PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-05-21, 01:34:48
News: Forum TIP:
The SHOUT BOX deletes messages after 3 hours. It is NOT meant to have lengthy conversations in. Use the Chat feature instead.

Pages: 1 2 3 4 5 6 7 8 9 10 11 [12] 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 ... 29
Author Topic: The Rosemary Ainslie Circuit  (Read 460628 times)
Group: Guest
@feynman

For a guy that puts down my sim because I might not have included the exact molecular structure of Rose's load resistor, it's kind of odd to hear that you would replace the whole affair with an arbitrary length of nichrome without regard to precise dimensions, turns, resistance or inductance. I clearly stated that my model has the same RLC as Rosemary reports she uses.  What have I left out that you consider so possibly essential for a proper model?  Zipon content and dipole heat velocity factors?  Doesn't seem quite consistent but what else is new around here.

I'm glad that Stefan's heavy mod act thrilled you so much.  Sounds like your Saturday nights may be just as drab as you suspect mine are.  You sound positively giddy over it.  Personally, I was amazed that he let 7 of my posts stand and did not ban me outright (seeing as how he perma-banned me four years ago and I made no bones or pretense that I was not one and the same as Humbugger).  

Incidently, I waas banned over there because I busted Ashtweth using edited quotations from a guy named David Kou who was, according to Ashtweth, successfully using his "Neon Extractor" to turn reactive power into real power and thereby get self-running operation from an RV setup.  I searched out and located the actual quote from David on another board and the words were just as Ashtweth reported except that Ashtweth chopped out the last sentence where Mr. Kou happily stated that he had not yet even tried to acheive self-running and was still far from it.  That was years ago and there is still no self-running RV, of course. Publishing Ashtweth's outright lies got me banned.

As for the rest of your crap and opinions about me personally, you can shove it, pal.  And we can still be civil just like real scientists in any future discussions.   O0  (starting after I finish this totally straw-man ad-hominem attack post, of course)   >:-)

Just realize that your disrespect and loathing of me is entirely symmetrical, although I don't probably care as much as you do.  I have explained my tendency toward arrogance when confronted with idiots claiming world saviourism before showing anything beyond totally ordinary lossy circuits and machines.  They love to hide behind mysterious bullshit; I love to pierce it incisively until they are naked, if not bleeding.  In Black Swan terms. I'd rather be the butcher than the turkey, if you catch my drift.  (By the way, thanks for the tip on the book.  I have not read it yet and will do so soon),  Nobody pays me to do what I do here or on any forum..  I'd take the job in a New York minute, though, if the money was good.

Humbugger
« Last Edit: 2011-03-20, 13:23:37 by humbugger »
   
Group: Guest
For general interest regarding inter-board politics and Stefan's moderation (spelled DELETION OF 8 POSTS of mine and several of Mark Dansie's),  I made one final post over there as Cheeseburger tonight and it went like this:
Quote

It is clear that Stefan has the right and the might to heavily moderate this forum and I respect that.  Whether I think his judgements are fair or not is not important and I appreciate that he left intact seven of my fifteen posts.  I assume that was because they were on-subject, without insult and quoted in other posts anyway.  Thank you Stefan for leaving at least some of my comments and suggestions.

Apparently, at least Mark Dansie and I are on "heavy moderation" at this time and Stefan has suggested that someone (I assume he means me) is a paid detractor sent by big oil or the MIB to destroy Rosemary's progress.  That is simply not true, I can assure you.  However, each of us has a right to hold and express our opinion; especially Stefan, as it is his blood sweat and finances that have built this forum.

That said, I will not be posting here again because I do not feel that my comments are welcome, despite their often acute and timely relevance.  As my final post here, I would ask and hope that Stefan have the sense of fairness and good sportsmanship to clear this for posting.

For anyone who is interested, I have tonight published a very revealing set of simulations under the Rosemary Ainslie thread at OUR forum.  I have discovered a couple of things that may be of interest even to those of you who put no faith in simulations.

First, the Ainslie circuit with five MOSFETs models quite easily and shows exactly the same waveforms and performance that Rosemary shows in her lengthy series of 'scope shots on her blogs when all of the various wiring inductances are included.  This indicates strongly that all circuit behaviors she has pointed out and shown are entirely explainable using only classical circuit models.

Second, I have shown that one single input step-function pulse of 1 microsecond duration will set the circuit into continuous oscillation.  Simply setting a tuned DC bias at the gate may or may not set the circuit into oscillation, but a single sharp transition past the threshold definitely does when the gate is subsequently held at zero or negative DC.

There are no ground currents involved.  The circuit oscillates continuously at between 1 and 1.4 MHz, depending on specific MOSFET types and inductance values in the range consistent with the wire lengths and load inductances as stated by Rosemary.

I think these facts and the ability in the simulation to probe inside the RL lumped components (the shunt, the load and the battery stack) to look at the waveforms as they exist across just the battery, just the resitive parts of the load and just the resistive part of the shunt to see the true current and voltages there are useful to anyone planning to replicate and or simulate the circuit for further study.

I will be doing some more sims to include actual measurements of the load heating power and the battery input power in the near future.  If these tests show anything like or close to overunity, I will proceed to an actual hardware replication to verify the results.  This work will be shown at OUR exclusively unless Stefan invites me to also show it here.

Thank you for your consideration of these sincere efforts to gain a fuller understanding of this circuit and its performance.  So far, the model tracks and agrees very well with all of what Rosemary has reported based on her waveform analysis.

Of final and very significant interest is the fact that the waveforms representing the shunt voltage change dramatically when measured to include or exclude the omnipresent inductance of the physical shunt.  

Yes, as Rose has reported. it appears that the current waveform has nearly identical areas above and below zero when the shunt inductance is allowed in the measurement and that the apparent current levels are many times higher than those measured across the purely resistive portion of the shunt impedance.  

However, when we look only across the true resistive portion of the shunt (leaving the inductance in the circuit so that operation is unaffected), it becomes clear that the real area under the current waveform is quite a bit larger above zero (coming out of the battery) than the current being returned.

cHeeseburger

It will be interesting (at least to me, anyway) to see if Stefan clears it for posting.   C.C
« Last Edit: 2011-03-20, 09:17:23 by humbugger »
   
Group: Guest
Rosemary has now posted a challenge (now that she has seen my simulation here and knows I won't probably be allowed to accept her challenge under Stefan's heavy moderation).

Here is her post at OU.com

Quote
Guys a quick word here.  I challenge anyone to do a simulation where the phase angles are at precisely 180 degrees.  If they are not precise then, as day follows night - they'll ring and cancel out at zero.  At 180 degrees they'll reinforce each other.

What intrigues me is this.  I've been called delinquent - ignorant - presumptuous - pretentious - deluded - self-serving - deceptive - optimistic - fanciful - manipulative.  Name it.  It's there.  In fact I've also been called an IDIOT SAVANT.  Poynty?  I have a really functional intelligence quotient.  Surely that much is evident?  And right now Hamburger is throwing a hissy fit with MileHigh as his praise singer.  So what?  What if I am all of those things?  What has it got to do with the issue at hand?  We are only showing results.  And those results are hugely promising.  But they unquestionably need research.  We're all, on our team - ONLY anxious that this get researched.  If any out there need to take the credit - FEEL FREE.  Just DON'T try and patent it.  I'll contest that to my dying breath and with my last cent.

Kindest as ever,
Rosemary

First let me say that I for one have absolutely no interest in any intellectual property that has ever spawned forth from Rosemary, so she needn't worry on that account at all.  Second, I dispute that Rosemary or her team of experts has shown any results other than the same very results I demonstrated in a thirty minute simulator exercise, which match her results quite closely.  

Furthermore, my simulation clearly shows that her measurements "results" which consist of polluted and inaccurate representations of both the battery voltage and the shunt voltage are not worthy of basing any further investigation of input power upon unless and until they are corrected to remove the artifacts and gross waveform distrortions and exaggerated amplitudes caused by the dominant inductances in the battery wiring and the shunt.

She herself has not demonstrated any waveforms that are clearly "precisiely 180 degrees out of phase" and, despite her insistance that that would be surprising. essential  or meaningful, it is not.  What is important are two things:

The current flowing into and out of the battery and the DC voltage of the battery.  These and only these will reveal the true input power being supplied by the battery.  Rosemary's obsession with the phases of the "battery" voltage and drain current are based upon the absolute fallacy that the battery voltage is 150VAC and the shunt current waveform is accurate in both time and amplitude despite it being taken across an element that has several times the inductive reactance as it has resistance.  Many times more.  My simulations clearly show this in an unequivocal way.

In truth, the battery voltage is virtually a constant DC voltage with less than a volt of AC ripple and the true resistive shunt voltage is far smaller and differently-time-aligned and differently shaped and differently centered about zero than the illusionary inductive shunt waveforms indicate.  So any statement regarding perfect phase opposition based on two incorrectly-represented critical variables is also bogus.  Not to mention that her own waveforms do not meet that criterion.

In general, it is absolutely to be expected and totally unsurprising that the basic drain voltage and drain current waveforms are by nature 180 degrees out of phase in a MOSFET circuit, whether operating linearly or as a true switch.  As the drain current goes up, the drain voltage pulls down.  This is obvious to anyone half-skilled in the art and is not evidence of a magical or special or unique situation or result.

Feynman and Harti have now taken over the thread and are making a barrage of demands for much more detailed documentation, scope shots, build specs, etc.  Rosemary is frantically trying to appease and accommodate these demands while at once cleaving to the false comfort of being sheltered from my revelations here by these two whose obvious agenda is to prolong the saga of Rosemary Ainslie's magical zipon technology for as long as possible and to help her continue the clouds of dark energy ink that the Octopus so loves to hide within when attacked.

The mystery, however, has been solved.  But you know what?  Folks like Stefan, Aaron, Lindemann, Bedini and Sterling make their livings by endlessly prolonging and convoluting and digging up new examples of these false claims of overunity inventors.  They could not survive if the light of day were shone brightly into each and every corner of false "promise" for free energy such that all could quickly see through the bullshit and delusions.

Talk about paid disinformation agents!  These guys do this shit for a living and have figured out how to play each episode out for as long as possible in front of the gullible believers.  I am completely convinced of that,  These guys know that 99% of this stuff is obvious bullshit yet they promote and prolong it on purpose to make a buck.  And when anyone sharp comes along and quickly shows the obvious errors and fallacies, they turn aqs one against them, banning them and accusing them of being paid misinformation artists,  Dat's da game, here folks!  And this forum is probably the only exception.

Idiot Savant?  Maybe...not too sure about the Savant part unless it's in regard to her near perfect ability to cloud, distract, avoid and obfuscate.  All is as it should be in Heaven.  The true distorters for financial gain are teamed with the idiot savant of the decade.  Don't expect any free energy any tme soon, folks!  Not from that motley crew of charlatans and pretenders.

Humbugger
« Last Edit: 2011-03-20, 12:00:32 by humbugger »
   
Group: Guest
Okay...back to simple clear reality again:

The first schematic shows the battery wiring inductances lumped into one inductor equal in resistance and inductance to the numerous (7) distributed inductances previously shown so that we can finally take a look at the real battery voltage.  Nothing else has changed and you may note that the waveforms are exactly the same as the first sim where the same total inductance is distributed.  The "battery voltage" is probed on the wrong side of the llumped inductance (not on the battery terminals), just as it was in the first sim and, by golly by gosh, the scope waveforms are exactly the same.  

Yes Martha, you can lump or distribute series inductances in a model or in hardware and it makes zero effect on the sim or on a real circuit.  It's just that it's trivial to do on a sim and a lot harder to do in real physical life, but one can come very close and then, as needed, simply use RC filters (not done here as yet due to potential whining) to produce accurate measurements in the face of real world distributed inductances that cannot all be moved into one "lumped" location wherever you want.

The sole purpose of this elementary post is to demonstrate that lumping the battery wiring inductance in the model does not change the circuit waveforms in any way.  The shunt trace, as in the first set of sim shots, is hooked up to include the spurious effects of including the shunt inductance.


« Last Edit: 2011-03-20, 12:30:50 by humbugger »
   
Group: Guest
Now, we move the battery voltage trace probe so that it is actually and truly reporting/recording/sampling the real battery voltage.  Notice that it is essentially, as completely expected and predicted, a DC voltage centered at 75.6VDC with only a small ripple.  I actually increased the modelled internal resistance of the batteries here twofold and thus the ripple is a couple of volts.  NOT 150Vpp AC.

Furthermore, to complete the task, I have moved the shunt probe now to measure the true current waveform appearing across the purely resistive part of the shunt while still keeping the shunt inductance in the circuit so as not to change its operating point.

Now, and only now, can we begin to think about making actual input power measurements that will provide real-world true numbers.  To make it easier to study, both wavefprms have been set for zero offset.  The center line is zero volts.

The DC battery voltage scale is still set at 20V/division and the current shunt voltage scale is gained up to 100mV/division.  Now it is absolutely clear that the area under the positive portion of the shunt voltage is far lqarger than the negative area.

Yes, the picture is worth ten million words.  A group of hungry current sharks are leaping up to bite the hell out of the battery.  They got tired of eating zipons, apparently.
   
Group: Guest
Stay tuned for tomorrow's episode, where we actually measure the input and output power using multiple approaches and see how it all turns out.  It's already quite clear that the battery is continually giving up net energy to the circuit and load.  The currents and power levels are, I expect, fairly low and so I will not be surprised at all if the circuit shows efficiencies easily upward of 95%.  

For this level of power output and input (just roughly estimating from the waveforms and using head-math) the five MOSFETs in parallel are waaaay overkill and represent, with the shunt itself, the only really significant loss factors.  It certainly appears that the MOSFETs are running in the linear region and not acting as hard switching devices, so my 95%+ estimate may be quite optimistic.

Tomorrow we'll peek at the voltage at the drain of the MOSFETs and see if they are hard switching (typically 90% or better efficiency) or running class A linear (50~75% efficiency typical) or somewhere in between.  We'll do that first and I'll reiterate my overall "guesstimate" of efficiency before we do a few tests using true multiplying power meters and plain old voltmeters/ammeters with RC integrating low-pass filters to see how the various approaches compare.


Now I really have to go eat something, get another pack of cloves and take a nap.  At last check there were 16 folks on the forum and 9 of them were reading this thread.  It's 5:22 AM here on Sunday morning.  West Coast Black Ops Headquarters time  :o
[being sarcastic and facetious here...don't get all snippy and paranoid now].

Yours truly,

Humbugger
   

Group: Administrator
Hero Member
*****

Posts: 3217
It's not as complicated as it may seem...
Good stuff Hum.  ;)

Is it possible to place component values on your schematics? I know you've denoted them in your post text, but it is much easier to get the "picture" when component values are shown.

Can you limit the transient analysis step-size in Tina? I noticed in all your scope shots except the last one with the true Vbat, that the traces are showing a bit of "under-sampling" distortion.

I've been busy working on the Lewin experiment stuff, but my last try with the Ainslie  sim was producing something close to Rose's wave forms. I have not included any inductance in the CSR though, and that perhaps may be the "problem". Maybe I'll try to get it going, and we can see how mine compares with yours. ;)

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   

Group: Administrator
Hero Member
*****

Posts: 3217
It's not as complicated as it may seem...
Hum,

What is your value for R2? I noticed in my sim, that if R2 was relatively low, i.e. about 0.5 Ohms, that the self-oscillation appeared, but when changed to 50 Ohm, the oscillations became highly damped and it would not sustain. We don't know what the function generator model is, so it is impossible to say what the output impedance truly is, although it would be surprising to me if it was anything but 50 Ohms.

Just a note. ;)

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   

Group: Administrator
Hero Member
*****

Posts: 3217
It's not as complicated as it may seem...
Hum,

So it would seem that your last post at OU (copied here) did not last, i.e. Stefan removed it.

Oh, and Rose, the "idiot savant" reference I made was not intended for or directed at you. You thinking it was is rather pretentious don't you think?

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   
Group: Guest
Poynt,

Too tedious to label every part, doable but tedious.  No control over sample rate on scope but i see no evidence of undersampling anywhere on any trace.  Gate resistor was 0.5 ohms, inductance was around 1uH based on long non-50 ohm twisted pair cabling to gates.  OU post to stefan from Rose has siggen at 50 ohms.  Funny thing is, it started oscillating heavily on the very first turn-on of the model.  I did no fiddling at all.  No tweaking to cherry-pick waveforms, etc.  I'm conviinced it would be hard to stop it from oscillating save by turning the MOSFETs on hard.  Fooled around with a few variables and didn't see a whole big bunch of difference .

Gotta go now.  I see AC and Exnihiloest are posting probably severe critiques.  Have to look tomorrow...burnt out...gotta eat and sleep.   :D

Cheers,

Bryan
   
Hero Member
*****

Posts: 2664
@Feynman
Quote
This seems like an awful lot of work, and I think its fascinating you've done this, but why go through so much trouble just to disprove her?
I disagree with your claim the 'simulation' is reality... it's just a useful tool, not the ultimate measure of reality.  The reality is on the bench.  

I would agree, I have always found it odd that many would seem to go to almost any length to disprove another persons work or discredit them personally to justify their own opinions. You would think they would have more constructive things to do with their time and expertise such as actually proving how a real device could work. I find it kind of comical that in the end, after all is said and done, that the critics will end up right back where they started which is nowhere, no working technology and no idea how a FE technology could work which would seem rather pointless in my opinion.
On the issue of simulators, I run a few simulators for simple standard technology circuits but it should be obvious that they are useless for anything more complicated than that. First simulators only model lumped sum inductance and capacitance which is the last thing I want, they do not accurately model external magnetic or electric field interactions between each or all of the components and the circuit as a whole, they cannot model unknown non-linear material properties, they cannot model the electric field emanating  from a spark gap nor the hard UV wavelengths producing thermionic and photoelectric effects. Essentially they are completely useless for all the things I need to know, they are great for beginners but for more advanced concepts which rely on complex field interactions, radiant effects and not simple conduction they are a complete waste of time and energy in my opinion. Here is a thought, I have long solenoid coil which is induced by a two turn primary and when I disruptively close the circuit on the primary, before a measurable current flow is detected in the primary a HV spike evolves from it inducing a higher voltage in the secondary. We may as well stop here because it should be perfectly obvious that no simulator will show you this transient voltage spike from the primary simply because the programmer of the simulator has no knowledge of this effect. So you see using a simulator in effect and practise holds a person to the knowledge base of the person who designed the software which pretty much ensures anyone using a simulator will never learn anything new or anything that really matters.
I'm not saying any of this directly applies to Rosemary's device, I am simply stating that if anything out of the ordinary was happening a simulator would not show you because a simulator is only as good as the person that programmed it.

Edit: I should also say I am not against criticism nor the critics, I disagree with people who make irrational one sided arguments not based solely on logic and science, this includes not only the critics but everyonwe else as well.
Regards
AC


---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   
Group: Guest
...
As Stefan says,

Quote
[Moderation] is due to the facts that now more and more paid "twisters" are coming onto this forum, who want to suppress free energy as there are now viable solutions.
...

Stefan is not a reference. In a totally biased attitude, he urges the skeptics to prove their affirmations against unfounded FE claims, or ban them, while authorizing  gullibles and scammers to assert the worst stupidities outside of any evidence.

As there are always many more gullibles, scammers, gurus with followers and so on, than skilled and skeptics people, it is better for him and his site to favour the first ones, question of popularity and profitability of the advertisements. Therefore when he is talking about "paid "twisters"", a legitimate question arises: isn't it his own attitude that he projects onto others? Imho, likely.

   

Group: Administrator
Hero Member
*****

Posts: 3217
It's not as complicated as it may seem...
AC, I must disagree with most of your post. Big surprise huh?

You are entitled to your opinion, and I am to mine. You said that SPICE has been useless to you, and that is fine. However, I must protest to your comments that SPICE is only as good as the programmer that designed it. That is just plain incorrect. That is analogous to saying that a programming language is only as good as the programmer that designed it. Yes, they both have their limits to what they can or can not do, but just as in software programming, the end result of your program or simulation model, is only as good as the programmer/SPICE user using the program.

In the case of SPICE, it is capable of more than what most people realize, including yourself. SPICE is a tool box and a simulation engine. It is not an AI program that reads your mind and enters the schematic accordingly with all the hidden effects built in. YOU, the user must know precisely what it is that you are simulating, and be aware of ALL the subtleties involved, including all the possible parasitics.

Your pre-current spike is one good example. I have seen this in my sims many times. I do not have an example at hand at the moment, but I will see if I can produce one for you.

Give it a break man. Perhaps you should consider the possibility that SPICE has been useless to you because you expect it to do all the thinking for you? Well, a reality check; that's not how it works.

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   
Group: Guest
...
I have explained my tendency toward arrogance when confronted with idiots claiming world saviourism before showing anything beyond totally ordinary lossy circuits and machines.  They love to hide behind mysterious bullshit; I love to pierce it incisively until they are naked, if not bleeding.
...

Well there is at least one person here who is delighted in advance by the show. I'm also a bit involved in this practice :)

   
Group: Guest
...
You are entitled to your opinion, and I am to mine. You said that SPICE has been useless to you, and that is fine. However, I must protest to your comments that SPICE is only as good as the programmer that designed it. That is just plain incorrect.
...

I totally agree. I'm a user of ltspice, and if properly done, a spice model is of great help in debugging experiments with curious results.
The difficulty is to not forget all the parameters of the "real life".

For example, for realizing the spice model of the Stiffler circuit with coils and LED that is discussed elsewhere, we must add parasitic capacities along the coil, i.e. the model for this single coil must be either a line with a distributed capacity or many coils, each with a capacity.

   

Group: Administrator
Hero Member
*****

Posts: 3217
It's not as complicated as it may seem...
Poynt,

Too tedious to label every part, doable but tedious.  No control over sample rate on scope but i see no evidence of undersampling anywhere on any trace.  Gate resistor was 0.5 ohms, inductance was around 1uH based on long non-50 ohm twisted pair cabling to gates.  OU post to stefan from Rose has siggen at 50 ohms.  Funny thing is, it started oscillating heavily on the very first turn-on of the model.  I did no fiddling at all.  No tweaking to cherry-pick waveforms, etc.  I'm conviinced it would be hard to stop it from oscillating save by turning the MOSFETs on hard.  Fooled around with a few variables and didn't see a whole big bunch of difference .

Gotta go now.  I see AC and Exnihiloest are posting probably severe critiques.  Have to look tomorrow...burnt out...gotta eat and sleep.   :D

Cheers,

Bryan

Yes, the ISOTECH GFG - 8216a. Apparently 50 Ohms. To be faithful to the actual circuit, I would suggest you change your R2 to 50 Ohms then.


I label every part on my schematics. It's part of the part placement process actually.

How or where do you enter the part values then? It is most likely a setting in your options to show the part value on the schematic, and it is currently "turned-off". You should not have to add the values again because you already have entered them. Please have a look at your options, there has to be a setting for this.

What version of Tina are you using? Here you go Hum:

Quote
TINA will automatically assign a label for each component you place
on the schematic. It will also display the numerical value of the main
component parameter (for example: R4 10k). Note that the value is
shown only if the Values option of the View menu is checked. For
files from the older versions of TINA, the Values option is turned off
by default.

Regarding the time-step, indeed it appears that Tina does not allow you to set this, how unfortunate. Anyway, it is not that critical for this circuit anyway, as there are no transients present in self-oscillation mode, and if there are, I would hope that the sim engine adjusts it's time step accordingly.

At least we got your displayed component value issue sorted out.  O0

Cheers,
.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   
Group: Guest
Yes about pSpice and other simulation packages, it's all about how well you do your modeling, and to what level of detail you want to do the modeling.  I think it's fair to say that lots of people on the forums can't understand the underlying issues here.  Simulation is just a Black Box that they don't understand.

I can't resist quoting Feynman from OU and OUR considering that he offered to prototype the Ainslie circuit and resolve the matter:

Quote
I can go build this on the bench in a couple of hours... I might just do that to settle this stupid diversion...

Quote
The problem I'm getting is that the output of pin 3 is completely inverted from the output as measured at the MOSFET drain.

For example, if my input is PWM with 1% duty cycle and a specific frequency, on the drain of the MOSFET I see a 99% duty cycle wave at the same frequency.

Quote
Okay , got it.   Does a P-channel MOSFET have the same inverting effect when driving the low side?   I know these are probably obvious questions , but I normally deal with microcontrollers and sensors.   This is my first time actually doing serious benching with MOSFETs.

LOL  Any long-timers in the Ainslie affair will recognize shades of Joit in that second quote!

I am sure that we would be more than happy to help you here also Feynman, just be real.

MileHigh
   
Group: Guest
@Feynman
I would agree, I have always found it odd that many would seem to go to almost any length to disprove another persons work or discredit them personally to justify their own opinions. You would think they would have more constructive things to do with their time and expertise such as actually proving how a real device could work. I find it kind of comical that in the end, after all is said and done, that the critics will end up right back where they started which is nowhere, no working technology and no idea how a FE technology could work which would seem rather pointless in my opinion.
On the issue of simulators, I run a few simulators for simple standard technology circuits but it should be obvious that they are useless for anything more complicated than that. First simulators only model lumped sum inductance and capacitance which is the last thing I want, they do not accurately model external magnetic or electric field interactions between each or all of the components and the circuit as a whole, they cannot model unknown non-linear material properties, they cannot model the electric field emanating  from a spark gap nor the hard UV wavelengths producing thermionic and photoelectric effects. Essentially they are completely useless for all the things I need to know, they are great for beginners but for more advanced concepts which rely on complex field interactions, radiant effects and not simple conduction they are a complete waste of time and energy in my opinion. Here is a thought, I have long solenoid coil which is induced by a two turn primary and when I disruptively close the circuit on the primary, before a measurable current flow is detected in the primary a HV spike evolves from it inducing a higher voltage in the secondary. We may as well stop here because it should be perfectly obvious that no simulator will show you this transient voltage spike from the primary simply because the programmer of the simulator has no knowledge of this effect. So you see using a simulator in effect and practise holds a person to the knowledge base of the person who designed the software which pretty much ensures anyone using a simulator will never learn anything new or anything that really matters.
I'm not saying any of this directly applies to Rosemary's device, I am simply stating that if anything out of the ordinary was happening a simulator would not show you because a simulator is only as good as the person that programmed it.

Edit: I should also say I am not against criticism nor the critics, I disagree with people who make irrational one sided arguments not based solely on logic and science, this includes not only the critics but everyonwe else as well.
Regards
AC

Reminds me of the old saying about March:  In like a lion, out like a lamb.  Ninety six lines about personality questions and simulator dissing ended with a couple of lines of timid qualifying remarks.  Yahooooooo!  Ever heard of a paragraph?  Makes it really hard for us old stodgy farts who know nothing of creativity or open-mindedness to read your stuff there AC  O0

Humbugger
   
Group: Guest
...


Stefan is not a reference. In a totally biased attitude, he urges the skeptics to prove their affirmations against unfounded FE claims, or ban them, while authorizing  gullibles and scammers to assert the worst stupidities outside of any evidence.

As there are always many more gullibles, scammers, gurus with followers and so on, than skilled and skeptics people, it is better for him and his site to favour the first ones, question of popularity and profitability of the advertisements. Therefore when he is talking about "paid "twisters"", a legitimate question arises: isn't it his own attitude that he projects onto others? Imho, likely.



Right on the money, sonny!   O0
   
Group: Guest
Yes, the ISOTECH GFG - 8216a. Apparently 50 Ohms. To be faithful to the actual circuit, I would suggest you change your R2 to 50 Ohms then.


I label every part on my schematics. It's part of the part placement process actually.

How or where do you enter the part values then? It is most likely a setting in your options to show the part value on the schematic, and it is currently "turned-off". You should not have to add the values again because you already have entered them. Please have a look at your options, there has to be a setting for this.

What version of Tina are you using? Here you go Hum:

Regarding the time-step, indeed it appears that Tina does not allow you to set this, how unfortunate. Anyway, it is not that critical for this circuit anyway, as there are no transients present in self-oscillation mode, and if there are, I would hope that the sim engine adjusts it's time step accordingly.

At least we got your displayed component value issue sorted out.  O0

Cheers,
.99


Yes, thanks Poynt.  I told you I almost never use sims and Tina is a real basic kind of slow dog, but I find it to be pretty close to accurate for most of what I use it for.  I'' try to remember to turn on the value thing next time.

The Ainslee circuit is rather triivial except for the unknowns like exact wire lengths etc.  I would take youir advice and insert the 50 ohms if I knew the exact details of the homebrew twisted pair transmission line she's using to drive the gates.  The 50 ohms has to go on the generator end, of course, unless the transmission line is 50 ohms also, which it isn't.  It does adjust the step and I didn't notice any sampling artifacts at all.

The real doggy thing about Tina Pro (really old version) is that it is so far from real time that it's sick.  At 1MHz, the waves take about 1 second each to compute and display.  No big deal until you start adding long RC time constant filters.  I had to wait 20 minutes to get an averaged DC reading for the shunt current signal!

Humbugger
   
Hero Member
*****

Posts: 2664
@Poynt99
Quote
In the case of SPICE, it is capable of more than what most people realize, including yourself. SPICE is a tool box and a simulation engine. It is not an AI program that reads your mind and enters the schematic accordingly with all the hidden effects built in. YOU, the user must know precisely what it is that you are simulating, and be aware of ALL the subtleties involved, including all the possible parasitics.
I think you may have just agreed with me :) or not.
Quote
YOU, the user must know precisely what it is that you are simulating, and be aware of ALL the subtleties involved, including all the possible parasitics
What if you do not know precisely what it is that you are simulating or ALL the subtleties involved?, what if we missed one single subtlety?, then we are for the most part simulating something based solely on the things we know not things which are unknown to us. To me this is akin to trying to simulate a UFO when the user inputting the parameters has no knowledge of how the damn thing stays in the air, I believe this is where science and real experiments come into play.
You know I have programmed, built and interfaced with computors ever since the day I bought my first Vic 20 a "few" years ago, lol, I designed and built computor interface boards, microcontrollers, robotics, power supplies and such--nerd stuff. In all that time I have seen no magic, no computors filling in my gaps (that one comma out of place in the 6000 lines of code I had written), in fact I wrote all the code on a gas turbine simulator at one time because no simulators would work for one of my projects. In all that time I have never seen a computor do anything truely unexpected because the software does what it is supposed to following the sequence of logic available to it. Computors do not fill in the gaps nor do they tolerate mistakes, they do what they do --compute--, which raises the ultimate question, if we fully understood all aspects of what we are trying simulate then why would we simulate it?. If we know precisely what it is we are simulating and all the subtleties involved then I would think we would simply be verifying what we already know in many respects. I understand your point and it is a very good one however I do not fully agree with all aspects of it which is no surprise, lol, usually when a computor gives me an answer I did not really expect I start the debugger. :D
Regards
AC


---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   
Group: Guest
Here is one last little presentation for those who just can't wait to hear whether the sim showed overunity or not.  I'm sure the suspense is killing everyone  C.C

Now, mind you, this is not intended to be in any way absolute or accurate to within 0.01% tolerance.  I'll do some more methods maybe later on (still no sleep yet but I had a steak and half a bottle of Smoking Loon Merlot (not to bad for $12)).

All I am giving here is a simple close approximation based on the actual waveform and voltage across the resistive part of the load versus the RC-filtered current as measured at the resistive part of the shunt.

I finally found the actual models for the IRFPG50 MOSFETs and plugged that in and tweaked a very little bit here and there to adjust the amplitude of the load waveform to be 60Vpp.  The main sensitivity I found was to the sum of the battery lead inductances.  As that varied over +/- 20% or so, I could adjust the amplitude easier there than anywhere else.  

That said, and having essentially left all else the same except the MOSFETs, the output wave looks quite like a sinusoid, although it certainly isn't a pure one...slightly tilted.

That said, and only trying for a close estimate, I simply used the standard sine calculations 60Vpp/2.828 to obtain RMS of 21V which, when squared over 11.11 Ohms (Rose's load resistor value sans inductance) comes out to 40Watts, the number she so often references as being the measured amount of heat in the load.

The battery voltage is 75.6VDC averaged (not shown but diligently measured over half an hour with a modest RC integrator on it and only a very mild ripple to begin with.

So, with 40Watts being pumped out as heat in the load, the RC-filtered voltage and current product was positive.

175mV average on the 250milli-Ohm shunt resistance yields 0.7A average net current drain from the batteries.  Muitiply that by 75.6 Volts and you get the picture of a net input power of 53 Watts. for a net efficiency of 40/53 or 75.5%.

The circuit is operating primarily in a linear mode between class A and class C, by that figure of merit, which is just about as I expected.  Here are the pictures:

Good night (morning...almost afternoon now...damn)

Humbugger
   
Hero Member
*****

Posts: 2664
@Humbugger
Quote
Reminds me of the old saying about March:  In like a lion, out like a lamb.  Ninety six lines about personality questions and simulator dissing ended with a couple of lines of timid qualifying remarks.  Yahooooooo!  Ever heard of a paragraph?  Makes it really hard for us old stodgy farts who know nothing of creativity or open-mindedness to read your stuff there AC  

I figured I would write everything without paragraphs to keep the old stodgy farts from reading my posts, obviously it isn't working out all that well. I don't think being a stodgy old fart means we lack creativity or open-mindedness it is more a matter of immaturity, you see I am really immature, so much that everyone including my kids are always telling me to grow up. Immaturity is like the fountain of youth ;)
Regards
AC


---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   
Group: Guest
Oh...before I give up the ghost for that long-awaited nap, I have two speculations (purely educated guesses) to add:

1)  When Rosmary talks about "that super-nova" mode that she fears entering and lingering in too long, I suspect that what is happeneing is that one or more of the MOSFETs gets turned fully on, placing the DC battery voltage of 75.6V directly on the load and shunt and RDSon path.  That would be just about 500W.

2)  The MOSFETS she's using probably are not matched in terms of VGS threshold and temperature and, while MOSFETs are known for a positive (non-thermal-runaway) aspect in their VGS/temp curves, this is not always the case at lower levels of current and may be backward down there.  Even at 500W, each MOSFET's share of current is only (500/75)/5 Amps or 1.333Amps each.

Come to think of it, maybe it happens just once every 2.7 minutes when the funkygen goes high!   >:-)

Okay..that's all for now and thanks to all for all the comments, good, bad, encouraging, critical or whatever.  It's all good stuff and helpful.  Cheers!

Humbugger

@AC  Well...we have something in common after all!  O0
   

Group: Administrator
Hero Member
*****

Posts: 3217
It's not as complicated as it may seem...
AC,

Stop expecting SPICE to make allowances for your oversights. If the results are bad, it is because YOU caused them, not the SPICE program. Know the limits of the program, and don't expect it to think for you. If you can not think of all the variables, then don't use it, but stop putting the program down.

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   
Pages: 1 2 3 4 5 6 7 8 9 10 11 [12] 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 ... 29
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-05-21, 01:34:48