In my previous thread on ISLab's Replication of Basic POC Effect, I had many useful learning outcomes and got a good "feel" for the behaviour of POC coils under various kinds of situations. With help and encouragement from all of you, I was able to get the Sawtooth Wave and "generate" more power than "induced" within the coils.
This thread will focus on optimising the POC coils and maximising their magnetic resonance for optimum power generation.
In this thread I will only be using large C-cores. I had already wound coils for this back in February. These are not optimal as my understanding has grown much since then. But I will use these for an initial study of magnetic resonance before winding better coils as a subsequent step.
What follows below is the study of magnetic resonance in these non-optimal coils.
The coils are:
L1: 29.4uH (in air), 2.6mH (on core), 0.3 ohm, 20 SWG, 24 turns, 378cm, CW.
L2: 0.376mH (in air), 49.4mH (on core), 0.6 ohm, 20 SWG, 104 turns, 1470cm, CW.
L3: 0.372mH (in air), 49.4mH (on core), 0.6 ohm, 20 SWG, 104 turns, 1461cm CW flipped.
Testing them for magnetic coupling coefficient shows:
KL1-L2 = 0.994
KL1-L3 = 0.763
KL2-L3 = 0.757
The cores with coils set up for measurement of magnetic resonance:
I decided to test magnetic resonance in all possible combinations just to see if it made a difference. And it did!
With a core separation gap of about 1.1mm there were two peaks of resonance, one around 185KHz and one around 500KHz.
Reducing gap to 0.1mm also gave two peaks but both became much narrower, rising suddenly and dropping suddenly within ± 2Khz. But it did not change the peak voltages by much when at correct frequency.
Below is the table of measurements for 0.1mm gap:
L2 IN and L3 IN are where the coil windings begin on the first layer of the coils, and L2 OUT and L2 OUT are where the wires exit the coils after winding the topmost layer. All layers are wound asymmetrically in a single direction left to right, then straight back, and again left to right, etc. L3 is flipped and exactly mirror image of L2 when placed on the cores.
The variations of F between measurements seem to be due to slight movement of coils on the core due to wire tension, showing how critical it is to completely lock down the wires and coils on the core. When hitting resonance, there is a significant reduction in peak by nearly 50% with F just 1KHz off from peak for core gap of 0.1mm; the variation is much less when core gap is 1.1mm.
Scope photos referred in the table follow:
Photo 1
Photo 2
Photo 3
Photo 4
Photo 5
I was surprised that the coils don't work the same both ways. Perhaps because these coils were wound asymmetrically, or perhaps dues to point of entry on layer. Maybe if the windings were left-right right-left as people normally do, it might be the same; or perhaps not if the difference is caused by point of entry being on lowest or highest layers.
Strangely though, I could not get the currents to be 180° out of phase -- they seem to be nearly exactly in-phase. (In a prior test in my previous thread with E-cores the current was obviously 180° out of phase.)
In showing how to test for magnetic resonance Chris had said to pick the F that gives the highest voltage.
In this case the highest voltage is in Photo 3, but the currents are very low. Photo 5 gives the best combination of voltages and currents. My logical thinking would point to this as better configuration. But is that correct?
Questions to Chris and team:
Any advice and guidelines will be much valued. Thank you! 😁
Hi ISlab,
All Valuable Experiments, but...
I think you've gone off on a bit of a tangent, missing the point of: Chris's Non-Inductive Coil Experiment
Of course, the Forward Mutual Inductance, will create an In Phase condition in the Coils, there will be no Opposition of Partnered Output Coils.
Maybe reviewing my Thread and videos will help.
Best Wishes,
Chris
Hi Chris,
>>you've gone off on a bit of a tangent, missing the point of: Chris's Non-Inductive Coil Experiment
I've already done a quick test and easily got the Sawtooth Waveform (SW) on this coil. To post that as is would add nothing new as insight or progress. My intention here is to explore the magnetic resonance in greater detail to understand its parameters so as to increase and optimise it. Then I will post the SW output with the "before" and "after" of the improvement.
>>Of course, the Forward Mutual Inductance, will create an In Phase condition in the Coils
This is new to me. So far I assumed that any magnetic resonance would produce out-of-phase current. What this means is the magnetic resonance can take place in many modes; we have to find the modes in which the current is 180° out-of-phase and then pick the one with the highest voltage. Is my understanding correct?
[Edit:] In your video "The Secret Revealed -- Resonance Magnetically" you demonstrate this resonance on your coils at about 3MHz, then you say that this is a harmonic of the actual resonant frequency which you show to be about 132KHz. but at this point you don't show the current, nor explain why you did not stay with actual frequency and went with the harmonic instead. Could you please elaborate this? This seems to be very important and deserves fuller discussion.
Thank you!
ISLab
Hi ISlab,
Magnetic Resonance is seen very clearly at the part marked Important:
Which must be tuned for optimum efficiency.
EDIT:
Not forgetting the definition I have laid out for Magnetic Resonance:
Magnetic Resonance is where the Currents in both Partnered Output Coils are 180 degrees out of phase. Equal in magnitude and opposite.
Ref: Magnetic Resonance
Best Wishes,
Chris
Method of finding magnetic resonance is as described in Chris' video:
My oscilloscope was limited to 30MHz, and its sensitivity was limited so that I could not see the low value currents in the previous post, and so got side-tracked to the 0° in-phase readings which alone were visible.
Taking my coils to a more sensitive oscilloscope I get the following results:
Photo1:
Photo 2:
Observation 1:
There are two frequencies of resonance: about 82Khz and about 510KHz. The 82KHz never gives any useful current in L3 during resonance. The 510 KHz gives good current which is 180° out of phase when voltage is measured on L3 IN (with L3 OUT grounded), but gives 0° phase when voltage is measure on L3 OUT (with L3 IN grounded).
Surprisingly, where I put the Signal Generator (L2 IN or L2 OUT) makes no difference to the phase shift in L3! It only makes a difference to the peak voltage which is at least 50% higher when Signal is on L2 IN.
Observation 2:
So I put the Signal on L1 and tried, getting the same result and almost the same F of resonance. Here the best voltage is with Signal on L1 OUT, and no clear current with Signal on L1 IN, although with good resonance. Swapping measurements on L3 gives 0° phase as before.
This is was very surprising at first. But it makes sense if L1 is inducing in L2 which is then resonating with L3 as before, and would explain why the inputs need to be swapped on L1.
Observation 3:
Swapping L2 and L3 gives the same result with same frequencies: 180° when measuring voltage on L2 IN (with L2 OUT grounded), and 0° when measuring voltage on L2 OUT, independent of where Signal is injected in L3. And the lower F gives no measurable current.
Observation 4:
Placing Jagau's SRO on L2 gave F=42KHz unstable, and on L1 gave F=104KHz. Neither gave any meaningful output on L3 which was hooked up as before. This is to be expected as most likely these frequencies represent LC resonance and not magnetic resonance.
Observation 5:
Flipping L3 around on the core, sitting in the same place, reverses the relationships for phase. Now I have 180° phase with voltage measured on L3 OUT (with L3 IN grounded).
Flipping L2 around changes nothing on L3 output phase, but flips the side which gives the highest value of Vpp is now measured when the Signal is on L2 OUT.
Observation 6:
In all these changes of Signal point and measurement points, the Resonant F changes very slightly, within 10KHz, but not because of coil movement as the coils were fixed firmly. Shifting the coil on the core by about 1/2 of its length shifted the resonant peak by about 7KHz. So it seems the changes are due to the entry point of current shifting from one end of the coil to the other when inputs are swapped.
Observation 7:
To maintain 180° phase shift when a coil is reversed we need to swap its leads. Hence the 180° phase depends only on the placement of the Ground in relation to the core and the other coil. The current is AC so we cannot speak of it as depending on the direction of the current. But it is intimately linked to the direction of coil winding. In this case it is the magnetic pole pointing to the centre which stays constant (for the given winding which is CW). Will it be the reverse if the winding were CCW?
Keep in mind that some of these observations would be specific to the peculiar asymmetric winding that I had on these coils (described in previous post). But other observations should be common to all coils.
Observation 8:
The slightest touch of hand or pressure on the core shifts the peak of F radically. As also position of coils on the core, and slight shifts of coil angle if the coil is loose.
Conclusions:
Hello ISlab,
Let me drop this in your Lap: "There is a Natural and also an Induced Resonance"
Frequency can be 10Hz or 10KHz depending on factors I will discuss later!
This topic will be discussed further at a later date! I will not be discussing more until then!
Best Wishes,
Chris
>>This topic will be discussed further at a later date! I will not be discussing more until then!
Noted. I hope it is ok with you that I continue to post my explorations on this though. Please let me know if you prefer that I don't post on this aspect. Also feel free to comment on other aspects of these explorations.
Continuing my explorations with the non-optimal coils
Started with rechecking the optimal F which was found to be 513.5KHz. Then connected the POCs the usual way and pulsed with F at 50% duty cycle and progressively less.
There was no useful waveform and no Sawtooth Waveform (SW) on the POC output. This I take to be because F is too high for L1 to induce in L2. Quick test shows F needs to be below 40KHz to get any useful induction from L1 to L2, and ideally below 25KHz.
I had expected that there will be some precise frequency F / N (where N is an integer) where I will suddenly get much better output. This was not the case, although I think I was able to tune quite closely to F=10.27KHz (which is 513.5 / 50). Duty cycle was reduced proportionally to some 2%.
In practice the entire range from 10KHz to 25Khz gave a nice Sawtooth Waveform (SW), peaking around 10Khz, and then losing the clear triangle and becoming curved as F drops below. For optimum SW, we need F > 10 KHz with duty cycle of maximum 3.5%.
But this did not come at first. At first I got a crooked SW (as below).
This was because the contribution of L3 to SW was so low that the diode was cutting off the bulk of it. Removing the diode completely from L3 (shorting it) gave a clean SW:
Separating the coils up to 1cm made no difference to the SW, but only lowered the peak by about 10%.
Raising the pulsing voltage required me to reduce the duty cycle further to allow SW to fully use up. So it seems I have reached the limit of what these coils can generate.
Now I have two limitations to overcome:
So as a first step I replaced L1 by a much larger and narrower coil -- 4 turns of 8cm diameter SWG 18 loop like this:
The L1 coil inductance is 7.6uH after placing on core and all previous coils.
Running now at the same F and duty cycle gave a much better SW response with much higher peak also!
This is very encouraging. In practice, the bulk of my previous L1 wire was unnecessary. This is much closer to Chris' recommendation of L1 made to only "tickle" L2/L3.
This also reminds me of the geometry used by Tesla in his Amplifying Transformer and also in Tesla Coils. I can appreciate why Don Smith used air coils without needing ferrite cores and yet got a higher output. The ferrites reduce the free resonance while allowing us to work at lower frequencies which are easier to manage.
At this stage I see no point in working further with these coils.
In winding new coils, I will follow these guidelines:
Perhaps with a looser POC, we will be able to tickle it at a higher F, and thus get better output.
I've documented each step in detail (even when I made mistakes) to hopefully help others to avoid them. I hope this is useful.
Hi ISlab,
Yes, no problem. I just don't want to confuse others, not while they are learning! Best to stick to one area and focus on that! 😉
I have added you to the successful Replications list: Here
Very nice work! You have a very good, very nice looking waveform!
You're right on the verge of Excellent Results if you're not already there! You could increase the Input to POCOne Coupling by winding the Coil over the top. Don't be scared of increasing your Magnetic Field slowly, to see more Output, and more Power coming back to your input from the Input Coil.
Increasing your Core CSA, and the Coil Geometry, will help in the overall Output for less Input also.
Excellent work! Thank You for Sharing!
Best Wishes,
Chris
I wound two new coils of SWG 18 on circular core former 44mm diameter, 21 turns covering one full layer only of width 28mm, with total length on coil 299cm, and total wire length including leads 320cm on each coil.
L2: CW, 19.2 uH in air, 0.9mH on core, 0.2Ω.
L3: CCW, 19.2 uH in air, 0.9mH on core, 0.2Ω.
They are exact mirror images of each other when placed as POC.
I chose to start with a single layer to see how it acts, and will add more layers as needed. Also chose to start with the loose 4 turns of 8cm diameter as L1 from the previous experiment.
Rough measure of coefficient of coupling gave approximately: KL1-L2=0.7 to 0.8, KL1-L3=0.8 to 0.9, KL2-L3=0.9.
Attempt to find the frequency of magnetic resonance was unsuccessful. Although I got natural resonance at 592KHz, then again from 1MHz onwards growing until 3.5MHz (limit of my SigGen), but could not find any point with clear 180° phase of current as the current was too indistinct.
Setting up the POC as usual like this:
gives clean Sawtooth Waveform (SW) for F > 20KHz, and slightly curved below those frequencies. Pulsing with 3.6V, F=25KHz, Duty cycle=4% with core gap 0.1mm gives SW like this:
The good news is that both coils are now better partnered as each is putting out a SW waveform independently. Disconnecting L2 makes the SW drop slightly. Disconnecting L3 makes almost no difference. So each is resonating and inducing in the other nearly equally, but L3 is slightly less.
The actual height of the SW is about 50mA. So next I tried to increase this level by several means listed below.
Raising voltage gives:
Raising duty cycle to 7% gives:
Increasing gap between cores to 3mm gives:
Next I took the loose L1 and wound it directly on L2 increasing from 4 to 8 turns. The SW drops due to reduced transformer induction:
Increasing Duty Cycle to 7% gets it back to normal height.
Further steps
So I seem to be stuck with what the coil is capable of giving. Since L3 is contributing slightly less than L2, I added two extra turns to L3. This makes their contributions nearly the same. But beyond this, it seems the only way to raise levels will be to wind more turns on each coil, or narrow the coils much more. My fear is that I may end up as before with too much saturation between coils.
Any suggestions or advice is welcome.
Hi ISlab,
If L2 and L3 oppose each other, then one must assist the Input Coil, so if you have the Polarities correct, then you will see this:
Its very easy to achieve this result.
Best Wishes,
Chris
After some more reading and thinking, I did the following tests to better understand the coils, and had some very happy results! 😁
Pulse Stability: First I inserted a 1000uF capacitor on the PSU output that is used for pulsing the coils. This made the reading on coil output much more stable. Clearly the PSU was strained by the pulses.
L1 Turns: Next I experimented with number of turns on L1 varying from 1 to 7 (somewhat loosely wound) on L2 which is 21 turns, taking detailed readings.
More turns allowed Sawtooth Wave (SW) to be raised to higher level by increasing Duty Cycle (D). Otherwise, on lesser turns increasing D has a limit on how much higher SW can be raised. The current maximum of SW rises proportionally to the turns, but also needs proportionally higher D for that advantage. Since in the final count, we don't want a high D, so this parameter is not critical.
Pulse Frequency: Generally speaking, lowering the pulse frequency (F) had the effect of raising the SW peak, the side-effect of F below 10KHz was a slight curve in the SW. Pressing the cores together slightly made the curve more straight. So I tied the cores together with elastic. Overall this allowed SW to remain somewhat straight even at 1.4KHz.
Pulse Voltage: Next I raised the pulse voltage from 3.5V to 15V in small steps. The SW current maximum rises proportionally, but the input current also rises proportionally. Unfortunately I did not measure the SW voltage at this point. But what is more important is the the duration of the SW was also increasing, possibly more than proportionally (could not make accurate measure of this). In effect raising the voltage raises overall generated power far more than increase in consumption. At several points a rough estimate gave a first impression of being lose to overunity (OU). But the input current measurement is fluctuating and unreliable.
Since input current is proportional to D, the key was to increase voltage while cutting back on D.
Raising the voltage also had the effect of increasing the coil hum/vibration quite dramatically for V > 9V,
Increasing Load current: Finally I played with increasing the SW current by reducing the resistance in the load. So far I had a 0.1Ω for measuring current followed by 4.7Ω as load. Reducing this to 1Ω immediately increased SW current, voltage and coil vibration. The coils felt more intensely active than before, and much louder.
Input pulse current increased also, so I reduced the turns in L1 to 3 turns.
Playing with D for a given F, gave a distinct impression that the coils bucked harder at specific points, from the nature of the sound. The "singing" of the coils hit points of what seemed like an auditory resonance. This was very interesting. I could not detect a dramatic drop in input pulse current though. I believe this point is what Chris is referring to for sudden increase in current. I think I need to be able to control D more precisely to get this. [To Do]
For F around 10KHz, the SW was never fully consumed, so I dropped F to match the load.
Live loads: Next I put a 12V Halogen lamp. Unfortunately the shop only had 50W, so the best I could get was a dim glow, but it got very hot quickly, for pulse V=16V, F=2.5KHz and D=7.5%.
Next loading with a bank of 10mm LEDs x6 in full brightness (+ 3 LEDs not fully lit) at pulse 16V, F=4KHz, D=4%. F and D were selected to fully consume the generated SW and with LEDs at full brightness. Interestingly, with the LEDs, when the SW current is fully consumed, the voltage drops below zero and goes into a negative pulse. I set D to avoid this.
While adjusting D, at several points some of the LEDs dimmed with other brightened. It seemed as if there are resonant waves in the load wires which change with D. Is that possible? Has anyone else observed this?
Here is the coil output for LEDs: [Yellow = coil output SW current, blue = coil output voltage across entire load]
The LEDs and coils:
I feel I'm very close to getting OU. I plan to fine-tune F and D for a given load, as well as tweak L1 further.
At this point I could do with some guidance on further optimisation for OU, and especially on measurements for input power and generated power.
Thanks to all of you for help so far! 😇
While adjusting D, at several points some of the LEDs dimmed with other brightened. It seemed as if there are resonant waves in the load wires which change with D. Is that possible? Has anyone else observed this?
hi islab!
this can be a good thing because this happens when there is a standing wave with voltage-nodes and current-nodes on the TRANSMISSION-LINE 😎
thank you for explaining your findings in such great detail.
regards
B
Hi Islab
It seems that you have experienced what doctor Konstatin Meyl suggests, the first frequency is about 1.5 times what he explains on the technology of scalar waves, in his books which I have placed here in pdf, to read.
Jagau
Thank you Baerndorfer! Yes, the LEDs were in straight line which would have helped the effect to be more visible! This is a line of experiment that is worth pursuing futher with more LED in the line.
Thank you Jagau! This book is very interesting and the experiments seems so simple to replicate!
What is not very clear is this: is he saying that any EM transmitter has a Scalar band of x1.5 frequency? Or is he saying that Scalar operates only at multiples of 6.7Hz?
Looking further for his work I found an updated version of 2014 of the same book, now much expanded with many more experiments and theoretical discusssions here: http://www.freepdf.info/index.php?post/Meyl-Konstantin-Documentation-Volume-1-On-scalar-wave-technology. (More of his downloadable here: https://freepdf.info/index.php?category/Meyl-Konstantin)
(for those who are interested, the printed books/papers can be purchased/read from his website which is in many languages: https://www.meyl.eu/go/indexb830.html)
I also discovered this gem https://web.archive.org/web/20170617045150/https://www.k-meyl.de/go/Primaerliteratur/Faraday-or-Maxwell.pdf which would have implications in understanding POC interactions.
Another gem: https://www.k-meyl.de/go/50_Aufsaetze/water_engine_4.pdf is the only satisfactory explanation I've seen for thunder and lightning.
Continuing Experiments
Changing Duty Cycle: I modified the pulsing circuit for finer control of duty cycle (D) and frequency (F), and then experimented with the point where the coil "singing" seemed louder or seemed to buck more, and tried to fine tune D. On the TL 494 I was able to step up 1Ω at a time. But could not find a dramatic change. Tried arbitrarily with various F. Perhaps there is a special F I should focus on? Raise voltages more?
Changing Load: All this was with the 50W Halogen lamp as load (0.6&Omega
or with a dead 1.2Ω 10W resistor, as these were the lowest values that I had. But the Sawtooth Wave (SW) was not getting fully consumed.
But when I inserted the LEDs bank (measured 54Ω on RLC meter) the SW was fully used up, even though the voltages dropped! This is counter-intuitive as one would expect higher lower usage on the higher resistance. But I now appreciate that in loading OU systems the normal Ohms relationship does not apply as the POC have a varying impedance that adjusts to the load. What is needed is a "live" load that draws more power as it "lights" up, and the POC generates more current as more current is drawn!
My understanding is that the higher LED bank resistance momentarily pushed up the generated voltage which kickstarts the current flow, which actives the lamps which draws more current, and by that time the voltage drops, but the current flow sustains the POC efficiency. In other words, the POC impedance dropped to match the load impedance.
At such times the current measurement levels are also inaccurate as voltage across 0.1Ω is not small enough against the load "impedance". Only the waveform is to be considered as indicative, but not the currents or voltages.
Paradigm Shifts: There are two paradigm shifts needed when dealing with POC and OU in general --
1) Do not think of POC as transformers, but rather view them as transmission lines and antennas with standing waves and resonant waves.
2) Do not think of the output power and load in terms of conventional Ohm's law of impedances, but rather as open, living and dynamically changing energy sources that draw from the aether as much as the load and POC can take.
I wanted to study the voltage changes as the LED load is inserted, but could not for reasons described below.
Timing: With the dead resistor load, I zoomed in on the timing of current in L2 and L3 (based on Chris' comments on precision), and found them simultaneous in rise time within the scope's resolution, varying only in peak and curve, L3 being lower and more curved.
It was not clear from Chris' observation whether L2 should first rise partially or fully to maximum and then L3 should should kick in and slap. So I inserted an extra diode in L3 to delay it. A slight delay was visible on the waveform, but the wave height dropped dramatically.
So I increased the input pulse voltage up to 16V and played with measurements. Before I could take any useful images, the oscilloscope waveform became strange. Further observation confirmed that some high frequency components in the scope have blown. Both my probes were on X1 setting to measure currents, but the maximum rating for the probes is 150V at X1 (300V at X10).
With input pulse of 15V pulse I had already measured spikes up to 20A (20V across the 0.1Ω resistor) with a dead 4.7Ω load. Now it would have shot up to much higher. Surely something in the scope also got overloaded from the spikes.
Lots of important as well as painful lessons learnt!
Questions: I would really appreciate guidance for going forward. In particular:
(Feel free to PM me if you prefer to remain anonymous in the thread.)
Thank you all for help and guidance! 
Hi ISlab and all readers,
Other than my reply Here, I felt ISlab was doing so well that no help was required. ISlab PM'd me asking where to go next, I apologise for this!
Each Partnered Output Coil needs to be treated as the Conventional Generators Rotor and Stator Coil, this means standard Electromagnetic Induction Rules apply:
and another component not mentioned in Electromagnetic Induction:
So with a little more work, the Delta T of the POC Magnetic Fields, in Superposition, Cancelling, because Bone + Btwo = Zero, will "Generate" an EMF in Each Coil, and at specific points, this can be of equal Magnitude!
So don't give up ISlab, work on, specifically, where your POC Slap Together, Magnetically, at TOn when the Mosfet turns on! There is Delayed Conduction here, from L1 to L2 to L3, there is a very important relationship here!
This point:
Is very Important!
This is: Asymmetrical Regauging:
At the point of Regauge, this is where your Partnered Output Coils need to do all the work! Not your Input Coil, we have a near total Work Offset from Input to Output!
This area needs much more study from all Researchers! Its very important!
Floyd Sweets modified Sine wave was also a tell tail sign!
I will help when needed, I saw such great progress that I did not think any help was needed. Congratulations ISlab, your thread is very impressive!
Best Wishes,
Chris
Key points to focus based on Chris' feedback:
What is totally new for me is the focus on TOn when the Mosfet turns on. So far I had only focussed on TOff when we see the current of the Sawtooth Wave (SW) rise and then slide down linearly. The actual pumping up taking place while Mosfet is On is hardly seen in the SW current which remains flat.
My previous report was posted prematurely by mistake. Then the site went down so I could not post my full readings. Below is the original full report, followed by very important progress which I made subsequently.
================
Key points to focus upon based on Chris' feedback:
What is totally new for me is the focus on TOn when the Mosfet turns on. At first I thought Chris had mis-typed this, but thinking further, I realised this is in fact very important. So far I had only focussed on TOff when we see the current of the Sawtooth Wave (SW) rise and then slide down linearly. But the actual pumping up which is taking place while Mosfet is On is hardly seen in the SW current which remains nearly flat, and then suddenly rises when the Mostfet turns off.
But it is clear that how much it rises depends on the duration of Mosfet On time during which something very important is happening which is not visible in the SW. But if we study the currents in L2 and L3 separately, we do see something happening in both coils as seen below where Yellow = SW, Green = L2 current, and Pink = L3 current:
First we note that L3 current (and therefore magnetic field) is distinctly lower than L2 current. We notice also that the non-linear curve of L2 at the start is perfectly compensated by the curve of L3 at the start, irrespective of their levels, to make their sum exactly linear.
How much stronger L2 is over L3 current is determined by what happens during the ON phase. Looking closer there, we see that L2 current is much lower than L3 during the ON phase, marked in ellipse.
Zooming in further:
separating L2 and L3 to study them more closely (White line overlaid to indicate zero level):
We see that L2 drops distinctly lower, while L3 is very close to zero. So, during this ON phase, we need to get L3 to drop as much as L2. The result should be that when the Mosfet switches OFF, they should both rise equally in level, or as close as possible.
Incidentally I made a test with the same settings but with lower side switching, and it gave a much worse waveform:
So stay with upper side switching when working with POCs!
Next I wanted next to check the current in L1 during the pulse and see what it reveals in relation to L2. The problem of course is that L1 is not fully transferring energy in to L2. This relates to the impedance of L1 vs L2 as also the length of L1 in relation to L2. Reading further at https://www.aboveunity.com/thread/antenna-theory-magnetic-resonance/ I discovered this hidden gen of a piece attributed to Ruslan which describes how to find the resonant frequency. This gave me the breakthrough that I needed which is documented next.
================
Recall that earlier I had failed to find the magnetic resonant frequency, or it seemed too high. Now this text gave very useful guidance on how to find it:
The first rule: Wind the coil 40 meters. 2. Find out its resonant frequency (1/4 wave) Inductor 1/4 = 10 meters of the same wire (for example, 2.5mm) Connect the generator to a 10 meter coil, drive the rectangles at a frequency of 1 MHz and crawl higher until 40 meters do not appear sinusoid. The maximum amplitude. The generator is desirable to take a normal, laboratory! With output adjustment from 0 to 20 volts. We achieve maximum amplitude and move tenths of a kilohertz until it starts to dance. This is your wave resonance !!! We fix the frequency and voltage. It is for this all have to do the generators. Further ... Tesla we shake under this frequency that the effect has turned out. Then we do everything as I did or the Shark. In this case, everyone wants to repeat this device. Forward! We fix everything beautifully and stiffly, without forgetting that the resonance and effect can escape in the case of fastenings on the snot. We need to get the effect itself and work, and not a ready device.
My L2 and L3 is already wound with exactly 299cm (+ 10 cm leads on both sides) with 18 SWG wire. So now I took exactly 74.75cm of 15 SWG wire (+ 5cm leads on both sides) and wound it tightly over L2.
Since total turns are less than coil width, I put them all on the left of L2. The reason: In the earlier tests I had observed that placing L1 on left of L2 made the Vpp of the initial spike much higher than placing in the middle which was slightly higher that placing on right which was lowest. Effectively the "kick" is given at the most extreme end of L2-L3 so that the point of contact of L2-L3 is allowed to resonate and collide freely.
A quick scan from 100KHz to 15MHz on L1 with square-wave pulse showed that induction in L2 drops dramatically after 6MHz. And the only point where L2 shows a clean square wave is around 650KHz. (Initially I made the mistake of leaving the load connected on L3. This ruins the waveform. Ensure L3 is free!)
Below optimum F the output waveform is distorted like this: (Blue: input pulse on L1 from Signal Generator, Yellow: output on L2)
and above optimum F, the output waveform tends towards a sine wave and amplitude drops rapidly:
I've deliberately picked F which is way off to document the nature of the change. In practice the waveform is quite close to square about 30KHz above and below optimum F. In fact choosing which F is the best can be quite subjective. But I did notice that the height of the initial rising edge maximises at at very sensitive point, and reduces before and after.
I realised now that I was using too low a pulse voltage. Ruslan recommends something closer to 20Vpp. Raising voltage gave a cleaner edge and the ripples on the top were less prominent but still present.
The slightest movement of coil wire or core shifted the F dramatically by up to 20KHz. So now I bound the coils much tighter. The ripples were less but present. Then I found that moving the leads emerging from the coils changed the ripples!! In other words the leads are causing the ripples due to curves and crossings. Straightening them removed almost all ripples:
Still, you notice the tiny notches at the start of the coils. These shift position with the position of the coils on the core!! In other words the notches represent reflections of magnetic field between the coils.
Moving the coils closer together changes the phase of these tiny notches! Wow!
The best result was with the coils next to each other with leads straight and not crossing:
Notice the shift in phase of the notches. It seemed like the first notch would vanish if I could get the coils to come still closer. But I could not as my leads were sticking out.
This whole experience has been amazing and very fulfilling!
Now it is clear that the entire coil system consists only of antennas influencing each other! Even the coil leads, their lengths and shapes are critical for best resonance. We need to first build the coils L2 and L3, then make L1 as a quarter-wave and then find the resonant F. The key is here. The pulsing circuit, etc, will follow. Build the coils right, then tickle them.
As an aside, when I left the load on L3 by mistake, my LEDs lit up quite bright. This was in spite of low voltage square waves. Clearly the performance at this F is far better.
The next step will be to fine tune F in still finer steps of 100Hz as described by Ruslan, until the coils start "dancing".
More to follow! 
Hey ISLab,
I am very impressed!
Don't give up! You are so very close! You're very advanced and you will succeed very soon if you have not already: 😉
Close, and thorough observations, which you are very good at, will give you a huge advantage!
If I may, there is Symmetry in the Magnetic Field, each Magnetic Field having a North and South Pole of Symmetrical Geometries. As each Field Changes, with equal Symmetry, at Time T, then you may find Symmetrical Placement may give you just a little better performance with a Closed Core. This is not critical, or Important, it may help some is all. Trying to include Symmetries where they do exist can provide advantages also.
Sometimes the Duty Cycle does not make sense, why having a shorter DC can improve POC Performance, but this is another thing to keep in mind! The Input Pulse can be confusing at first!
The small spikes, create a much bigger response: Action, Reaction and Counter-Reaction!
Dedicated Focus on this:
This image is marked Important for a very Important Reason!
Again, I am very impressed, please do not give up!
Best Wishes,
Chris
Hi Chris, thank you so much for your encouraging words!
>>you may find Symmetrical Placement may give you just a little better performance with a Closed Core.
I will keep this in mind. But am wondering if your comment came with regard to my placement of L1 on L2 or the placement of L2/L3 on the core? Do please elaborate.
Continuing
Following the last report, I put hot glue on my coil wires to fix them firmly and made them come straight out from the coils so that the coil forms can be kept close.
The next step is to find the exact frequency of resonance FR by sweeping the frequency F in steps of 100Hz around what seems visually to be optimal at about 680KHz.
The previous test to find the frequency of resonance was made using the output of a Signal Generator (SG) which has a wide range, but its best fine tuning slides in large steps of at least 1KHz at these frequencies.
My existing pulsing circuit based on the Swagatam design around TL494 triggering a Mosfet driver IR2110 to drive an IRF840 Mosfet has an upper limit of around 300KHz for the TL494.
So I turned the Rigol scope which has a built-in frequency sweep which gives a clean output but is limited to 4Vpp, whereas I need 20Vpp at least. I modified the IR2110 driver circuit to make it sensitive to 4Vpp by changing the level of Pin 9 VDD to 3V using a simple resistor divider of values 1K and 330. This works fine giving somewhat clean output pulse on the Mosfet IRF840 to drive a pulse voltage of 3V on L1. But raising the L1 pulse anywhere above 6V makes the IR2110 very hot very fast.
With the best that I can pull from this arrangement, the actual output is not as clean as I was getting earlier, as seen below (Blue=IR2110 output pulse, Yellow=L1 current, Teal=L1 voltage, Red=L2 output):
It looks like I'm hitting the limits both of the Mosfet (which is not as clean at this F) and the driver (which gets very hot at this F). Raising the pulse voltage to 12V does not change the quality of the waveform, but only "clicks" the two cores tighter.
I tried to drive the IR2110 using the output from the SG. This was problematic as the SG gives AC square wave pulses. But while playing around, several times the Mosfet triggered hard and PSU voltage dropped to 0V as if the Mosfet was causing a short ciruit. In retrospect though could have been due to a ground loop, so I removed the separate PSU that was being used for the pulsing only. This seems to have solved that problem.
I also realise that the previous test was made with square wave alternating current between positive and negative pulses, whereas my current circuit is only capable of giving positive pulses.
Next Steps
Further progress is delayed because I'm hitting limits of my present circuit and equipment. Several options are before me sorted by immediate to long-term:
1) Proceed with looking for FR by scanning at 1/4 the frequencies around 170KHz at 12.5% Duty Cycle in steps of 25Hz. Perhaps the easiest option to move forward in the short term.
2) Wind new coils with higher impedance and so lower FR. But this is a long shot as I don't really know enough about what makes the coils work best. Eventually I will do this also, but only after drawing out the full learning that the present coils have to offer.
3) Hack into the SG to insert an external potentiometer. I don't feel comfortable messing with this though as this is a precision instrument with shielding, etc. Would like to avoid this.
4) Replace the TL494, IR2110 and IRF840 chain by higher frequency equivalents (such as IRF840LC for Mosfet) which means effectively building a whole new circuits base eventually.
5) Build a new circuit base such as https://www.aboveunity.com/thread/reliable-and-flexible-switching-system/?p=2 or several related circuits listed on AboveUnity. This will be necessary anyway in the long run, as I would need far finer control of Duty Cycle and Frequency preferably with software control.
In any case I would like are more sophisticated, versatile and isolated circuit. But there are several projects listed on this site. Which would be most versatile and convenient for long-term experimentation? I request those who have used them to please advise.
Your IRF840 only has a pulsed drain current of 32A that's not going to go far in a circuit like this, look for a mosfet with a low gate charge and that should help out your driver, Put a DC link capacitor on the driver with a low esr film cap across it, have a look at the datasheet for recommendations, you can also lower the series resistance to the gate within reason, use a ferrite bead on the gate (highly recommended)
You should have a large low ESR DC link capacitor in parallel with a film cap on your power supply connected with good thick conductors as short as possible.
I recommend the use of snubbers, something as simple as a bulb and a diode connected across the coil so it activates on inductive collapse / reversal will help eliminate ring and make your switching losses lower, you may find the heat output considerably reduced, you will blow mosfets all the time without this in my experience.
Your frequency doesn't matter for observing the effect, reduce the frequency and work on duty cycle and rise and fall times only for now. I have had a look for resonance and what i observed was LC way to high of a frequency to think about getting into.
Increase the load on your output, if the triangle is not straight then use more load on the other side and vice versa, until you have them pushing with equal force.
I can only offer opinions of what i have done in the last week, so take it with a grain of salt.
Hope this helps.
Hey ISLab,
Your comment:
I will keep this in mind. But am wondering if your comment came with regard to my placement of L1 on L2 or the placement of L2/L3 on the core? Do please elaborate.
Yes L2 and L3 Symmetry, because each North Pole and South Pole have their own Symmetry.
Go back to the Sawtooth Waveform, like you have, adjust Frequency and Duty Cycle, until you see maximum Amplitude on the output, then adjust Input Power, and Load, Load first.
Fine your Coils Resonance, this is where you want to be.
Best Wishes,
Chris
HI,
I like how you word the tuning of the POC. Reminds me of when I worked on Big radio transmitters. Some people understood tuning a 50kW tube amplifier and other didn't. And once you understood what was happening, well, it was no big deal. I suspect this is how I need to approach the POC tuning.
Donovan
Very important fact that I did not document at the beginning of this thread is that in my previous thread all experiments were done with circuits on breadboards. But in this thread all circuits were made on soldered boards.
I consider this important as one of the conclusions from the previous thread was that wire lengths and precision of frequencies are critical. Hence all connections need to be stable and fixed for consistency and stability of results.
Recent experiments
Following Chris' advice, the following were my further steps:
After playing with settings, general observations on F and D were as follow:
I decided to focus on the overal Coeffcient of Performance (COP) rather that bright lights and high output power, which are pointless if they draw proportionally higher input power.
After a lot of playing around, I decided to keep D as low as possible to get the highest voltage possible. With my PSU, this was 16V at D=0.6%.
I played with load from a bank of LEDs (needs low voltage to light, but don't know the total power rating) to single and double Halogen lamps of 12V and 5W, 16W and 50W in cobinations. While LEDs lit up and made for dramatic visuals, the actual COP of power generated was higher with Halogen lamps. But these lit up only with high D>10% which also draws 10x more current on L1. This is because the voltage on POC output needs to be near 12V to get the Halogens to light up.
I decided not to worry about visuals, and instead focus on current generated in the SW as seen on the scope.
Increasing the load from 5W to 16W to 50W made for higher and longer SW. I finally settled on 50W x2 in parallel as giving the best power generation. This gave the best SW peak and duration which were comparable to a full short circuit on the POC output! Putting Halogens in series made for much lower power.
Next I found that separating the cores dramatically raises the induced peak current in POC. At F=575Hz the peak was with gap=5mm.
Here I've set my scope amplification on all channels consistently to 2V so that waveforms can be compared easily to assess COP.
Blue is: V on L1, Yellow is: I on L1. Pink is: V on POC output, Teal is: I on POC ouput.
The same out with multiple waveforms:
Both of the above are for F=575Hz, D=0.6%, pulse V=16V, input current in attached ammeter is 15mA, load=50Wx2, gap=5mm.
Ouput measurements are set to show Vtop which is the useful peak of each waveform ignoring short spikes. (For current, read the V value as Amps). POC output current Itop = 4.3A and POC Vtop=4.8V.
Separating the cores further beyond this ponit, the peak stays the same, then slowly reduces.
At F=2.5KHz the peak drops quite a bit with POC Itop=1A, but separating the cores to 10mm raises it again to Itop=1.5A.
Changing the load does not change the peak that is got from a gap separation. That is, the maximum peak for a given load is at the same gap separation for a fixed F.
So my best POC currents were with this significant core gap. Then separating the coils on the cores to their maximum possible distance dropped the peak slightly but made the duration of SW slightly longer than the drop.
Observations and Conclusions
I was expecting a sudden sharp drop in the input current at specific F and D combinations as if some special resonance would suddenly kick in. This did not happen. Rather the optimum was gradual found as described above. I believe this is because in this case we are forcing currents in 180 degree oppostion with diodes using "induced resonance" instead of "natural resonance". (Phrases used by Chria in a post ealier above. I hope I've understaod them correctly.)
Or perhaps I need finer control of D and F for very precise values for sudden resonance to kick in? @ Chris, please advise.
So my best power generation for given input power was as in the first image with peak power at about 4.8Vx4.3A = 20W with the SW triangle surface seeming to be quite a bit larger than the input pulse surface, although it is difficult to calculate the input pulse power due to sharp spikes.
I can and have generated much higher power, but with much high input currents (and power) also, and hence lower COP. But this seems to be the best COP that the current POC coils can generate. Remember they are one layer of 21 turns each only!
To get the Halogens to light up, I will need to raise the POC output voltage much higher. This can be done by raising pulse voltage or it can be done by winding new coils with more turns.
To generate higher COP, I would need to wind new coils with more turns, keeping in mind important lessons from these experiments. In this post I believe I have got a sense of how to find the best F & D & core gap combinations for maximising COP from these POCs.
Circuits
Througout these experimenst with pulse V=16V, the Mosfet stayed always cool, but the IR2110 was getting quite hot even with low currents in L1. I have yet to implement BigMotherWhale's suggestions in this regard. Any other suggestions are also welcome.
Next steps
Wind similar coils with more turns.
Modify circuit to reduce driver heating.
Modify D and F controls to be able to make much finer adjustments.
@Chris and others, please feel free to comment and guide.
Hey ISLab,
Your early waveforms were very good! I would review your thread and my post Here. It appears you have moved away from your previous excellent Waveforms, like this one:
This is excellent work, all I can recommend is don't loose sight of the original goal.
What are the fundamental laws that govern a voltage to be generated? Knowing that your Input Coil Does Not Generate your Output Voltage!
Best Wishes,
Chris
Thank you Chris!
I was chasing the bare "effect" independent of all "flashy" appearance of lights. By effect I mean the maximising of Sawtooth Waveform (SW) surface for minimum of surface from input pulse with maximum of COP ratio between the two.
With the same setup of V=16V and same load as before I can easily change F and D to get clean successive SW as before. Overall COP is much lower though due to higher F.
Changing gap spacing gives the following result:
With no gap
With gap of thin plastic
With 1mm gap
(Blue is: V on L1, Yellow is: I on L1. Pink is: V on POC output, Teal is: I on POC ouput. Note that I have changed Teal units to 1V to make the waveform larger than in previous post.)
The average current on input coil as seen in a mechanical ammeter is 40mA for no gap and thin gap, but shoots to 82mA for 1mm gap. So effectively, in this case, thin gap has higher COP, although visually similar to 1mm gap.
All these data are now my baseline against which I hope to compare the next set of wound coils. In this way I hope to get a clearer sense of which factors in the coils impact POC output in which ways.
Hey ISLab,
There appears to be something wrong with your Input Coil, the Current and Voltage don't look quite right. The Input should be Off, or in a Negative Power State, when your POC are Pumping Current.
Blue is: V on L1
Yellow is: I on L1
The Input Coil should have a waveform something like this:
Here is an example:
Where:
- Purple Trace is the Math, showing Positive and Negative Power.
- Pink Trace is the Gate Signal to the Mosfet.
- Yellow is the Input Voltage.
- Teal Trace is the Input Current, both Positive and Negative.
I must apologise, I have better examples of this, but do not wish to confuse everyone. This example is sufficient to show what I am talking about.
Again, marked in Red, Positive Voltage and Negative Current, you have Negative Power. Not Negative Energy, Negative Power, I hope people do not confuse this as I believe people have in the past.
I hope this helps others when doing experiments, knowing what to look for is very important!
Remember: This is the very reason you can NOT Use RMS Measurements on the Input! See Measurements Thread and see the above Figures:
- Average: 95.7 mW
- RMS: 1.28 W
A Huge error here! 1.28 - 0.095 = 1.185 Watts. 13.474 times!RMS is totally Wrong! Remember, your Zero Graticule Line is very important:
- above: Positive
- below: Negative
Its worth noting, there is 1.28 Watts in the System. An analogy, inaccurate as it stands, is, Power Delivered to the Coil might be in the order of 1.28 Watts. Then the Coil returns 1.28 - 0.095 = 1.185 Watts back to the Power Supply. So the Total Power used is only: 0.095 Watts. Power Returned, is not Power Used!
You need to look for Good, Solid, Fast Turn Off, it appears your Input Coil is staying on far to long.
This will make your COP go through the Roof when you fix this problem! Input Coil configuration and Switching is a very important aspect, which we saw in the early Akula Work.
NOTE: I see you are starting to get the Negative Power Effect occurring! Just not enough yet!
Best Wishes,
Chris
Hey ISLab,
I should add, you want a short Input Coil Duty Cycle for the 100% Output Duty Cycle. You have Output for 100% of the time!
Knowing that your Input Coil Does Not Generate your Output Voltage!
For example, I gave everyone in my video, a 100% Output Duty Cycle with only a 10% Input Duty Cycle.
It appears your input Duty Cycle is upwards of 90% of your 100% Output Duty Cycle?
The Lower your Input Duty Cycle is, the less average Power Used of course.
Best Wishes,
Chris
Hi Chris,
Wow! Your keen obervation is amazing!
Yes, my IRF2110 MOSFE driver had blown. But it seems it degraded gradually -- the only sign being that it was getting unusually hot, and then was getting hotter faster and longer. I believe it started from the point I was playing with 16V on the input pulses. Surprisingly the MOSFET is fine.
In retrospect, it could have happened when I changed the POC load while the POC was live, possibly creating strong back-EMF on the input side.
I checked all past readings in detail going back to the first images where the input current is seen to be leaking. These seem to be only in the last post where I documented the difference with core gap spacing with 3 images. All previous readings, reports and posts are fine.
Note that the scope does not show accurate duty cycle info when the waveform is not a pulse. I'm tracking actual duty cycle and frequency from the pulse going to the MOSFET driver, where the values are accurate.
For the record I want to document the difference between the bad and the good driver here, so that I as well as others can watch out for the signs in future.
Bad IRF2110 MOSFET driver:
With good driver:
I made an animated GIF which makes it easy to see and get a feel of how the POC output changes:
Chris' Second Observation
NOTE: I see you are starting to get the Negative Power Effect occurring! Just not enough yet!
Chris' second observation is very important. I'm highlighting it below for future reference of all.
Although seemingly small matters, I consider both these observations of Chris to be of extreme importance as they are central to optimising POC output and debugging poor POC performance.
I had not paid any attention so far to the Negative Power curve on the input voltage waveform. From now on I will consider it essention to watch (and internally document) the input pulse current and voltage info. This will require zooming into the pulse on the scope before saving the image, as part of my SOP after saving the main output waveform from the wide view.
So far I have tested and posted with two separate coils. The last coils were:
L2 and L3 is wound with exactly 299cm (+ 10 cm leads on both sides) with 18 SWG wire, circular core former 44mm diameter, 21 turns covering one full layer only of width 28mm.
L2: CW, 19.2 uH in air, 0.9mH on core, 0.2Ω.
L3: CCW, 19.2 uH in air, 0.9mH on core, 0.2Ω.
They are exact mirror images of each other when placed as POC.
L1 exactly 74.75cm of 15 SWG wire (+ 5cm leads on both sides) and wound it tightly over L2.
Based on lessons learnt, I've taken the next step of winding new coils with same length of wire as before but on half the width of coil, now having two layers going L to R, then R to L. All else is as before.
This is so that I can learn to get a feel for what aspects of the coil make what kind of difference.
Coil Measurements
New coil meansurements to compare with previous:
L2: CW, 26.1uH in air, 1.623mH on core, 0.3Ω.
L3: CW, 24.3uH in air, 1.614mH on core, 0.3Ω.
Exploring the change when placed on the core, I found that the gap between cores makes a critical difference. Here are the measurements for L3:
Cores with thin plastic gap sitting on table: 1.614mH
Cores with thin plasticgap pressed together by hand: 1.684mH
Cores separated by 1mm plastic: 0.5mH
Cores with no gap: 2.30mH
The gap acts to reduce effective core permeability! We can use the gap to tune the exact point of when the coils hit core saturation. This will be critical for over-unity to extract energy from the aether.
Also the slightest pressure on the cores changes the frequency of resonance. Hence it is critical to maintain stable constant pressure between the cores for getting a stable resonant F.
For now I placed L1 loosely over L2 coil for basic tests of resonance. In the next post I will discuss its optimisation.
We have two methods for finding resonance mentioned in my previous posts. I first used the method described by Chris (quoted in previous posts) pulsing Sine waves into L2 and reading voltage and current on L3 looking for 180° phase difference at maximum V.
I got a 180° phase for input on L2In and V on L3Out at F=437KHz with very fuzzy current image:
Very noisy and difficult to see unless one watches it while F is changing. Then the peak point is seen shifting. I had similar issues with the previous coil also. Likely this is because I'm measuring current on a 0.1Ω resistor and need to use 0.01Ω for better resolution.
But I got a clear current waveform for F=3.41MHz for Signal on L2In and V on L3In as well as Signal on L2Out and V on L3In. The latter image:
In general I got clear currents like this for high F in MHz and fuzzy corrents for F in KHz in all combinations. Not sure why this is so.
Holding the two coils without the core together in the air I could 180° phase only with F=3.448MHz with signal on L2Out and V on L3Out.
Modified Ruslan Method
Next I tried to test with the method described by Ruslan (also quoted in previous posts) with AC square waves pulsing on L1 and reading voltage on L2. But this time I modified it to also test with reading on L3, as well as pulsing on L2 and reading on L3. In each case I switched both sides of the coils for pulsing as well as reading. The results were consistent for frequency of resonance, and varying for intensity of effect depending on the sides of coils used.
Pulsing the input with 20V AC square waves, the output was AC square wave (with ripples at the top) all the way from 40KHz to 3.4MHz, distorting only at the resonant F. The effect was always an unusual distortion in the output with the flat tops curving then tilting to form a linear Sawtooth Wave (SW) shape at peak resonance, then returning to square wave.
The peak output of SW was with signal on L1In and Scope on L3Out at F=466KHz:
The video below shows the full sweep of the distortion:
The exact peak point shifts a little between readings as my coils and cores are not yet tightly fixed. But the pattern is consistent and only happens at that particular F. It did not happen at the higher range of 3.4MHz.
The pulses were 20Vpp, that is +10V and -10V. I reduced the pulse voltage slowly and found that the ΔV of the SW was steady above 16V but rapidly dropped below that. This is a useful measure for how much voltage is needed on the coils to get optimal resonance and generation.
Finally, I found that in all cases the highest waveform voltage was with both coils touching each other. Separating them kept the resonance F the same and the waveform also the same, but the height dropped, with lowest being with coils maximally separated.
The logic for why the SW linear diagonal appears at resonance F should be obvious. This is clearly a point of magnetic resonance between L2 and L3.
The modified Ruslan method of finding magnetic resonance would be useful to most as it is less demanding on equipment and measurement. It can then be confirmed with final validation using Chris' method.
@Chris: If you think it useful, you may consider putting this as an option in the builders' guide info, or in the Magnetic Resonance thread.
Ripples
I also observed the ripples on top of the square waves and how they change based on the position of the coils on the cores as well as distance between the coils. What is fascinating to see is that the distance between the ripples are not dependent on changes in frequency -- they remain steady on the scope as F drops or rises! Clearly the ripples are purely caused by magnetic induction or collisions from the core path and shape in relation to the coils and their positions. Most interesting! The ripples are least and lowest when the POC coils are touching each other in the centre of the cores.
For all these observations I took detailed scope images and videos. They would be too many to post here and possibly confusing. So I'm reported the main results and using here the minimum of images for illustration. If there is special interest for elaboration of any of these, I can share more images. But the above with descriptions should suffice.
Negative power
I put a UF4007 in parallel to the Mosfet as recommended on this forum, but could not find any significant change in the negative power waveform. In any case I will solder it in in view of its potential benefits.
L1/L2 matching
I decided to postpone the planned L1/L2 optimisation of length and impedance matching, in order to get the magnetic resonance going first.
So I fixed L1 on top of L2 using 5.2 turns (coil length = 70cm + 5cm + 5cm leads). The resistance is 0.2Ω; impedance without core = 2.8uH and placed on the core is 59.7uH.
Audio resonance
When pulsing the coils in audio frequency range, one hears them hum or sing increasingly louded with increasing voltage (V) or increasing duty cycle (D) or reducing Frequency (F).
During several past sessions, while turning knobs I had occasionally touched a particular spot of F or D or V when the coils briefly "chirped" with a modulation on the sound. But the spot was elusive and too narrow to get again except while turning slowly and that too only in passing. On a couple of occasions the spot stayed on and the coils chirped repeatedly on their own for many seconds. But later I could not recover those spots.
So I replaced all my potentiometers with 10-turn pots with several values in series to allow fine control. On my TL494 Swagatam circuit, I have F control now with 100K+10K+1K of 10-turn (with 1Meg regular to set baseline) and on D control I have 10k+1K of 10-turn.
Timing of pulses
Checking the pulse from the circuit on the scope I easily get a clean rectangle down to 150nS pulse. The rising edge is clean but the falling edge varies in timing by up to 50nS. Connecting to IR2110 without resistor in between makes the top part of the edge cleanest and sharp. Any resistance in between brings a curve on the top part of rising edge.
The Mosfet output placed on short circuit gives clean edges for the first 150nS then is followed by heavy ringing for nearly 500nS.
The D control of 1K 10T pot allows fine adjustement of pulse width of about 100nS every five steps.
These measurements are critical so that we know the limitations of our pulsing circuits before proceeding with the next steps.
Calculating Magenetic Resonance Harmonics
I set up a spreadsheet where for a given resonance FR we calculate pulse duration and duty cycles for various lower frequencies of which FR is a harmonic.
Assuming my coils are resonant at FR=460KHz the required pulse duration is 1063nS which is within reach of my present circuit. But since my core is not responsive at high frequencies, and since I need to hear the coils sing to tune them, I would like F to be within audio range, the lower the better.
F should be selected so that its even numbered harmonic is FR so that the end of one cycle of F aligns to the end of FR resonance cycle.
For example, pulsing the coils at F=7.3438KHz gives FR as its 64th harmonic. This requires D=0.781% to get the correct pulse duration. Since this D is too short to build up any useful current / magnetism in L1, we need a multiple of this. We must use an odd multiple (as even multiples would only fill the additional half of the resonant cycle) such as D=2.34% or D=3.91%.
Interestingly, for a given harmonic number, the possible values of D are fixed and independent of the frequency. This is very convenient for the TL494 which has a duty cycle control which is independent of frequency.
I chose Harmonic H=256 for preliminary experiments so that F=918Hz for which optimum D=0.098%. A suitable odd multiple will be 2.05% (which is x21). Since my exact FR was somewhat variable due to reasons of wire, coil core, pressure, etc, I used these figures only as starting points, following from then on the audio resonance of the coils to let their song lead me to magnetic resonance.
Following the coils songs
I played with F and D combinations and also raised V. Above 11V suddenly the song is much louder so I kept V=16V which is my maximum, and swept F and D alternately and pretty much at random, listening for louder sound or chirping or warbling modulations. Suddenly raising V created this, but only momentarily.
At various points the audio became much louder and the sometimes the coils and core were felt vibrating to the touch. But at no point did I find the input current drop on the mechanical ammeter. [In retrospect perhaps I should look only on the scope?]
Then while sweeping F, at the specific F=945.2Hz the scope image was started shaking left and right, and only at that F. This sounds like Akula's description of the waveform "dancing". Is this it? Sample:
So I explored this further. For that F the shaking continues even with D varying from 1.5% to 3.8% and with V ranging from 6V to 16V. The only result of changing these values is variation of the song volume. But every 30 seconds or so, the F drifts off and I need to re-tune slightly using the 10K pot to get back the shaking image point.
Tapping the coils or cores, or even the table, makes the chirping sound, but I cannot get it to stay, even with fine control of F and D. Sample:
@Chris: At this point I need some guidance of whether this image shaking is what I should look for? Or the warbling sound? Or the vibrations on the coil and core? Or something else altogether? And what do you suggest to get it stable?
Hey ISLab,
Aboveunity comes when one reaches optimum Magnetic Resonance. This is when your POC are 180 Degrees Out of Phase, approximately equal in Amplitude of each POC Current.
This is where you see your input decrease dramatically, and your Output becomes much greater than your Input.
Best Wishes,
Chris
Thank you Chris! But all of the examples you've posted are from a mechanical motor. Not helpful for POC and my current situation. Presently my currents are 180 degrees in opposition due to diodes, and so I have the Sawtooth waveform. The question is only in tuning of F and D as far as I understand. Is this image shaking is what I should look for? Or the warbling sound? Or the vibrations on the coil and core? Or something else altogether? And what do you suggest to get it stable?
Hello ISLab,
But all of the examples you've posted are from a mechanical motor. Not helpful for POC and my current situation.
No ISLab, you're thinking about this the wrong way!
ALL of these machines work the exact same way, they just use a different approach to get to the same basic working! That's exactly why I just yesterday Posted this:
Tinman was the first:
Others followed, like Captainloz:
Others also, some may not want their names used? Security for people is important to me, but eventually, we must do this as a Team and make a Stand!
As I said in the PM I sent to you:
Have a go at explaining why I said:
Your Input Coil does NOT "Generate" your Partnered Output Coils Output!
When you understand this and can answer this statement with the why, then you will have a much greater understanding of what's occurring.
Don't let frustration creep in, maybe have a break and think on my current posts a bit more before you try again.
POCOne and POCTwo Oppose, so one must assist the Input right? It must! So how much is the Input Dropping when you disconnect and reconnect the Assisting Coil!
Have you identified the Assisting Coil?
Focus on the fundamentals and forget all the other trivial data. Focus on the content I am posting in my posts, forget about the rest! Forget about the rest! My posts here do not focus on much of what you're focusing on ISLab, so best ignore that for now and focus on what is important!
My Apologies if I have come across abrupt, I get very tired of repeating myself when I have already given all the information required!
Best Wishes,
Chris
ALL of these machines work the exact same way, they just use a different approach to get to the same basic working!
Yes, I've got that! Sorry that I was not clear enough. When I said "all of the examples you've posted are from a mechanical motor. Not helpful for POC", I meant exactly this, that although they use "the same basic working", they are very different in their "approach" and implementation. The coils, trigger, core, etc, are all so different. And presently my problems are at the implementation level on details that are specific to these.
Hence my questions were also very specific. When you do not answer them, I assume that either you don't wish to answer for reasons of secrecy (perhaps to manage trolls) or that you prefer that I work it out on my own. But it would be of great help if you or others who have successfully done this can give some indication that would help in course-correction. If you prefer to keep certain things secret for now to manage trolling, a PM would be good also.
Focus on the content I am posting in my posts, forget about the rest!
I thought this is what I've been doing! 😇 Everything that I've worked on and documented here is based on your posts and content from the discussion-threads here. Your last advice to me in this thread was to:
Go back to the Sawtooth Waveform, like you have, adjust Frequency and Duty Cycle, until you see maximum Amplitude on the output, then adjust Input Power, and Load, Load first.
Find your Coils Resonance, this is where you want to be.
So this was my main effort in the previous post.
Please don't misunderstand my words or my intentions. I value every word you say, and all the wonderful information on your site. Right now, I could do with a little hand-holding on specifics in the implementation.
how much is the Input Dropping when you disconnect and reconnect the Assisting Coil!
Will work on this now.
I'm also winding another coil with double the wire length to raise the overall magnetic field and interaction, as I feel this may be at the root of my problems --- insufficient field strength.
Hello ISLab,
With respect, all questions, already have answers in my pages!
My Friend Wistiti has shown everyone exactly how to do this for a long time:
I have referenced Wistiti's excellent video many times. My Friend, please study this video and replicate this set of important effects.
YES, making a point of getting your Magnetic Fields up, both B and H, will result in greater Output Power for obvious reasons.
Best Wishes,
Chris
Hi,
Just watched this video, again......with any learning......I watched it before doing any experiments......thought I understood it.....did some experiments again, and then more experiments......and watch the video, now, again.......and now I have more questions, and need to do more experiments.......Like anything......there's a learning curve......and no lazy way to learn.....you have to do it.....
Donovan
Hello Donovan,
You are absolutely correct! The only way to learn is by close observation of experiment on the bench! Wistiti is a very gifted Researcher!
The most successful Members have generally done the most experiments! The Most Successful Members have done the experiment and the study in what we are sharing.
Little steps for little feet!
Ref: Sir Richard Feymann
So dont rush it, take your time because time is also important, just dont ever give up, because this does work exactly as we all have shared!
Best Wishes,
Chris
Hey ISLab, sorry for not giving any input, just watching your experments in amazement hoping you "break through". I don't want to point you in a direction I'm not sure of myself. But will try to share what I can, easily referencable, though I'm on my phone.
I think you definitely do need more turns, in the past videos i have posted, I was using small ferrite core, with around 135 turns of 21-24 Awg.
-Also, I may be wrong but, in your setup I see no capacitors. I would recommend getting a few adjustable capacitors or work with what you got, though set a point of resonance, certain frequency with different cores will have better Q factor etc....
I use this though live in North America: a.co/d/h1RevbZ">Electronics-Salon 1nF to 9999nF Step-1nF Four Decade Programmable Capacitor Board. https://a.co/d/h1RevbZ
As I learn with mine I'll try to correlate with you see if there's a best combo or harmonic.
I keep seeing get L1 and L2 at 80% and L3 should put you where u need to be. For me that isn't the case, though my wires may be crossed from how I thought I had everything hooked up haha. As you see in one video on my Rep. Attempt, while running as a normal transformer I was using a lot of power to drive L2(5.3watts i think), when L3 ws hooked up, I did get the effects of L3 assisting L1(believe the input dropped 33-50% if i remember correctly) , without the output of L2 diminishing. Though most my configurations, the output on L2 doesn't occur unless L3 is connected. Like you see in CaptainLoz videos(when he disconnected L3, L2 had nearly no output(though his input dropped as well - maybe we had 80% idk).
But the Fr will change, depending on the load. And got few experiments the Fr was up to 10khz lower than what was hooked up and measured (L and C). But I see now with new meter where i can select any frequency(and remember a post with vna) showing the inductance and/or X of L will change dramatically with frequency. 2H shown on my 70mH coil near its Fr.
As of now I'm winding new coils as well, but just FYI, on my AMCC-0200 core ive tested pulsing L2 and powering a load with a diode on L3 and got rly close to around 90% - something that may matter in the futrure
I'm still struggling on making this circuit and operating frequency to not be load specific, though probably need quite a few more things for that....
What core are you using in ur last setup?
Hi,
My two cents worth on capacitors and inductor resonance. I spent about a dozen years doing maintenance and installation on Transmitters for the public broadcaster here in Canada. So I saw a lot of different transmitters.
Hardly ever did I see an inductor that wasn't accompanied by a capacitor in a resonant circuit. If you didn't actually see one, it was something that was "inherent" in the design, ie, a center conductor a certain distance from the metal cavity of an RF filter.
The major exception to this, of course, was antenna's. For instance, an AM radio antenna would be hundreds of meters tall, and the complete tower was the radiator, because that is the 1/4 wavelength at that frequency. Simply a single piece of wire. And say, for instance, an FM radio antenna would be just dipoles on a tower, with the dipoles just a couple meters long, to be 1/4 waveength and resonant at that frequency.
Donovan
Hey Donovan, thank you for sharing that, I believe I've read in some areas the tower is lifted above ground, with grounding cage(or something under it.) Is this what your referring to? Also, what do you mean "like a center conductor? Also if you have info you can share as to what that conductor was connected to? Instantly made me think of kapanadze's device..... Thanks
Thank you AlteredUnity for your suggestions. I may revert with questions as I try these out. I've so far avoided capacitors, trying to get the most learning possible from the natural behaviour of the coils first. I view capacitors as a last step of optimisation only, including capacitor on L1, but only once the F is fixed.
Thank you Donovan for sharing your experiences. We will surely need help of your antenna expertise as we go along!
Hi,
The memory is a funny thing.......the more I talk about a subject, the more I will remember!
All antennas are actually about 1/2 wavelength.......in practice.....of course always exceptions......if the 2nd 1/4 wavelength isn't obvious, then it is incorporated into the ground plane......as would be in an AM antenna. Where the "tower" is 1/4 wavelength and the other 1/4 wavelength is made up of ground radials of about 36 (every 10 degrees) 1/4 wavelength long copper wires plowed into the ground . And then at the feed point, just above the ground, is a massive ceramic insulator about 40 cm in diameter and 50 cm tall to support and insulate the tower from the ground. Then the coax feed time is "matched" to the tower with coils and capacitors, but basically the center conductor goes to the tower and the shield goes to the ground. It was amazing to work on the "big" stuff.
Compare all that to the much higher frequency FM radio, where every is much smaller.......
Center conductors, for example, in a coax, would "look" like a series of inductors from input to output, with "capacitors" to ground, distributed along its length. And then, say for a "filter cavity" which is basically just a metal box with the size determining the operating frequency. The input (and output) would be basically a heavy metal loop, one end at the input connector and other end at ground and the length would be an "inductor" and the distance from the ground or case, would be a "capacitor" Once again, an amazing science.
In the RF world, most chassis are just plain aluminum, while the "center conductors" are copper rod, or pipe, with the higher powered stuff being silver plated in some equipment.
The biggest coax, "hardline" I worked with was 8", so the outer conductor was an 8" copper pipe, and the inner conductor was a 3" copper pipe. With the diameter ratios giving you a 50 ohm impedance. And I use the term "pipe" loosely, it is precision engineered, the pipe diameters and thickness and insulator spacings are all critical for keeping the impedance stable.
Most installation we worked on, were works of art!!!
Donovan
Following Chris' direction to check L3 support for L1 and the reference to Wistiti's experiments, I took my efforts into troubleshooting in diagnostic mode.
So far I always had a clean Sawtooth Waveform (SW) and assumed that all was well. But persistently L3 has not been making any difference to reducing input current, although it played a significant role in make the SW linear. Increasing or reducing L3 turns by just 3 or 6 made a huge impact on SW, so it was definitely strong enough.
The complete block in progress has forced a review from scratch. I went through many checks of which I offer a summary here.
Rechecking currents using the right-hand rule reveals that diode D3 is connected in the wrong direction, and so POC fields do not oppose. But connecting D3 correctly kills the SW. At a very early stage when I had checked this, I assumed the CCW somehow reverses the field. That was wrong.
So how was I getting such a clean SW with wrong diode polarity and without coil opposition?
The answer is in Chris' post here: https://www.aboveunity.com/thread/the-sawtooth-waveform/?order=all#comment-25542b1e-aabc-453c-8083-ac8d001b8b05 where he compares the POC SW with the Flyback Converter SW.
Due to the wrong polarity of the D3, I just had built a very efficient Flyback Converter instead of a POC!
I'm documenting this in detail so as to warn others of making the same mistake. Getting a good SW is not enough! Ensure the correct polarity of your magnetic fields first, then aim for SW.
My sincere apologies to all for this serious lapse! And thank you Chris for your patience. I assure you, that I will more than make up for this mistake in subsequent work.
All that has been documented so far in this thread is now of limited value and perhaps partly wrong since there was no opposition in the coils. I thought of starting a new thread with fresh data. But in retrospect, I feel that the documentation of steps and learning experiences are all valid, including the consequences of such a basic mistake. And the thought process of exploration and validation will stand in spite of all else.
So I will continue on this thread with the necessary corrections that follow.
Getting L3 to oppose L2 is simple -- just turn diode D3. But this kills the SW and makes it flat. How to get back the SW, and then how to get it to support L1?
I placed the existing coils on the following circuit using a breadboard:
.jpg?width=690&upscale=false)
In this mode it is easy to isolate the effect of L3 on input current.
For purpose of consistency of measurements, I took an arbitrary pulse of F=1KHz, D=3%, V=6V and used about 1Ω 5W resistor as load. Also removed the 1000uF on L1 side so that the mechanical Ammeter on L1 is unbiased.
Unlike in the image above, I will refer to the POC components and measurements by the number of their corresponding coils. So D2, D3, I2, I3, V2, V3, etc. to maintain consistency across the thread.
Below are the currents (Yellow=I3, Teal=I2) and voltages (Pink=V3, Blue=V2) on L2 and L3, which I origianally did with the wrong direction of diode D3:
Below are currents (Yellow=I3, Teal=I2) with input pulse (Pink=V1, Blue=I1):
With diode D3 correctly directed we get:
The little notch on the Teal line is actually a straight line! I totally missed seeing this the first time while trying out all combinations (before I realised my error).
Input coil current is I1=80mA for D=3%. Raising the Duty Cycle to D=25% makes input current shoot up dramatically to I1=700mA, but the SW is clearer:
Now to check if input current L1 is supported by L3.
Removing L3 makes the L2 current rise and more like a SW, but the input current drops to I1=650mA:
(Just to compare, with D3 wrong giving the Flyback Converter SW there is no change in input current with L3 removed or even with L2 removed.)
But with the correct diode direction and POC SW, the input current still drops when L3 is removed, instead of rising.
This means that although we have the POC magnetic bucking, the L3 coil is not yet supporting L1 coil.
I tried every possible combination of input/out and diode positions but could not get L3 to support L1 as we want it to, or as shown by Wistiti. So next I turned to the Wistiti's configuration.
In his post and video at https://www.aboveunity.com/thread/some-coils-buck-and-some-coils-dont/?order=all#comment-92df39b3-7baa-4b4b-9d37-a88c00147dde, Wisititi offers this as circuit diagram:
.jpg?width=690&upscale=false)
Wistiti wires his L1 equally across both L2 and L3. So first I tried this, and immediately got the effect of L1 current reduction with L3 on. But L1 wrapped only on L2 for POC gave even better results which are documented here.
In the subsequent scope shots Yellow=Voltage and Teal=Current from coils output, and D3 is the diode at the centre of POC, and D2 is the diode at the extremes of POC.
With correct positioning of both diodes the input current I1=320mA:
Shorting (removing) D3 correctly raises input current to I1=380mA:
Shorting (removing) D2 also correctly raises input current to I1=380mA with similar waveform as above.
Shorting (removing) both diodes kills the SW and raises current the highest to I1=470mA:
It is obvious that the POC did not get a chance to build up to opposition, and there is no current pumping effect.
Reversing both diodes raises input current to I1=440mA and without SW:
With reversed diodes, shorting one or both of the reversed diodes raises the current slightly more with I1=460mA with similar waveform.
Here I have successfully and entirely replicated Wistiti's experiment.
There remains one slight glitch. The SW is not yet a pure straight line. I realised that during the rearragement of coils, I had dropped the very thin piece of plastic from the gap between the cores. Re-inserting it gave the perfect SW:
This result is deeply satifying. We have a clear and significant reduction in input current with full POC function.
The load configuration of Wistiti is very new to me, and I don't recall having seen it anywhere on this website.
What is very unusual is that shorting out any one of the diodes kills the effect of L3 supporting L1. In our normal way of thinking, a single diode on one side of the load would be the same as having both diodes on both sides of the load, especially since the load is a totally passive resistor. This is very strange and needs deeper thought.
The slight gap of core made a huge difference to the SW linearity. This will now be my reference for optimum core gap -- linearity of SW! The plastic piece is 0.065mm measured by micrometer. Putting a 1mm piece made the SW very narrow and slightly distorted.
This configuration is still too new, and I will have to explore is further to get a "feel" of how the coils behave in this. But what I find promising is that the peak voltage on SW is already 3.6 volts for an input pulse of 6V. This means scaling up will be much easier and the input drop much more rapid.
At this point I have succesfully corrected a major mistake that had taken me off track, and now feel confident of further progress.
I'm actually surprised at the ease with which I was able to get all the effects once back on track, and can better appreciate Chris' frustration when people don't get it. I hope that my detailed documentation here will make it easier for many more to replicate as well as avoid the most common pitfalls.
This is very exciting and promising, and I look forward to playing with this configuration and reporting progress soon.
Hi,
I am just reading your last post.....I believe this is where I am at.......I appreciate that you have posted in so much detail......tonights efforts will involve studying your work while I troubleshoot mine.
Thanks,
Donovan
Hello ISLab,
This is by far your best post and set of experiments! Thank You for your very professional Approach!
NOTE: POC Function is broad and more than one configuration exists to achieve the same Effects! Input Current going Down Under Load is the critical aspect of all Free Energy Machines! Floyd Sweet used 33 Micro Amps under loaded conditions!
I agree:
But in retrospect, I feel that the documentation of steps and learning experiences are all valid
I agree, this is a Learning Curve and we must always force our brains to NOT assume, because that's the main issue with all of Science Today!
Yes, there are several Polarities that need to be taken into consideration, which your experiment shows! The Correct Polarity must be obtained, which once done, becomes easy! Everyone must follow the Big Red Arrows for conformity!
You are right, we must think about this the right way, to make this work, it is VERY Easy, but not as intuitive as many think, again the assumptions many make, hold them back from a whole new, hope filled future!
I believe you will now understand better why I have said:
Knowing that your Input Coil Does Not Generate your Output Voltage!
I am very pleased you have achieved these goals and now all you have to do is some deep study and documentation, and look at getting the Magnetic Fields up so as to bring the Output Power up, of course one follows the other as we all know!
Well Done ISLab! I am impressed with your effort! You have achieved Asymmetrical Electromagnetic Induction, a feat that 99.99999% of all Researchers don't even know exists!
Don't give up, because you're on the verge of big breakthroughs, breakthroughs others here have also gained and achieved! Breakthroughs the Other Forums will never achieve, because We are Light Years Ahead of the other Forums!
If only more would join in and learn, we could free Humanity tomorrow from the Chains of Ignorance!
Best Wishes,
Chris
I soldered a UF4007 diode on the MOSFET as suggested by many here and added a finer control on Duty Cycle for future tests.
Activating the circuit which was exactly the same as the last time, gave the SW as before. Then shorting D3 to check for rise in input current gave no change!
I tried again, checking every connection. All was ok but I got no change, or a barely visible 3mA or so increase.
I disconnected one side of one coil, then both sides of either coil, and the SW remained unchanged and the input current I1 stayed unchanged!
Disconnecting all coils broke the SW but left I1 exactly the same.
Finally, removing the thin plastic from the gap got back the effect and I1 rose by 20mA. The difference was caused purely by the now increased magnetism that clicked together the cores and held them quite tightly. The thin gap was enough to break this magnetism.
The difference between yesterday and today was a slight drop either in pulse voltage V or in Duty Cycle D, just enough to bring the coil magnetism below the threshold required for current reduction effect.
Looking back I was lucky to try with 6V last time, as that is the minimum required for my coils. Today's failure was due to a very slight reduction just on the edge of the effect.
Slightly raising V a little got back the effect suddenly and I could put back the thin gap to make the SW linear again. The effect also returned with a large increase in D.
The difference between non-interacting coils and magnetically interacting coils is heard by a sudden increase in the singing volume (I still have F=1KHz).
In fact there was a sudden increase twice from 6V to 7V pulse, each giving better interaction seen in an increased drop in I1.
All this shows that a minimum level of magnetic interaction is required before current reduction is suddenly seen. This is best done by raising voltages rather than by raising D, as seen below. You will know this effect has kicked in when your cores will attract and click together, making it difficult to separate them by finger pressure.
For convenience I will refer to the change in input current I1 with D3 shorted as ΔI (Delta I) hereafter.
Some observations:
For example:
These figures are approximate as everything is still on breadboard and the shorting is done by touching wires by hand. With the last settings the D2 and D3 diodes got very hot.
At one point, a much lower D and higher V, I got a distinct gain of about 50%, when input currents were low in the range of 30mA. I did not record the output power in any of these, since this is only a preliminary exploration to get a feel of how the coils behave in this circuit.
With the last setting of V=14V and D=10% the following are scope shots for various core gaps.
No core gap gives:
Core gap of thin plastic of 0.065mm gives:
Core gap of 1mm gives:
The peak current in the first and second images are over 1A for a nearly steady generated voltage of 3.2V. This is without optimising the load or input F and D. Shows promise.
Hi,
I am new here......where did you place the UF4007, on the MOSFET?
Donovan
where did you place the UF4007, on the MOSFET?
Between Source and Drain. See the section called Avalanche Protection here: https://www.homemade-circuits.com/mosfet-protection-basics-explained-is/
It is not necessary for the effect, nor even essential for protection unless you are handling high voltages and currents on the Input, AFAIK. Must be ultra-fast though for our purpose.
Although I made some interesting observations with the Wistiti circuit, I've held back posting as I wanted to improve my setup first. (I will report on those later.)
My progress has been blocked, I believe, because my TL494 based frequency generator does not have sufficient precision on duty cycle, and my high-side IRF 2110 MOSFET driver circuit is getting very quickly hot when I raised pulse voltages beyond 12V as the high voltage is leaking back into the driver output due to issues with the high-side bootstrap circuit, etc.
So I decided to build the https://www.aboveunity.com/thread/reliable-and-flexible-switching-system/ to at least be able to push higher voltages without hassle.
I've soldered one channel on the board to test:
Voltages are ok at a stable 15.3V within the circuit for an input voltage of 7.0V or above. MOSFET is switching fine. So everything seems to be ok.
Except the pulses are badly varying in frequency as seen here:
My original input pulses are rock steady and clean. They enter the IL610 at about 3.6Vpp. But the output from the IL610 is coming like this with the shaky pulses as if the frequency is varying all the time around the actual F.
The above photo shows the placement of probes measuring just at the output of IL610.
@Chris, and others, could you please indicate what could be the problem and how I might try to fix it?
Hey ISLab,
My first thought was some kind of supply ripple causing a reference trigger voltage to change but looking at the IL610 I'm not sure that is plausible. I'm still curious though what the voltage looks like across pin 8 and 5 of the IL610 if anything is showing up there.
Nice PCB by the way, I have been trying to get some made through JLCPCB but when I upload the files I get a digital equivalent of a blank stare in return. Lets see if I can get it to work today.
Matt
No one online at the moment
In physics, scalars are physical quantities that are unaffected by changes to a vector space basis. Scalars are often accompanied by units of measurement, as in "10 cm". Examples of scalar quantities are mass, distance, charge, volume, time, speed, and the magnitude of physical vectors in general.
You need to forget the Non-Sense that some spout with out knowing the actual Definition of the word Scalar! Some people talk absolute Bull Sh*t!
The pressure P in the formula P = pgh, pgh is a scalar that tells you the amount of this squashing force per unit area in a fluid.
A Scalar, having both direction and magnitude, can be anything! The Magnetic Field, a Charge moving, yet some Numb Nuts think it means Magic Science!
Hello my children. This is Yahweh, the one true Lord. You have found creation's secret. Now share it peacefully with the world.
Ref: Message from God written inside the Human Genome
God be in my head, and in my thinking.
God be in my eyes, and in my looking.
God be in my mouth, and in my speaking.
Oh, God be in my heart, and in my understanding.
We love and trust in our Lord, Jesus Christ of Nazareth!
More than anything else, your contributions to this forum are most important! We are trying to actively get all visitors involved, but we do only have a few main contributors, which are very much appreciated! If you would like to see more pages with more detailed experiments and answers, perhaps a contribution of another type maybe possible:
They REFUSE to tell me why!
The content I am sharing is not only unique, but is changing the world as we know it! Please Support Us!
Thank You So Much!
Ere many generations pass, our machinery will be driven by a power obtainable at any point of the universe. This idea is not novel. Men have been led to it long ago by instinct or reason. It has been expressed in many ways, and in many places, in the history of old and new. We find it in the delightful myth of Antheus, who drives power from the earth; we find it among the subtle speculations of one of your splendid mathematicians, and in many hints and statements of thinkers of the present time. Throughout space there is energy. Is this energy static or kinetic? If static, our hopes are in vain; if kinetic - and this we know it is for certain - then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature.
Experiments With Alternate Currents Of High Potential And High Frequency (February 1892).
