spice or Pspice, etc.  And I will always want to see a version of any product that comes to be marketed with these ideas.  So I can sample it and use it.  Without any expense to me to procure.

My finalized models for you will come to you with Acceptable Use Terms that you must abide by for using the concepts in research.  And you must agree to compensate me if you obtain a research grant for my innovations, and for if you want to consult with me.  And I reserve the right to charge a fee for consultations.

#### TinselKoala

• Hero Member
• Posts: 13958
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #20 on: May 05, 2009, 01:21:53 PM »
Your work is interesting, but there's a lot to get through. So that we are on the same page, can you confirm that we mean the same things when we are talking about certain concepts?
Joule = the fundamental unit of ENERGY, and is equal to Watt-second.
Watt = the fundamental unit of POWER, and is equal to Joules per Second, that is , (Watt-Seconds)/Seconds.
And the term "per", as is your phrase "Joules per second" is normally to be interpreted as the division operator, as in Joules per second = (Joules)/(Seconds), which of course equals Watts.

Am I with you so far?

I am asking this, because it sounds to me like you are still trying to compare Watts (that is, instantaneous power) in and out, when you should be comparing Joules (total energy) in and out. When waveforms are considered, the proper computation of energy goes like this: You do an instantaneous multiplication of your current and your voltage waveforms. This gives you an instantaneous power waveform. Then you integrate that waveform over the time of the measurement period. This integral of the instantaneous power will be the Energy in Joules, and is the appropriate measurement for Coefficient of Power (CoP) or whatever overunity / underunity ratio you might want.
At least, that's the way I understand it. Of course I could be revealing my hide-bound ignorance here for the world to see. But it's not really my field, after all.
Anyway, when you say "Joules per second", why don't you just say "Watts" if that's what you mean?

#### WilbyInebriated

• Hero Member
• Posts: 3141
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #21 on: May 05, 2009, 02:49:25 PM »
tk, it's my understanding that 'watts' is used more often in high school level books/discussions on the subject and 'joules per second' is used more often in more advanced books/discussions. i have no idea why. personally i think 'watts' should be done away with, but that's just me.

#### TinselKoala

• Hero Member
• Posts: 13958
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #22 on: May 05, 2009, 03:14:18 PM »
tk, it's my understanding that 'watts' is used more often in high school level books/discussions on the subject and 'joules per second' is used more often in more advanced books/discussions. i have no idea why. personally i think 'watts' should be done away with, but that's just me.
It's because "Watts" doesn't capture the whole story. It's easy to conceptualize, using the hydraulic analogy and DC currents, that Watts is Volts times Amps (pressure times flow) and represents some total ability to do work over a period of time. But with more complex AC waveforms or duty cycles, the situation is more, well, complex.
The (joules) = watt seconds, and (joules/second) = watts, these are Definitions and are part of the standard system (SI) of physics notation.
See http://physics.nist.gov/cuu/Units/units.html Table 3

#### WilbyInebriated

• Hero Member
• Posts: 3141
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #23 on: May 05, 2009, 03:20:00 PM »
It's because "Watts" doesn't capture the whole story. It's easy to conceptualize, using the hydraulic analogy and DC currents, that Watts is Volts times Amps (pressure times flow) and represents some total ability to do work over a period of time. But with more complex AC waveforms or duty cycles, the situation is more, well, complex.
The (joules) = watt seconds, and (joules/second) = watts, these are Definitions and are part of the standard system (SI) of physics notation.
See http://physics.nist.gov/cuu/Units/units.html Table 3
sorry, i said that wrong. i have no idea why they choose to use a term that, as you eloquently stated, doesn't tell the whole story to jr. and high schoolers. i think they should start with 'joules per second' and 'watts' shouldn't even be brought into the discussion until it is at a more advanced level.

#### D.R.Jackson

• Jr. Member
• Posts: 58
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #24 on: May 09, 2009, 01:27:52 AM »
Well the sonic resonator is perfected now, and you can download the final model below to verify the statements and the wattage plot of this post.  Despite some of the comments made in reply to me, check all of this out for yourself.

Consider the following based upon a DC power input of 5.29 Watts pure steady state DC.

First of all the DC signal output has been worked on and perfected into a DC signal square wave output.  The amplitude of the DC signal is 8.23 Watts.  The rms, average and peak of a square wave are the same.  And please note that the output at this point across the 120 ohm load resistor is a DC Peak signal output of 8.23 Watts.  This has not been converted to peak to peak AC in the model but I have a model I have done this with.

Also, extra energy is stored in the half wave cycle by it being extend past the 180 degree half cycle point by +50 degrees to the 230 degree point in the wave cycle.  So energy has been stored in the half wave by extension of the half cycle of 0.5 wave cycles to 0.64 wave cycles.  As can be seen in the below plot.  And you can download the model to run and verify that this is the case below also.  You can easily see that the half wave in proportion to the entire wave cycle, is reaching over into the negative half cycle and so, has its half cycle period extended.

Since this is a DC output signal, the positive half cycle is the On state, and the negative half cycle is the Off state.  And so, can be viewed the same way as a digital signal.  The Period of the On state then is extended +50 degrees to the 230 degree position of the wave cycle.  So the the period of the On state of the power cycle is thus now 64% (230 degrees) of the wave cycle rather than 50% (180 degrees) of the wave cycle.  So more energy is added into the power cycle this way, and not just only in the unusual peak power output amplitude that the model demonstrates in software simulations.  The power then is stored in the wave cycle both by amplitude ~ as well as by extension of the period of the half cycle to 230 degrees.

Now there is no way with ordinary circuit concepts you are going to get a square wave output of 8.23 Watts output with only 5.29 Watts DC input from a DC power supply.  If you do the math on a purely lossless system model (theoretically 100% efficient system) you will not get a  8.23 Watts square wave calculated out of an input of 5.29 Watts DC input ~ unless there is more to the circuits concepts and principles than meets the eye.  Such as conservation of losses and restoration of those losses back into the circuit to do useful work.  And the use of electromechanical resonance.

If you convert 5.29 Watts DC to a DC Signal Peak of 5.29W*1.414 = 7.48006W in a lossless  model (and this is based upon a sine wave peak rather than a square wave).  So the rms value of the sine wave peak will be 5.29 Watts, and our rms value of our square wave is the same as its average and peak which equals 8.23 Watts.  Understand?  See?  Thats over unity.

Ok now imagine that you upscale this device into the hundreds of watts range or the kilowatts range.

Example:  Input power is 529 Watts, the output square wave power then equals 823 Watts.  And so, for lighting and heating we have a unique scenario for cost saving.

Or lets upscale this to look like this:

5290 Watt steady DC In ~ 8230 Watts DC square wave signal Out.

(By adding a capacitor in series with the output and a light bulb or heater you have a simple method of converting the DC to AC for such loads.)

So, the idea seems to hold some interesting potential especially for industrial lighting and heating cost savings.  But would be something nice to use out here in the private residential sector for our lighting and heating needs.

For industrial manufacturing facilities that use a few thousands watts per day merely for lighting, and then in the winter uses infra red space heating for large open areas.  I am sure that a good amount of cost savings can be accomplished by use of these circuits to enhance the power.  And so, I would classify these devices as "electrical power enhancement circuits".

Summed Up:

Here is how it all adds up.  After the software simulation runs the circuit's simulation the DC power input starts to become steady after 3 seconds.  So the initial charge up of the circuit is complete and the current and voltage of the power supply become steady at around 5.29 Watts steady state DC.  Now since there is no periodic wave cycle power in the power supply at this point.  The rms, average and peak power input of the power supply are all one and the same, 5.29 Watts pure DC steady state.  And so, the square wave output is likewise to the DC power input, the rms, average and peak power of the square wave are one and the same, 8.23 Watts.  So we have a power enhancement of: 2.94 Watts.

Extended Half Cycle Period:

Now its hard to tell how the extended period of the half cycle will average out in terms of the power that is placed in the extended +50 degrees past the 180 degree half cycle.

It just so happens that 5.29 Watts / 8.23 Watts = 0.6427

When rounded off, this equals 0.64 or our extended half cycle period.

Since there appears to be a direct relationship between the square wave amplitude and the extended period of the half cycle.  I can not say that we can extend the period of the half cycle anymore.  And so, there is a limit.

If it can be extended I do not believe that it can be extend so that the angle of the period is 270 degrees, which would reach the negative peak.  But, if it is possible to extend it more, then I suspect that it will not go past 0.7 wave cycles at most.  Perhaps 0.68 might be the limit if that amount of periodic extension can be accomplished?

It would perhaps take a super computer able to analyze and also make design changes and tweaks, to come up with some sort of improvement.  Given a standard model to begin with, the software then would analyze the circuit, change a parts unit value and analyze the results of the various changes made and tabulate the data ~ and then formulate the most efficient model from its data.  So thats what we need to happen somewheres, someday.  Its kind of hard for one person to keep track of all of the various changes and their effect and data.

Innovation:

I guess all we are going to get is baby steps here, and so, the rest is up to a whole army of researchers and university students to improve and make something out of the math of it all.

I am hoping though that someone with some Spice library models for vacuum tubes might make this into a higher powered device model.  And maybe realize a greater ratio of input to output power for use in such things as industrial electric lighting and heating.  My tube suggestion for this is the 4-65 A beam power tube.  And if anyone comes up with a LTspice library model for this particular vacuum tube, then I would like to have a copy of that to model circuits with.

I would have to say that with all of the circuit components, that would normally result in allot of losses in such a circuit.  And with a DC power input that in no way under conventional thinking can equate to a square wave output as this model demonstrates, that this model best demonstrates what some refer to as over unity.  Which in some engineering circles is a taboo and hence forbidden term.

You can download the model here and run it and who knows, somewheres, sometime ~ someone else may advance it even more.

As for the ways I have found to load the circuit to take power off, you can use a capacitor, a resistor or a coil or transformer.  However when you add a diode or two to rectify the output and then a filter capacitor you have immediate losses.  I believe that one of the things that happens is that the diodes pulse the half cycles and so, these pulses feed back into the transistor collector circuit of L1 and L2, etc, as a reflected wave and upset the circuits timings.  And the filter capacitor draws allot of charge current.  This does not mean that in time a remedy will not be found.  However if no remedy is ever found, then the application of the circuit is such that it will work with resistive loads such as electric lights and heaters.  But these things account for allot of our power use.  Electric heating is used in manufacturing on a daily basis and not just in the winter.

Applications:

Given enough output voltage, the circuit I believe could change the way we light our buildings particularly when using fluorescent lights that respond to higher frequencies better than they do to 60 Hz.  Some fluorescent lights can be fired up without filament heaters in each end.  Tesla did this using H.F. frequencies.

Also, in some of my models I have been able to effect high instantaneous power peaks.  Such peaks I believe would be useful with fluorescent lights and with quartz (infra red) heaters as well as with oil filled heaters that store heat in the thermal mass of the oil.  The idea of the oil heater is to use the regular 120V AC line to bring the oil up to hot, then this device can replenish heat losses with instantaneous high wattage peaks.  And so, sort of stroke the heat, and stoke it up.  While resulting in a cost savings.

However the square wave output when used in a higher voltage and higher powered model is able to flat out run a quartz or oil filled heater and bring it up to hot, and then can be switched, manually or in a self automated manner to provide instantaneous power peaks which is more applicable with oil filled heaters than with quartz types (the instantaneous peak mode I mean).

And well, the device can be used with incandescent light bulbs also.

Since the device works on a DC power supply input, in an emergency the device can be powered by a few automotive batteries and so, provide light and heat.

Since the device can be used with batteries it then would make a great addition to solar electric systems since the device can run straight off of the storage battery bank.  And so, enhance the lighting and heating capability of solar electric systems making them more efficient.

The margin of efficiency that my low powered simulation model under discussion here demonstrates is equal to:

8.23 Watts / 5.29 Watts = 1.55:1

Now to calculate the improvement:

5.29 Watts / 8.23 Watts = 0.6427 ~ equal to the extended period of the half cycle

0.64 wave cycle - 0.5 wave cycle = 0.14

0.14*100 =14%

14% then equals the power enhancement improvement that this model would provide to a solar electric system.

More Software Analysis:

I have seen applications written for LTspice by hobbyist that run in the program.  My thoughts is that someone might become familiar with these circuits.  Or a team of researchers, may write an application for use in LTspice, or an whole new Spice software, that can analyze this circuit (and this circuit only)  and self adjust and tweak around the parts values so that the software can come up with a design with the best figures of Input versus Output.  And such things as transistor or vacuum tube max voltage and current ratings not be exceeded.

Conclusion:

I believe in sharing things that are important with others.  And I know allot of us study and try hard to accomplish things with regards to potential over unity or alternative energy innovations.  And so, more minds working on somethings leads to something in the world sooner.  And who knows, someone may be working on something that we can not conceive of at the moment that will achieve something even more for us when combined with these ideas.

Then again, we may not ever find anything else that seems for the moment to be feasible at least in software analysis.  And I know I don't have any 1 Henry to 0.5 Henry transformers at the moment to prototype a test bench model with, but someone out there may.

I of course do hope that something else will come along somewheres to help round out the world of alternative energy concepts.  And whether or not their ideas can be combined with mine for a more complete and effective concept.  Well, is not the point, if someone else has something that is working or at least can be modeled in software, then I hope that their innovations also will add to the world of things that we need.

More than anything, we need people to become focused on something that appears to be feasible in software analysis.  And so, allow as many as can and want to, to do those analysis themselves and so, pool together.  And then as a world wide team of interest, we can make test and if all goes well, we will have something that I hope will work on the test bench.  An who knows who will be the first to get all of the parts together to test it all?  Sure wished today that I had the parts to test it all.

If I could have a transformer to test these circuits with, I would like one of those new toroid core power supply and audio transformer types  with a 1 Henry primary and a 0.5 Henry secondary.  And there is a new core material that is used with audio transformers that consist of powdered ferrite with piezio electric crystal powder impregnated into the powdered ferrite core.  It is said that this core enhances the audio quality and so, I wonder what it might do in this circuit?  I wonder if the piezio electric crystal might be capable of adding its electrical energy to the sum of energies?

Ok, when I saw that my concepts worked in software simulations, I thought then that I had accomplished it.  But I had to work at it all to get something substantial out of it.  And so, days it looked all good and then days it did not.

And so, I concluded early in my testing that if all I could get out of this is enhancements for electric lighting and heating then thats perhaps what we need the most at this time.  And so, though the device fails when I attempt to convert the output to DC for obtaining an energy efficient power supply it does shine with potential for use with lighting and heating.

So, when I realized that my square wave output held up to the idea of a notable and demonstrable energy enhancement.  I was satisfied with that.  I did realize in the end what I was looking for.  And its hard to argue with the results when it comes to us in a square wave output in watts where the rms, average and peak power of the square wave are all one and the same.

LTspice IV analysis then demonstrates that the over unity device concept works in software, and is feasible in software analysis.  And anyone can run the simulation for themselves and analyze the DC input versus the square wave output power ratio.  And you can download those files here at in this post.

#### TinselKoala

• Hero Member
• Posts: 13958
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #25 on: May 09, 2009, 03:00:54 AM »
Power is not energy!!
A device isn't overunity unless it outputs more Energy than input.
I realize that is simplifying a bit. But it's still true.
Can you show, computed with your LTSpice simulation, the time integral of your watts waveforms?
Also, there can be a lot of energy gained or lost in those little wiggles at the top of your "square" wave, plus the rise and fall times of the "square" aren't all that steep. So I believe you are overestimating the energy in your output, even though your math, comparing some kind of average watts and assuming perfect squareness of your waves, might be right.

#### TinselKoala

• Hero Member
• Posts: 13958
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #26 on: May 09, 2009, 03:10:52 AM »
I just took a closer look at the numbers on your graph. That wave is nowhere near "square"! My oscilloscope's calibration trace is a more accurate square wave than that, with less ripple on the top and 20 nanosecond rise and fall time. Your rise and fall times appear to be on the order of 50 microseconds or so!!  That is, if I'm reading your numbers right. 300 microseconds between vertical divisions on your scope, right?
That's not a square wave.

#### TinselKoala

• Hero Member
• Posts: 13958
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #27 on: May 09, 2009, 04:11:01 AM »
This is a square wave.
Horizontal scale 50 microseconds/division. Vertical scale 10 milliVolts/division. Note the fast rise times and lack of ripple, compared to the "square" wave in the LTSpice simulation that you use in your power calculations under the assumption that it is perfectly square. On this scale your rise slope would probably be about 60 degrees, and I don't believe anybody would call it "square".
But I wouldn't even trust this square wave to be accurate in a real energy calculation.

But, using your numbers, again: I estimate that the deviation from true squareness in your output waveform brings its power down to something under 8 watts. But, unfortunately, 5.29 watts continuous input power, divided by the 64 percent duty cycle of the output waveform (an overestimate, by the way) is 8.265625 watts.

(A note on accuracy and significant digits: If you say that an answer to a computation is "8.265625" you are saying that it is NOT "8.265626" and NOT "8.265624" and so on. This accuracy is impossible if the input variables are not also known to at least the same accuracy. Since the input variables in any real circuit are impossible to know with this accuracy, at least without very special techniques and equipment, it is simply FALSE to state so many significant digits in an answer about the real world. Your answer cannot have more precision than the LEAST precise variable that goes into the computation.)
« Last Edit: May 09, 2009, 05:09:26 AM by TinselKoala »

#### TinselKoala

• Hero Member
• Posts: 13958
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #28 on: May 09, 2009, 05:28:41 AM »
"If you convert 5.29 Watts DC to a DC Signal Peak of 5.29W*1.414 = 7.48006W in a lossless  model (and this is based upon a sine wave peak rather than a square wave).  So the rms value of the sine wave peak will be 5.29 Watts, and our rms value of our square wave is the same as its average and peak which equals 8.23 Watts.  Understand?  See?  Thats over unity."

No, it's not. It's an error. You can confirm that it is an error by simply drawing the waveshapes (a constant DC at 5.29 watts, and your "square" wave, at sufficient resolution) on a piece of graph paper, cutting them out across the same time interval, and weighing them, or otherwise comparing their areas. (The area of the watts waveform is equal to the energy content). Your statement above does not even take into account duty cycle. It's easy to get high peak power levels from a low-power continuous input, as you show later when you do compute duty cycles. Your output waveform isn't square enough to justify your assumptions, the true duty cycle is less than 64 percent and the true power level is less than 8.29 watts, and once again, the appropriate measurement is ENERGY, which is power integrated over time, and your input energy is greater than your output energy. SO no overunity here.

For example, if we correct for your slow risetimes, the duty cycle might be, say, 62 percent instead of 64 percent. This means that the 5.29 watts constant input power could deliver as much as 8.532258064516129032258064516129 watts. No overunity, even if one accepts your own model's power calculations for that ripply signal,which I do not.

And see how ridiculous it looks when you just grab all the digits your calculator spits out? The proper answer from that computation is 8.5 watts.

Understand? See?
« Last Edit: May 09, 2009, 06:14:40 AM by TinselKoala »

#### TinselKoala

• Hero Member
• Posts: 13958
##### Re: LTspice IV Computes That Over Unity AC Circuit Works!
« Reply #29 on: May 09, 2009, 06:15:50 PM »
Of course, you could just start yet another thread. But one wonders why you haven't addressed a single point I've made (and others have made) in this one.