Wilby
I appreciate TK's efforts ,without men like him this forum would be all talk
He takes action gets to the point [no BS] ,can take on almost any job
,skills in many areas

TK I SALUTE YOU AND THE BIRD
Chet

so taking action by "replicating" a circuit and NOT using the correct/exact mosfet on the build and then saying it's "exact" ISN'T BS?

huh? well you and i seem to have a differing definition of BS i guess...

yes i agree on the almost any job statement, he is surely a "jack of all trades, master of none."

The Dude has skills
As far as Rosemary goes he's just Coasting [not breaking a sweat]

Chet

Thanks for the support, Chet. Sometimes it feels like nobody cares.
 

Thanks for the support, Chet. Sometimes it feels like nobody cares.
 

TK I have a lot of respect for you and your methods.

My advice to you is to ignore folks like this. Sometimes their only aim is to entice you into battle. Usually it's not worth the time or effort.

Regards,
.99

TK
The big man in the house
Rosemary said
TinselKoala - I viewed your Utube on our circuit. It was interesting but my first point is that the waveform on both the first and the second are nothing like our own. It did not go into resonance and it is nowhere near as complex as the one's we generate across the load. Something between your circuit and ours is out of synch.

I'm so sorry I can't get a picture of our waveform. I'll ask my co-author if he can perhaps organise something but suspect that it will take a little time. Michael John Nunnerley was bang on in pointing out that we were probably using a resonating frequency. Indeed we are. We sweep the duty cycle until it first goes into oscillation. That is the point that we usually get the best results. The waveform is not periodic - which makes for some tricky calculations of power - hence the need for that Fluke scopemeter and for those calorimetric values.

The other thing I did not see was the diode return to the positive teminal of the battery? I presume this was included? But I have a real problem in watching the battery voltage collapse on the second video. The load is light - and with the best will in the world, even with a 90% on duty cycle, I cannot understand how a fully charged 24 volt lead acid battery can deplete within the first 10 minutes of running. Were you using a flat battery? Certainly one would expect that it's capacity would enable a current flow of a resistor at 10 ohms (was it?) - therefore not more than 2.4 amps for longer than 10 or even twenty minutes even without any applied switching cycle. I also noticed at one stage that the battery seemed to lose it's voltage entirely - then go into a negative voltage value and then spring back to 24 volts. I can only say that such is really strange and in the years that I've been testing our circuit have never seen the like. I am reasonably certain that your battery was nearly flat or that its rated capacity may be somewhat questionable.

I am also concerned that you used a different mosfet. Not that it needs to be identical to the one that we used - but I am just not sure of the properties of the one that you used.

Regarding the 555 switch as opposed to the function's generator. There really should be no real difference between the two. However it is easier to adjust the 555 switch to enable that resonance which is both the object of the circuit and the main object of the thesis. Why you are not able to get the circuit into oscillation I do not know. Perhaps you must vary the frequency better. Incidentally I could not make out the frequency you applied on that demo.

The niceties regarding the actual published switch and the one that you built - here I cannot comment. What I do know is that if the switch is set at 5% on and the load shows 5% on then it cannot default to 90% on. It is that simple. I could not make out the positioning of the probes in relation to the load resistor. Again. Your questions seem earnest - but your references not so easily detected on that video.

In any event, the other problem I have is with the value of that spike which your referenced in the second video - I think it was. Our spike is generally far higher, upwards on 120v but is largely dependant on the applied duty cycle. In any event the amount of energy in the on cycle is always marginally more than the amount delivered by the spike. The value to the energy gain is in that this energy is repeatedly returned to the battery and to the load. This can be seen if you use 2 x 12 volt batteries as we did. If you run the test on the one battery and connect the second to the first with a common negative rail - then feed the diode to the positive of the second battery, you will see an immediate recharge to that second battery. That test was done to prove that the returning energy does, in fact recharge the energy source.

So it is that we justify the value of the energy delivered by the battery as the sum of the on and off cycles. The energy dissipated is the product of both cycles. Therefore is there a gain. And at this fast resonating frequency the gain is really substantial.

I do hope this addresses those points that you repeatedly refer to through these threads

Incidentally, TinselKoala - there's another point. We actually ran our tests with a control. The reason the published article and the paper deal with a test period of 10. something hours is because that is how long it takes the control battery to deplete its energy. For some reason, both in the quantum article and the paper I was specifically advised that any reference to battery duration was essentially irrelevant. Apparently battery draw down rates are subject to too many vagaries?

In any event, the actual draw down rate of the tested batteries is consistent with the energy measured to be delivered by the battery as the difference between the energy measured and calculated from the two cycles of each waveform being above and below zero. At the end of that test period the test batteries are more or less the same as at the start of the test period. The control is entirely flat.

We then recharged both battery sets (always used typical 12 volt car batteries) and swapped the control with the test. Variations of this was called for by BP to enable their accreditation of the tests. It was exhaustive and painfully repetitive.

CHET
PS
TK thank God you showed up ,or they wouldn't have had anything to talk about
You make us proud !!

I'll reply to some of these points here.

Rosemary said
TinselKoala - I viewed your Utube on our circuit. It was interesting but my first point is that the waveform on both the first and the second are nothing like our own. It did not go into resonance and it is nowhere near as complex as the one's we generate across the load. Something between your circuit and ours is out of synch.

Scope traces can show very different features depending on the scope settings. It is possible that I am emphasizing different features than you are. It's hard to reproduce a scope signal...if there isn't a model to work from. I hope you can show us what your scope traces look like. It does not mean that "something is out of synch."

I'm so sorry I can't get a picture of our waveform. I'll ask my co-author if he can perhaps organise something but suspect that it will take a little time. Michael John Nunnerley was bang on in pointing out that we were probably using a resonating frequency. Indeed we are. We sweep the duty cycle until it first goes into oscillation. That is the point that we usually get the best results. The waveform is not periodic - which makes for some tricky calculations of power - hence the need for that Fluke scopemeter and for those calorimetric values.

It's hard for me to see how this circuit could achieve any kind of real resonance at the frequency used. I have been able to reproduce what LOOKS like aperiodic waveforms but this is due to false triggering of the Fluke 199 ScopeMeter. The analog scopes are not as sensitive to this.
Also, the idea of aperiodic waveforms and resonance are, shall we say, oxymoronic. Like jumbo shrimp.
And I have swept both duty cycle and frequency through the entire range allowed by the 555 circuit as shown in the Quantum article.

The other thing I did not see was the diode return to the positive teminal of the battery? I presume this was included?

Please take a closer look.

But I have a real problem in watching the battery voltage collapse on the second video. The load is light - and with the best will in the world, even with a 90% on duty cycle, I cannot understand how a fully charged 24 volt lead acid battery can deplete within the first 10 minutes of running. Were you using a flat battery? Certainly one would expect that it's capacity would enable a current flow of a resistor at 10 ohms (was it?) - therefore not more than 2.4 amps for longer than 10 or even twenty minutes even without any applied switching cycle.

Yes, those batteries were depleted at the start of the video. And they are 2 Amp-hour, whereas according to your article yours are 20 A-h. And they are surplus pulls from decomissioned UPS units. No, the load is not light, it's heavy, at 10 ohms and 96 percent ON, as produced by the published 555 circuit.

I also noticed at one stage that the battery seemed to lose it's voltage entirely - then go into a negative voltage value and then spring back to 24 volts. I can only say that such is really strange and in the years that I've been testing our circuit have never seen the like. I am reasonably certain that your battery was nearly flat or that its rated capacity may be somewhat questionable.

As I thought I demonstrated in the video, cheap digital multimeters (and midrange scopemeters like the Fluke 199) are confused by spiky inputs and will read all kinds of strange things. It does not mean that is what's really going on in the circuit. Many of the folks doing this kind of research have, in the years they've been testing, seen the like many times.
And yes, as I said before, and in the video, the batteries were flat, but that doesn't have anything to do with the wild and sometimes negative DMM readings.

I am also concerned that you used a different mosfet. Not that it needs to be identical to the one that we used - but I am just not sure of the properties of the one that you used.

You can look up the data sheets and compare the properties. The main difference is in gate capacitance and turn-off times--the latter is a significant difference, because it means that, driven by the 555 timer in the published circuit, the IRFPG50 mosfet will be ON even LONGER than 96.3 percent of the time when driven by the 96.3 percent ON duty cycle that the timer generates. On even longer.
But don't you say in your papers and patent applications that the mosfet isn't critical? (And I now have tested and compared 4 different mosfets in the circuit: 2SK1548, IRFP450, 2SK1603, and 2SK5138. All behave substantially the same with the exception of the IRFP450 which has by far the longest turn-off time. My supplier does not stock the IRFPG50, and I would have to order a minimum of 10, and they are a bit expensive. If someone wants to send me one, PM me and I'll give address details.)

Regarding the 555 switch as opposed to the function's generator. There really should be no real difference between the two. However it is easier to adjust the 555 switch to enable that resonance which is both the object of the circuit and the main object of the thesis. Why you are not able to get the circuit into oscillation I do not know. Perhaps you must vary the frequency better. Incidentally I could not make out the frequency you applied on that demo.

The top instrument in the stack, with the little red numbers that say 2.404 or so, is a Fluke frequency counter, monitoring the output of the Interstate F34 sweep function generator. It is indicating the frequency of the pulse output of the FG. It is saying 2.404 kiloHertz. I agree that there SHOULD be no difference between the FG and the 555 output--but due to the duty cycle inversion, there is actually a HUGE difference. The 555 circuit as published in the Quantum article CANNOT be adjusted to provide a 3.7 percent ON duty cycle.

The niceties regarding the actual published switch and the one that you built - here I cannot comment. What I do know is that if the switch is set at 5% on and the load shows 5% on then it cannot default to 90% on. It is that simple. I could not make out the positioning of the probes in relation to the load resistor. Again. Your questions seem earnest - but your references not so easily detected on that video.

And yet, these "niceties" as you call them, call into question your entire experiment. And the probes are positioned and labelled exactly as in your diagrams in the Quantum article and the EIT pdf paper. The problem seems to be this: at point A, where the load is being monitored by the oscilloscope, when the voltage goes HIGH the load is actually OFF, that is, non-conducting. So if you are looking at this point with the Fluke-o-Scope set properly, it will report a 3.7 percent ON duty cycle--but in this case it means the load is OFF 3.7 percent of the time, allowing the voltage at point A to go HIGH.
Conmffsuing? Appraently.

In any event, the other problem I have is with the value of that spike which your referenced in the second video - I think it was. Our spike is generally far higher, upwards on 120v but is largely dependant on the applied duty cycle. In any event the amount of energy in the on cycle is always marginally more than the amount delivered by the spike. The value to the energy gain is in that this energy is repeatedly returned to the battery and to the load. This can be seen if you use 2 x 12 volt batteries as we did. If you run the test on the one battery and connect the second to the first with a common negative rail - then feed the diode to the positive of the second battery, you will see an immediate recharge to that second battery. That test was done to prove that the returning energy does, in fact recharge the energy source.

Again, experimenters here can tell you that pulse charging a lead-acid battery with HV pulses will cause the battery to indicate high no-load voltages even when its energy store is mostly depleted. I do not dispute that your circuit returns HV spikes back into the powering battery, so that this battery will indicate a no-load voltage that makes it seem to be fully charged. This is a well-known phenomenon around here.
As to the magnitude of the spike, it is here that the faster Fluke ScopeMeter does outperform my slower analog scopes at home. I have no doubt that the magnitude of the inductive spike is greater than what my 10 MHz scope could resolve.

So it is that we justify the value of the energy delivered by the battery as the sum of the on and off cycles. The energy dissipated is the product of both cycles. Therefore is there a gain. And at this fast resonating frequency the gain is really substantial.

And of course this entire claim depends on the correctness of the duty cycle numbers that went into your calculations. Nobody, under any circumstances, has been able to produce anything like the "really substantial" gains you have claimed--and I believe the reason is that your calculations are in error. Until we can resolve the issue of the actual duty cycle you used, this has to be the most likely reason.

I do hope this addresses those points that you repeatedly refer to through these threads

Not really.

Incidentally, TinselKoala - there's another point. We actually ran our tests with a control. The reason the published article and the paper deal with a test period of 10. something hours is because that is how long it takes the control battery to deplete its energy. For some reason, both in the quantum article and the paper I was specifically advised that any reference to battery duration was essentially irrelevant. Apparently battery draw down rates are subject to too many vagaries?

That's right. I have no issue with your control experiment. It is in line with my own. And you are right, batteries are difficult to measure in terms of energy content and "draw down rates" as you say. But with some understanding and the right equipment, useful measurements can be made.

In any event, the actual draw down rate of the tested batteries is consistent with the energy measured to be delivered by the battery as the difference between the energy measured and calculated from the two cycles of each waveform being above and below zero. At the end of that test period the test batteries are more or less the same as at the start of the test period. The control is entirely flat.

We then recharged both battery sets (always used typical 12 volt car batteries) and swapped the control with the test. Variations of this was called for by BP to enable their accreditation of the tests. It was exhaustive and painfully repetitive.

In the paper you say the batteries used were rated 12 volts, 20 Amp-hours. Most automotive batteries are rated in the hundreds of amp hours. So there is a discrepancy here. (EDIT: should read 80-100 Amp-hours for car batteries. Sorry. Still a discrepancy with Ainslie's report.)
And I gather you are attempting to measure the state of charge of your batteries by measuring the no-load voltage. Typically, a fully charged 12 volt lead-acid battery will read 13 or even 13.5 volts no-load. If it's only reading 12 volts no-load, it isn't charged fully. And in addition, if it has been subjected to HV spikes during charging it may indicate 13 volts even when its true charge state (energy content) is low. Only repetitive full-load high current discharge tests can truly indicate the energy content of a lead-acid battery.

CHET
PS
TK thank God you showed up ,or they wouldn't have had anything to talk about
You make us proud !!

Thanks, Chet, we shall see what we shall see. It seems from Rosemary's response that she may not be the electronics specialist in her research group. I hope that whoever was helping her can still reproduce the original circuit, or reconstruct it according to the directions they gave in the Quantum article and the EIT paper, because there are several issues that seem to need resolving...

" witsend  wrote:
Tinselkoala - I have no intention of answering any further posts. In truth I'm not sure that you wrote the last post as your standard of language is different to the previous. I think Gauss answered in your name. How do you do that? Do you share computers?"

So, there you have it. "My" 555 problem, according to her in another post, is "extraordinarily irrelevant".

Which says to me that she may have some cognitive difficulties--because the 555 problem completely invalidates the overunity claim, if the incorrect duty cycle figure goes into the calculations.

But she has refused, or dodged, all specific questions about the circuit.

She claims the batteries were car batteries--but the publications say they were 12 volt 20 Amp-hour rated--a far cry from a car battery around here.

She claims resonance and aperiodic oscillations--simultaneously.

She apparently does not understand that instrument readings sometimes are inaccurate, and some of her reported findings sound exactly like instrument artifacts, that I have reproduced in her circuit with my Fluke 199 ScopeMeter. But no scope traces from her experiments, showing the "aperiodic resonance",  seem to exist. And, of course, NOBODY using a short duty cycle has seen heating of their load, nor have there been any reports of resonance or especially simultaneous aperiodic oscillations with resonance.

Now, I've tried to play nice here. Anybody reading my posts, who is familiar with me, like 0c, can tell you that I am being quite polite and restrained, compared to my usual response to bullshit. But seeing the flack I am getting, from people who cannot even be arsed to assemble seven dollars worth of components to see how they behave...that's starting to piss me off.

And I am coming to the conclusion that Rosemary Ainslie is another one of a type that we know all too well. She's got a theory, and refuses to be distracted by facts, and whoever challenges the theory, based on correct experimentation, is by definition part of the suppression conspiracy and is evil. Never mind the new experimentation that shows that the original data from which the theory was derived is wrong... it is extraordinarily irrelevant.