OverUnity Research

Quote from: verpies on 2017.04.20, 23:07:38
Typo in subject
"Bifiler" should be "Bifilar"

Quote from: TinselKoala on 2017.04.21, 03:22:47
So there must still be a substantial alternating magnetic field happening in the TBF coil.

Quote from: TinselKoala on 2017.04.21, 07:33:49



It's hard for me to see how either the magnetic or electric fields can be alternating so nicely when there is no current -- or more precisely voltage drop -- evident in the CVR. Unfortunately when the "secondary" pickup coil is loaded with, say, an LED, the TBF's CVR current trace no longer flatlines. It gets small and still is at minimum at the TBF's resonant frequency, but is definitely no longer at zero.

Quote from: partzman on 2017.04.23, 13:36:29
TK,

If you would measure the inductance of just one of the bifilar windings in your TBP and also the capacitance between the windings, I would attempt a sim to see if the TBP is functioning as a cross coupled symmetrical transmission line as depicted in the attached schematic.

Regards,

PM

Edit: Added "cross coupled".

Quote from: TinselKoala on 2017.04.23, 20:55:18
OK, so I disconnected the centertap and measured the two windings individually. They measured 176 uH and 172 uH. Capacitance between windings is 2.80 nF (but we know this isn't the distributed capacitance when the centertap is connected). When centertap is reconnected the total inductance measured 680 uH. The DC resistance is 3.90 ohms.

Quote from: partzman on 2017.04.23, 23:24:28
TK,

Here is a sim of your TBP modeled as a symmetrical 12 section transmission line with the parameters you supplied.  There are no parasitics included nor couplings between sections so the accuracy may reflect this when comparing these results to your bench tests.

TBP_1A shows the plot sweep from ~50kHz to 1.25MHz and as can be seen there are 2 dips and 1 peak in Vout across the 1 ohm sense resistor.  There is also a trace showing the sum of the all the currents in the odd inductors L1-L23 over time representing one winding of the TBP and the other line is identical but not shown for clarity.  Also, the 10v peak input voltage Vin from the generator is divided by 100 for plot clarity.

TBP_2A shows an expanded view of the first dip in output voltage/current in Vout across/thru R1.  The frequency is ~277kHz, Vout is ~561uv rms, while the total current in the odd L's is ~140ma rms.  This shows the apparent output current to ground of the TBP to be nearly zero while there is still induction in both windings which would be detectable with a separate sense coil in proximity.

TBP_3A shows an expanded view of the 1st peak in output current at 505kHz plus the levels of output voltage and the odd L's winding current.

Based on your measurements,  your TBP has a Td=1.38us and Z=492 ohms.

Regards,
PM

   

Quote from: JimBoot on 2017.04.24, 11:08:44
there's a vicious rumour going around that someone is onto a sweet deal with a pancake coil supplier. Without mentioning any names <cough>Partzman</cough> Can anyone get in on that action?

Quote from: TinselKoala on 2017.04.24, 12:26:04
Verry interesting indeed. IIRC I measured 273.2 kHz for the resonant frequency of this coil, so your sim is not far off at all. Or rather, my measurement is not far off at all! Maybe even just my scope probe capacitance is enough to account for the 4 kHz difference.

Quote from: TinselKoala on 2017.04.25, 06:46:36
Dear Friends:
Magnetic field (AC) comparison rendering now and should be viewable in a couple of hours.   8)


Dear TinMan:
Saw your "BEMF" comparison; 100 percent Thumbs Up!    O0


Ever faithful, I remain
--TK

Quote from: TinMan on 2017.04.26, 10:15:52
T.K

I will be sending you a link to a video i am about to shoot-via PM.

This is in regards to something i have found with the BPC,that i have never seen in any other coil-and i have carried out this test with other coils.

We have been looking for something different about the BPC,and i think i may have just found it--never give up looking  O0

AFAIK,this go's against all the books tell us-AFAIK.

I have a theory as to why i am getting these result's,but we can discuss that after-after you confirm what i have found,to be correct.

Be in touch soon.


Brad

Quote from: TinselKoala on 2017.04.26, 19:54:49
I can confirm that the effect you showed in the video is real, correct, and relatively easy to reproduce with my own TBFP coils. I also think I have an explanation, as I said in the PM I sent you. But I'll wait until you post your demonstration for everyone to see and discuss. I think Partzman will be especially interested and I am curious if he will hit upon the same explanation that I think I have. I also have a demonstration of what I think is another version of the same effect, in a different system that has been discussed here not too long ago. I'll try to get a video of this demo up, but I'll wait until after you've started discussing the effect in the open thread.
Cheers, mate--
--TK

Quote from: TinMan on 2017.04.26, 20:51:15
As a seco d test TK

If you get your BPC to once again show 0 current input-what do you get across !what would be! R2 in my video?

Brad

Quote from: TinselKoala on 2017.04.26, 23:18:59
Well, with the resistors in place I can't quite get down to "zero" input current because R2 represents a power dissipating load, but pretty close. At resonant frequency (which did not change due to the resistors) the input current read 1.8 mA rms and the "R2" current read 8.6 mA rms. This is with the FG set for 5 v p-p sine.

Quote from: TinMan on 2017.04.27, 10:07:44
1-Using your current value(1.8mA),and then the voltage RMS value across the coil as a whole-what is your total P/in?

2-What are the values of your resistor's?,so as we can calculate the power dissipated by the two resistors.


Brad

Quote from: Itsu on 2017.04.27, 11:07:13
More dramatic when using the resonance frequency (300Khz)
input current (flatlining) 1.16mA rms, middle tap current (constant) 13.95mA rms

Measuring the voltage across the TBP coil shows we have 20Vpp at the input, and half (10Vpp) at the middle, so probably there lies the explaination.


Itsu

Quote from: Itsu on 2017.04.27, 14:06:56

The only time that P=I*U with AC sine wave, is when we are in resonance, else you need to include the phase difference between U and I.
I measure around 81° phase shift when not in resonance.

Itsu

Quote from: TinMan on 2017.04.27, 12:13:42
So you halve the voltage,but the current value is 1202% more  ???
I dont think the half voltage value is it at all.

So P/in is 7.07VRMS @ 1.16=8.2mW-->?
At mid point,power is 3.535VRMS @ 13.95mA=49.3mW-->?

Itsu

Below is a drawing of a test i just carried out,where i calculated power dissipated by each individual component.
Could you carry out the same test?

Cheers

Brad

QuoteMy transmission line demo video is up now:

http://www.youtube.com/watch?v=f4T5KKQjz0s

Quote from: TinMan on 2017.04.27, 12:13:42
So you halve the voltage,but the current value is 1202% more  ???
I dont think the half voltage value is it at all.

So P/in is 7.07VRMS @ 1.16=8.2mW-->?
At mid point,power is 3.535VRMS @ 13.95mA=49.3mW-->?

Itsu

Below is a drawing of a test i just carried out,where i calculated power dissipated by each individual component.
Could you carry out the same test?

Cheers

Brad

Quote from: Itsu on 2017.04.27, 15:04:09
Tinman,

here a similar drawing as you presented with my measurments in it.
I have no resistors, but used my current probe and in the drawing i have shown where.

I did not made the calculations like you did as i doubt that can be done the way you did.

The screenshot is the total input measurment, we are at 1MHz , so NOT at resonance.


For what its worth:


Regards Itsu

QuoteThings to think about.

Quote from: Itsu on 2017.04.28, 08:39:21

  :D You can say that again!


Something in my measurements must be amiss as the current in the total measurement and the one across L2 should be the same (same position current probe).
Unless my putting the voltage probe ground lead at the L2 - position influence this current,  have to look at that again.


Itsu

Quote from: TinMan on 2017.04.28, 08:46:17
Itsu

They wont be the same  :o
Place a couple of 100 ohm resistors in there,and try the LED test--the LED wont lie  O0

See my post above as well.


Brad

Quote from: Itsu on 2017.04.28, 08:58:43
Right,  i will test that later today, just need to clip on my ground probe in the middle to see if anything changes.
Else its like TK mentioned already, this transmissionline effect what makes his leds turn off and on (standing waves) in his excelent video above.

Regards Itsu

Quote from: TinMan on 2017.04.28, 08:44:11
OK,so i have made this very easy for some one well versed in calculating power factor into the below results.

You will see that i raised or lowered the voltage value by small increments of each channel,so as both wave forms are even-close to V/D line-making the job easier  O0

Each pic below shows the scope positions,and attached is the associated scope shot with the values of each channel.

So,who is well versed enough to calculate-->
1-Total P/in
2-Dissipated power of R1+L1
3-Dissipated power of R2+L2


Brad

Quote from: TinselKoala on 2017.04.28, 11:21:55
Brad, can your scope measure the Phase Shift between channels? Or can you post one shot with just a single full cycle of both channels horizontally so we can estimate it visually?


Anyhow... for the power calculations.... I'll take a shot at it.

I think for the moment we can disregard the dissipation in the Coil portions and just look at the resistors themselves.

In your first shot, to get the input power, we are looking at the voltage across 200 ohms so this is shown by the CH2 measurement, which is 6.20 Vrms, giving a current of 0.031 Arms.
So the total power IN is P=I²R = 0.192 watts.

EDIT: Do we need to make any corrections for possible phase shift at this point?

For the second shot, looking at the Vdrop across R1, again we read CH2 at 1.36 Vrms, which gives a current value of 0.0136 Arms, and a power dissipation of P=I²R = 0.0185 watts.

For the third shot, looking at the Vdrop across R2, we read CH2 again, at 2.56 Vrms, for a current of 0.0256 Arms, and a power dissipation of 0.0655 watts.

So we have Pin = 0.192 W and total dissipation in the resistors of 0.0185 W + 0.0655 W = 0.084 W dissipated in the resistors, compared to 0.192 W input. 

HOWEVER... The picture changes if we take the R+coil readings on CH1 and do that math. I get a total "dissipation" value of 0.2371 W which would lead us to conclude that we have an OU COP of about 1.23.

Cheezburger Time !!!


But is this really correct? We know that the pure resistors are dissipative loads that give off power as heat, lost to the system. But what about the coils? Neglecting their DC resistance of course... is the inductive load actually dissipating power like the resistors do ... or is it _storing_ it?



I've been working on this for about an hour, and it's late and I am coffee-deprived, so PLEASE check my math and assumptions everybody!

Quote from: TinselKoala on 2017.04.28, 12:38:41
Your Atten can't do trace math? ISTR something about that. Well, I don't really trust the Rigol's trace math very much either, really, and I didn't include RMS values in those shots up above because the scope's values are a little smaller than they should be for pure sine waves... wait, maybe that's telling me they aren't really pure sine waves and the "p-p x 0.3535" calculation doesn't apply.

For the FG pumping out the power, it's not hard at all to cobble together a simple powered transistor amplifier that can be driven by the FG and put the same sinus signal at greater voltage into your coils.

As far as a 4ch scope with isolated channel references... that's a toughie. I've used some 2 channel scopes, portables, with isolated references like the Fluke 199 ScopeMeter and the Tek THS720P (has power quality analysis built in!) but I can't think of any quality 4ch units with isolated references. However... the cure for that is Differential Voltage Probes !! And real Current Probes !! They can be costly too though. I wish I had one of each myself.  But there are pitfalls.... we are dealing with a situation where inductance can have a great effect on measurements and that's not even including phase shift/power factor.

Remember this blast from the past?
http://www.youtube.com/edit?o=U&video_id=KWDfrzBIxoQ

Quote from: TinselKoala on 2017.04.28, 12:45:04
Brad, does your scope have cursors? If so can you read the values of the Math trace using the cursors?

Quote from: TinselKoala on 2017.04.28, 12:49:08
Argh, I got the wrong link, I blame YT changing things.
Here's the one I meant:

http://www.youtube.com/watch?v=KWDfrzBIxoQ

I'll also edit my original post.

Quote from: TinselKoala on 2017.04.28, 12:53:31
Yes, this one was fun:

http://www.youtube.com/watch?v=hvf9Uo7UVx0

Quote from: TinMan on 2017.04.28, 13:22:44
Yes-thats the one.
Brightest LEDs ever-without blowing lol.
Must be negative current  :D


Brad

Quote from: web000x on 2017.04.27, 14:11:10
I would appreciate it if somebody could confirm or deny my suspicions that this is a distributed, parallel LC network,

Quote from: TinselKoala on 2017.04.28, 11:21:55
EDIT: Do we need to make any corrections for possible phase shift at this point?

Quote from: TinselKoala on 2017.04.28, 11:21:55
But is this really correct? We know that the pure resistors are dissipative loads that give off power as heat, lost to the system.

Quote from: TinselKoala on 2017.04.28, 11:21:55
But what about the coils? Neglecting their DC resistance of course... is the inductive load actually dissipating power like the resistors do ... or is it _storing_ it?

Quote from: verpies on 2017.04.28, 13:55:50
No, because the current and voltage are always in phase in a resistor.
Yes
A pure inductive load can only store energy and give all of it back.  It is lossless in that respect.

The current and voltage across an ideal inductor are not in phase so the formula for average dissipated power P_AVG=V_RMS*I_RMS does not hold (is false in that case).
Actually the power dissipated in an inductor is zero over entire cycles, because the power factor for a pure inductor is always zero, too.  This is because cos(90º)=0.
The same goes for capacitors because cos(-90º)=0, too

EM radiative losses are another story.

Quote from: TinMan on 2017.04.28, 14:26:44
Could you take a look at post 40,and let us know what you think.

Quote from: TinMan on 2017.04.28, 08:44:11
2-Dissipated power of R1+L1
3-Dissipated power of R2+L2

Quote from: verpies on 2017.04.28, 13:44:31
Yes, it is a coiled distributed LC network, that exhibits all of the behaviors of a transmission line when its length approaches or exceeds the wavelength of the signal applied to it.  Like in this video:

https://www.youtube.com/watch?v=ozeYaikI11g

Quote from: web000x on 2017.04.28, 17:16:51
That would be a wavelength of 300 meters. None of those BP coils look to me like they are within that range... please feel free to correct me if I'm wrong

Quote from: verpies on 2017.04.28, 14:49:03
I agree with TK's calculations
With respect to this, I must add that these questions do not have a clear answer because L1 and L2 do not dissipate any energy - they just store it, only to give it up later. 


The resistors R1 and R2 will dissipate electric energy as heat, of course ...just like TK has calculated.

QuoteIdeal inductors and capacitors are non-dissipative (lossless).

QuotePerhaps a better question to ask would be: "What is the ratio of stored to dissipated energy in R1+L1 ?," etc...

Quote from: verpies on 2017.04.28, 14:49:03
I agree with TK's calculations
With respect to this, I must add that these questions do not have a clear answer because L1 and L2 do not dissipate any energy - they just store it, only to give it up later.  
Ideal inductors and capacitors are non-dissipative (lossless).

The resistors R1 and R2 will dissipate electric energy as heat, of course ...just like TK has calculated.


Perhaps a better question to ask would be: "What is the ratio of stored to dissipated energy in R1+L1 ?," etc...

Quote from: TinMan on 2017.04.28, 23:04:34
Yes, but we are far from ideal.

Quote from: TinMan on 2017.04.28, 23:04:34
Well,as we know what the dissipated energy is in both R1 and R2,the better question would be-what is the dissipated energy in L1 and L2.

Quote from: TinMan on 2017.04.28, 23:04:34
So how to calculate the dissipated energy by L1 and L2 ?.

Quote from: TinMan on 2017.04.28, 23:04:34
If an inductor/coil is radiating a magnetic and electric field,is it not dissipating !some! power?
It is my understanding that it takes energy to produce a magnetic and electric field that changes in time.
This being the case,then L1 and L2 must be dissipating some energy at all times.

Quote from: partzman on 2017.04.28, 20:23:05
Regarding the velocity of TK's TBP using his overall inductance of 680uH and distributed capacitance of 2.8nfd with the formula v=1/√LC, we get v=724.71e3 m/s.  The velocity factor would then be Vf=1/c√LC=2.416e-3 where c is the speed of light or 300e6 m/s.

With the simulation first dip at 277kHz, the wavelength using the above velocity would be 2.61 meters or 1.31 meters for half wavelength.

However, this may not tell the accurate story due to the cross coupling of the distributed capacitance.  This is easier to visualize when looking at the attached schematic with the lumped equivalent TBP sections redrawn.  This is not a conventional transmission line.

PM

Quote from: partzman on 2017.04.28, 20:23:05
Regarding the velocity of TK's TBP using his overall inductance of 680uH and distributed capacitance of 2.8nfd with the formula v=1/√LC, we get v=724.71e3 m/s.  The velocity factor would then be Vf=1/c√LC=2.416e-3 where c is the speed of light or 300e6 m/s.

With the simulation first dip at 277kHz, the wavelength using the above velocity would be 2.61 meters or 1.31 meters for half wavelength.

However, this may not tell the accurate story due to the cross coupling of the distributed capacitance.  This is easier to visualize when looking at the attached schematic with the lumped equivalent TBP sections redrawn.  This is not a conventional transmission line.

PM

Quote from: Chet K on 2017.04.29, 13:40:26
Russ is exploring here also [Thx to Matt Watts]

https://www.youtube.com/watch?v=-HDwOwfIHns

Quote from: ION on 2017.04.29, 22:57:18
I agree, the video was rather inconclusive, more just like a fleeting overview on power factor without addressing the meat of your question.  ???

I'm truly glad that Russ is fervently open source. Hopefully his benefactor feels that way too.  :)

Where could I find a benefactor?........ :)

Quote from: TinselKoala on 2017.04.30, 09:54:46

Ok.... it would still be better if your scope could compute the phase shift itself. But in your First shot, if we take the rms current through R1 (CH2) as 0.0618V/10R = 0.0062 A_rms as the current being supplied to the system, and the voltage drop across the whole circuit (CH1), and estimate the phase shift as 72 degrees, we then calculate the total input average power as
P = V_rms x I_rms x cos (72 degrees)
P = 6.60 x 0.0062 x 0.309
P = 0.0126 W or 12.6 mW

And if we calculate the power dissipated in R1 as
P = V_rms²/R
P = 0.071²/10
P = 0.0005 W

And the power dissipated in R2 as
P = V_rms²/R
P = 0.348²/10
P = 0.0121 W

So we get the total power dissipated in the resistors as
0.0005 + 0.0121 = 0.0126 W or 12.6 mW.

Coincidence?

QuoteThis result is very sensitive to the Phase Shift in part 1, so we really need a good measurement of that.

Quote from: TinMan on 2017.04.30, 10:30:45
...any current that flows through a resistance creates heat,and that is dissipated power.

Quote from: verpies on 2017.04.30, 11:29:45
Yes, and current is always in phase with voltage in an ideal resistor, so power factor is not applicable to calculating power dissipated by such resistors.

However, power dissipated in input resistors is not the same as input power....even if all of the electric input energy is delivered through these resistors.  See MPTT (https://en.wikipedia.org/wiki/Maximum_power_transfer_theorem).
If a load is purely resistive then the power dissipated in that resistance is the same as output power - but that is a special case.

Quote from: TinselKoala on 2017.04.30, 11:52:07
Brad, can you confirm that your FG "black" output lead is not grounded back through the chassis and to the mains ground pin?

Also, my estimate of 72 degrees phase shift is just that: an estimate based on the scope's graticule markings, the trace zero crossings and peaks.  Also, your coil has relatively few turns of pretty heavy wire I think, so your coil's DC resistance is probably very small. Can you get a reading of that somehow? A  more careful estimate of the phase shift, and including the coil's DC resistance, may make some difference in the calculations.


. Maybe Verpies can help out here. Really, what we are calling Input power should actually be considered the total power dissipation of the load circuit, and I think my calculations breaking down the power dissipations of the various parts of the circuit show that this is correct.

QuoteBrad, can you confirm that your FG "black" output lead is not grounded back through the chassis and to the mains ground pin?

QuoteSo we are really back to square one, since we don't actually know the impedance of the power source (the FGs we are using). I'd like to believe that my ElCheepo DDS is 50 ohms impedance but I don't think it actually is.

QuoteI'm not sure how the Max Power Transfer Theorem affects these calculations or conclusions, except that the true Input Power is likely to be larger than what we are calling Input Power

Quote from: TinMan on 2017.04.30, 12:33:35
TK
I have always stated that my FGs black lead is grounded,and so is the scopes ground lead's.
This is why i have to use my inverter to run my scope,so as i get isolation between the two common ground's of the scope and FG.

With my FG,you can either choose 0 or 50 ohms impedance.

Unless the power factor is 1,then i think the real power value will be less than the apparent power we are measuring.

I carried out a load test today,using a simple load.
The load is a FWBR,with a cap across the output of the FWBR,and an LED as a load(also across the cap)

With the FG placed on the AC input of the FWBR,i can get 2.82 volts across the cap/LED-BPC circuit not in play.
This is the maximum power the FG can deliver to the load(LED)

If i then hook the FG leads to the BPC circuit,minus the input resistor R1(R2 still in play-10 ohms),and then place the AC inputs of the FWBR at a certain turn of L2,and the other at the input of L1,i can get 2.91 volts across the cap/LED,and 700mV across R2

Not only is the dissipated power greater in the LED,when the BPC is in play,we are also dissipating 49mW of power from R2. But this is only when one side of the FWBR is on a certain turn of L2--i think it was the 3rd or 4th turn in from the outer turn.
If i go a turn either side of the correct one,then the power being delivered to the load drop's to a value below that of what the FG it self could deliver to the load.

Thought that was interesting.
Will shoot a video tomorrow after work,and post here.


Brad

Quote from: TinMan on 2017.04.30, 12:10:15
It dose however,give you the input current.

Quote from: TinMan on 2017.04.30, 12:33:35
The load is a FWBR,with a cap across the output of the FWBR,and an LED as a load(also across the cap)

Quote from: TinselKoala on 2017.04.30, 19:53:17
Partzman, I don't have the right kind of dual-strand wire to make a coil like yours for testing. I do have precision 1 ohm 1 percent non-inductive current sense resistors (also have 0.1 ohm) Ohmite WN series type.

I took the liberty of making explicit the ground connections in your circuit diagram; I hope this is a correct reflection of how you had things connected in your tests. Please let me know. As you know, even a few inches of wire in probe or other ground connections can have enough inductance to throw off results.

I am also not sure that measuring input and output current through the same resistor at the same time is a legitimate measurement, and I am somewhat troubled by the negative phase angle. Does this negative angle mean that you are testing at a frequency that is higher than the resonant frequency of the coil?

Quote from: TinselKoala on 2017.04.30, 19:53:17
But the diagram indicates...

Quote from: TinselKoala on 2017.05.01, 04:59:16
IN:
Math "average" computed by scope is 1.60 mW. 

Quote from: verpies on 2017.05.01, 08:16:31
I don't see any sensing of the current delivered by the AC voltage source to the entire circuit ( including L1 ).  Wouldn't that be prudent?
Multiplying this input current by the AC voltage of the source would yield the true power delivered by this AC voltage source.

Quote
Do you mean the average of the Math channel AVG(Ch1*Ch2) ?   I did not see that on the scopeshot...

Quote from: partzman on 2017.04.30, 20:40:53
TK,

No need to wind a special coil although you are welcome to do that.  I was hoping you or TM could try you existing TBC's with my schematic layout.  The operating frequencies should be lower and it will be interesting to see the results.

The ground connections are correct as you show.  I have a coax cable from the signal generator (set at 50 ohm impedance) which connects to a BNC male and then short leads for all connections.  I am not using a probe and clip on the 1 ohm sense resistor but rather I remove the probe and use just the inner probe tip and a supplied Tek spring contact placed on the ground sleeve.  This probe connects directly across the TO-220 packaged sense resistor.

I looked at the data sheet for the Ohmite resistors and with an inductance <1nH @1MHz, they should be fine.  The 50 ohm load resistor should also be non-inductive as you know.

The output current measured in the sense resistor is the same as the input current drawn from the generator source.  This can be confirmed with a current probe placed in the input line to the transformer assembly.  KCL also demands they be the same unless there is an outside source of energy or current entering the circuit which in this case there is none.

The negative phase angle is due to my specifying that the phase measurement be taken from CH1 to CH2.  If I had stated CH2 to CH1, the angle would be positive but in either case, the current leads the input voltage.  This is a result of the fact that when the primary is driven at one end while the other end is open, the generator sees a reactive load due to the distributed capacitance between the primary and secondary.  The resulting displacement currents create induction in the loaded secondary.

Edit: OOps, I see I missed your last question on the wiring.  No, this is not connected in the standard Tesla manner.  The two windings are separated electrically and the dots indicate the start of the windings or the inner connections.  So, the generator connects to the start of L1 with the other end left open.  The start of L2 then connects to the top of the load resistor R1 with the finish connecting to the junction of R1 and R2.  If these connections are reversed, the operating characteristics will be completely different!

Pm

Quote from: partzman on 2017.04.30, 19:24:27
TK and TM,

I'm curious to see if your Tesla bifilar coils will produce a COP>1 when following the attached schematic if either of you or anyone else has the time.  The coil polarity is important and a 50 or 100 ohm load could be used.  Select a 1 ohm for current sensing that is non-inductive and 1% in tolerance if possible.  When connected as shown, find the resonant frequency and then take measurements at a frequency that is 60-66% of the resonance frequency.

The test coil used below is a 33 turn bifilar wound with 2-28awg ribbon cable on a form that has dimensions based on the golden mean ratio.  I have not had success with round coils for this type of induction.  A difference in the windings (apart from lower inductance and capacitance) is these are side by side where yours are planar.

The scope pix BPT_1 shows the basic measurements which reflect an input power of 29.33mw and an output power of 1.493^2/51 = 43.7mw for a COP = 1.49.  These measurements were taken with a 64 sample average and a record length of 1e6 points. 

Pix PBT_2,3 allow one to check the pin accuracy against the measured phase angle between CH1 (input voltage) and CH2 (output/input current).

Pm

Quote from: web000x on 2017.05.01, 11:16:10
Have you tried pulsed DC with this setup to see if sharp transients produce more pronounced COP>1 results?

THanks,

Dave

Quote from: Itsu on 2017.05.01, 10:29:09
Here my measurement ( at 1.4MHz) on a Bifilar Pancake coil (no solonoid like TK used).

I used my current probe to measure / calculate the input green current, (current controller at 100mA/Div.) yellow is FG voltage.
So input power is 256mW, see screenshot 1


Output power measured to be 187mW using the math and the blue / purple traces, see screenshot 2.

When i calculate the output power like TK did CH3 (purple) Vrms² / 51,   i get (3.274² / 52 (in my case) = 206mW


Food for thought.

Regards Itsu

QuoteI can't seem to get a COP that is NOT "overunity"

Quote from: TinMan on 2017.05.01, 09:35:03
Also,is not the schematic missing something?--see modified schematic below.

Quote from: verpies on 2017.05.01, 11:51:53
I would think so.
Also the CSR (R2) will not "catch" the current flowing through those capacitances you've added.  This allows some of the input energy to appear at the output without being accounted for.

To fix this measurement error the sensing of the current delivered by the AC voltage source to the entire circuit ( including L1 ) should be added.

Quote from: TinMan on 2017.05.01, 11:55:58
How can that be,as all the input energy must go through the 1 ohm R2.

Quote from: verpies on 2017.05.01, 12:29:48
The current flowing from the AC voltage source to L1 is not "caught" by R2 because L1 forms a capacitance with the ground (including the free end).

Quote from: TinselKoala on 2017.05.01, 07:27:11
I can't seem to get a COP that is NOT "overunity" from this thing. Setting FG to 15 v p-p and 1.00 MHz I get:

IN
4.00 x 0.0329 x cos -74.58 = 0.03499 W

OUT
1.46²/51 = 0.04179 W

COP 1.19

TK,

Quote from: verpies on 2017.05.01, 12:29:48
The current flowing from the AC voltage source to L1 is not "caught" by R2 because L1 forms a capacitance with the ground (including the free end).

Quote from: TinMan on 2017.05.01, 13:06:16
The ground is in series with R2...

Quote from: TinMan on 2017.05.01, 13:20:22
L1 forms a capacitance with L2....

Quote from: verpies on 2017.05.01, 08:16:31
I don't see any sensing of the current delivered by the AC voltage source to the entire circuit ( including L1 ).  Wouldn't that be prudent?
Multiplying this input current by the AC voltage of the source would yield the true power delivered by this AC voltage source.
Do you mean the average of the Math channel AVG(Ch1*Ch2) ?   I did not see that on the scopeshot...

Quote from: TinMan on 2017.05.01, 09:35:03
I'll give it a go PM,but my scope(or maybe me using the scope)is not so good for calculating power when phase angles and power factor have to be taken into account.

Also,is not the schematic missing something?--see modified schematic below.
Would this bring any light as to why TK is see all this OU ?.
Looks to me like a current loop,when the missing parts are added.

I did try the !one end of the coil disconnected! thing some time back,and indeed could get a good voltage across a 100 ohm resistor,and light up LED,from the two ends of L2.


Brad

Quote from: Itsu on 2017.05.01, 10:29:09
Here my measurement ( at 1.4MHz) on a Bifilar Pancake coil (no solonoid like TK used).

I used my current probe to measure / calculate the input green current, (current controller at 100mA/Div.) yellow is FG voltage.
So input power is 256mW, see screenshot 1


Output power measured to be 187mW using the math and the blue / purple traces, see screenshot 2.

When i calculate the output power like TK did CH3 (purple) Vrms² / 51,   i get (3.274² / 52 (in my case) = 206mW


Food for thought.

Regards Itsu

Quote from: web000x on 2017.05.01, 11:16:10
Have you tried pulsed DC with this setup to see if sharp transients produce more pronounced COP>1 results?

THanks,

Dave

Quote from: verpies on 2017.05.01, 13:23:16
"Ground" is a point. It cannot be in series with anything...anymore that it can be in parallel with anything.
Only multi-terminal components can be connected in series or parallel.
L1 forms a capacitance with much than L2.  This bypasses R2.
Designate the other side of R2 as the "ground" point and see for yourself

Quote from: partzman on 2017.05.01, 13:34:36
Brad,

OK, great. 

I am not understanding where you see a ground loop (not to say there isn't one) so could you perhaps be a little more specific.  Otherwise the equivalent circuit is as you show.

Pm

Quote from: TinselKoala on 2017.05.01, 13:29:11
Partzman, the solenoid coil is #33 green enamelled magnet wire, 376 turns (188+188) wound on a wooden dowel 12.67 mm diameter. I have not accounted for the DC resistance of the coil halves which is about 5.10 ohms per half.

Quote from: TinselKoala on 2017.05.01, 14:14:23
Another result from my PBF flat pancake coil.

FG set to 15 v p-p, tuned to 502 kHz for approx. -80 degree phase shift

IN: 4.45 x 0.0267 x cos -81.41 = 0.017747 W or 17.7 mW

OUT: 1.34²/50.58 = 0.03550

COP 2.00


So if the fact that my 50 ohm resistor (actually 2ea. 100 ohm 5% in parallel, measuring 49.58 ohms) is not noninductve is the cause of my "success", what's wrong with that? Maybe "Realni OU" (tm TKLabs) requires a little bit of inductance there.

(where's that tongue in cheek smiley anyhow...)

Quote from: TinMan on 2017.05.01, 14:14:51
I was waiting for this to come up.
So,our R1 is directly across L2-->so what is the true value of R1 that we are measuring the voltage drop across.

In my case,L2 has a very low resistance.


Brad

Quote from: TinselKoala on 2017.05.01, 11:40:41
I think you don't have enough phase shift. Try tuning the frequency until you get around -75 to -85 degrees of phase shift between input voltage (yellow) and input current (green).

Quote from: gyula on 2017.05.01, 10:57:10
Hi Itsu,

Sorry to chime in, did you measure the load resistance to be 52 Ohm? 

(and in your formula you first wrote 51 Ohm and then calculated with 52 Ohm,  with 51 Ohm the output power would be 210.2 mW)

Gyula

Quote from: partzman on 2017.05.01, 13:43:28
Itsu,

There appears to be something wrong with the current to input voltage phase in your scope pix for some reason.  The current should be leading the input voltage by ~80 degrees or so and yours appears to be considerably less than that.  Compare your results with TK's post #97 for example.  What are the specs on your coil assembly like the inductance of each winding, inductance of the windings in series, and the capacitance between the windings?

Thanks for doing the test.

Pm

QuoteItsu, what happens if you move your current probe to be right next to the 1 ohm resistor, preferably on the ground side of it?

Quote from: Itsu on 2017.05.01, 14:42:40
Hi Gyula,

my load resistor is 51 Ohm 1%,  the csr is 1 Ohm 1%, totaling 52 Ohm when using the CH 3 (purple) voltage.
TK used 51 Ohm in his formula (50 +1), i adjusted that to 52 in my situation as i have 51 + 1.


Itsu

Quote from: verpies on 2017.05.01, 13:23:16
"Ground" is a point. It cannot be in series with anything...anymore that it can be in parallel with anything.
Only multi-terminal components can be connected in series or parallel.
L1 forms a capacitance with much more than L2.  This bypasses R2.  Also, for input power calculations, the input voltage should not include the voltage drop of R2.
Designate the other side of R2 as the "ground" point and see for yourself

Quote from: TinMan on 2017.05.01, 14:01:33
PM

I never mentioned a ground loop.
I did mention a current loop though.

Brad

Quote from: Itsu on 2017.05.01, 14:38:39
Ok, did that, had to go down from 1.4Mhz to 190Khz to get -80° phase shift, see screenshot

The input now calculates to be 22.5mW, see red math trace (current controller set to 10mA/Div., so similar as to the green vertical setting on the scope).
Output calculates to be 871mV² / 52 = 14.5mW

Be aware i measure the input current with my current probe at the red FG lead!

My 50 Ohm resistor is 51 Ohm 1%,  my 1 Ohm resistor is 1 Ohm 1%.

Itsu

Quote from: TinselKoala on 2017.05.01, 17:20:51
But I'm using 1/4 watt resistors for the 49.58 ohm "load". Are they getting warm or not? Yes, they are perceptibly warming very slightly by touch. I have no way of measuring this though.

Quote from: gyula on 2017.05.01, 17:37:02
Hi Partzman and all,

I wonder whether you have already considered to test your bifilar coil circuit to embed into a looped circuit?

IF not yet, then I would propose to build a simple one transistor oscillator that would have a normal output coil inductively coupled to the main oscillator LC tank circuit and would feed your bifilar coil circuit as the FG does now. 
And at the output of your bifilar coil circuit a full wave bridge rectifier (made from say Germanium diodes) would drive a DC voltage stabilizer, this latter should have very low idling current of course. Then the output of this DC stabilizer would drive the one transistor oscillator. Such simple looping would eliminate any measuring issues... if there are any, that is. 

The DC voltage level for the one transistor oscillator could easily be chosen to be very close to the output DC level of the stabilizer (chosen strategically) so that a linear, micropower regulator could be used without too much loss.

With the COPs around 1.5 to 3 or so as the present measurements indicate, such looping should be possible and even if the power levels involved are in the some ten to some hundred mW range, that would be enough to maintain the operation of the looped circuit, perhaps even a dimly lit LED could be run from it.

Gyula

Quote from: TinselKoala on 2017.05.01, 17:37:25
Partzman, I am assuming that the corrections you calculated up above being small, will also be small in my setup. Do you have enough information to calculate the corrections for one of my datasets?

Also, what effect do you think my casual probing and my non-noninductive resistor might be having on my results? I would like to have some confidence that I am seeing the same effect you have seen but I'm worried that my results may be artifacts, whereas yours may not be, since you are using tighter probing and you have your proper resistor.

I'm happy that the effect can be reproduced even if it's not due to the same cause... but I think it probably is. However this brings up some other issues. Why did you not get similar results when you tried round coils yourself, but only got them with the Golden Mean proportions? Can Itsu reproduce the results using his current probe, wherever placed?  And of course we have the issue of whether or not our probing scheme is actually valid for this circuit. At present I am just looking at data, I have not started trying explain it or to find errors if they are there.

I've got to say though, that I haven't had this much fun in the lab in quite some time!   >:-)

Quote from: partzman on 2017.05.01, 16:20:38
Itsu,

Let's start with the deskew of your current probe which your scope will allow but may be limited in range.  A good check of this since you have a non-inductive 1 ohm sense resistor and you also should have the HF spring contact to use on your scope probe for a direct contact across the sense resistor, place your current probe on a short lead between the load resistor R1 and the top of sense resistor R2 and compare the current probe trace with the scope probe trace and adjust the deskew to align the traces if possible.  If you are using a current probe amplifier, set it to the highest bandwidth possible.

If the traces do not align, simply use the voltage across the sense resistor R2 in your Math calculation.

I just re-checked your scope pix on this post and I see that your measurements are taken across the entire screen which shows and incomplete number of cycles which will affect your overall measurements.  You can either adjust the frequency for a precise number of cycles or turn on your vertical cursors and activate measurements between cursors.

Pm

Edit: I'm curious if you can achieve a COP>1 at this frequency!

Quote from: gyula on 2017.05.01, 17:37:02
Hi Partzman and all,

I wonder whether you have already considered to test your bifilar coil circuit to embed into a looped circuit?

IF not yet, then I would propose to build a simple one transistor oscillator that would have a normal output coil inductively coupled to the main oscillator LC tank circuit and would feed your bifilar coil circuit as the FG does now. 
And at the output of your bifilar coil circuit a full wave bridge rectifier (made from say Germanium diodes) would drive a DC voltage stabilizer, this latter should have very low idling current of course. Then the output of this DC stabilizer would drive the one transistor oscillator. Such simple looping would eliminate any measuring issues... if there are any, that is. 

The DC voltage level for the one transistor oscillator could easily be chosen to be very close to the output DC level of the stabilizer (chosen strategically) so that a linear, micropower regulator could be used without too much loss.

With the COPs around 1.5 to 3 or so as the present measurements indicate, such looping should be possible and even if the power levels involved are in the some ten to some hundred mW range, that would be enough to maintain the operation of the looped circuit, perhaps even a dimly lit LED could be run from it.

Gyula

Quote from: Itsu on 2017.05.01, 18:59:35
After a fresh calibration, these are the signals from my current probe (green) placed between the resistors R1 and R2 (17.36mA rms) and the yellow
probe placed across the R2 resistor (19.36mA rms).

This is with the same setting from my FG (20Vpp sine wave @ 190KHz) as used before.

The yellow signal is kind of weak, so the between vertical cursor test is not very stable.

Itsu

Quote from: Itsu on 2017.05.01, 19:34:25
Using a different TBP coil (with similar properties) i redid my measurements using the same 51 Ohm and 1 Ohm resistors.

Input calculated by the scope/current probe is 22.7mW.
Input calculated by using the R2 current is: 6.81V rms * 0.0189mA rms * cos -80° =  22.3mW
See screenshot 1

Output calculated by scope is 15.4mW
Output calculated by Ch3 (purple)² / 52 Ohm = 0.8682² / 52 =  14.49mW
See screenshot 2

So this different coil also shows via 2 different methods a similar cop<1 as the first TBP coil.


Itsu

Quote from: ION on 2017.05.01, 19:38:30
Looping is always the final proof. I have explored circuits as Gyula proposed and it is one good way, but can be too lossy if COP's are not high enough to account for all losses in the loop. As Gyula states you will need some COP of 1.5 to 3.0 with perhaps a minimum of 1.2 to 1.3 if you are careful with your DC to DC converter design. I have outlined some efficient looping schemes on my bench.

But there is another less lossy way, and that would involve direct looping with appropriate phase tuning network from output to input. Also a 1:1 common mode choke can be used if isolation is required, the important thing is that the output phase be tuned to exactly correspond to the input phase, and make sure the phase tracks as parameters change slightly when a load is applied that causes the wire to heat up and add extra resistance.

Hypothetically, a passive system that runs with a very small power gain could start on thermal noise, or the swipe of a magnet, and not even need a signal generator for startup.

Far fetched? I think SM would disagree. In one of the videos, we see SM waving a speaker magnet around one unit that wouldn't quite start up while DR. Schinzinger observes.

Quote from: partzman on 2017.05.01, 20:04:53
Itsu,

I'm really not surprised you are not seeing a COP>1 at this low frequency ~200kHz.  If you raise the input frequency however, the COP should increase until you reach COP>1.  This should occur without the phase angle decreasing as seen on your previous test but if not, I'm not sure as to what the reason might be at this point!

Pm

Quote from: Itsu on 2017.05.01, 20:32:11
When increasing the frequency and to obtain again a -80° phase shift between input voltage and current i have to go all the way to 26MHz.
Doing again the measurements there i get:

input measured by current probe/scope 7.689mW
output Ch3 (purple)² / 52 = 1.26² / 52 = 30.5mW
Cop = 3.9

But i really do not trust these figures at this frequency

Itsu

Quote from: TinMan on 2017.05.01, 13:56:29
You cannot make a length of rope tight by pulling on one end,while the other end is not fixed (grounded)

Quote from: TinMan on 2017.05.01, 13:56:29
All the current flowing from the source(FG),must pass through R2-there is no other path for the current to flow through.

Quote from: TinMan on 2017.05.01, 13:56:29
If you remove R2,and connect the ground to R1,then you have just altered the circuit.

Quote from: partzman on 2017.05.01, 20:10:03
ION,

Could you give an example of how this phase tuning might be achieved.  There is also normally a difference in output voltage level that is lower than the input voltage so this would have to be compensated for as well.

Thanks.

Pm

Quote from: verpies on 2017.05.01, 21:22:18
Untied rope does not behave like an untied coil.
There is - through displacement current.
With that change (shown on the lower schematic below), only the polarity of the Ch2 signal will be inverted and the voltage drop of R2 will not be included in the input voltage.
This probe placement constitutes an improvement of the power measurement technique - not an alteration of the working circuit.

Quote from: ION on 2017.05.01, 21:30:44
I'm working on some schematics for this. If the voltage is lower than input, then the current must be higher if it truly has a COP=>1. There are a few ways to do a passive loop back. I'm working on using a matching transformer or autotransformer with slight step up for V so the output can pump current into the input and also adding a phase correction device.

I haven't been following too closely  the tests you guys have been doing , as I was working on the other project. How much phase correction is actually needed for the highest COP device? And which direction?

What I was getting at is this: a DC loopback is not really needed for an AC system. You don't need to go through all those conversions. If the system indeed has an AC power gain, it can be looped in the AC domain if you are careful about phase and provide a correct impedance match, as long as the input and output frequency are the same. This should have much less loss as it would only involve a small variable capacitor or variable inductor depending on the requirement of lead or lag.

I'll post some schematics soon.

Regards

Quote from: partzman on 2017.05.01, 21:52:36
The lower schematic is a cleaner connection layout for measurements but the signal generator must have isolated ground or an isolation transformer could be used on the input.

Quote from: partzman on 2017.05.01, 18:01:26
Gyula,

I have considered looping the device in the past but have not done so however, your idea certainly has merit and should be given a try.

Pm

Quote from: gyula on 2017.05.01, 22:35:35
Hi Partzman,

One more question if I may:  do you think your circuit would behave in the same way if the output would be taken from a third coil, L3,  coupled to L1 & L2?  I mean making a trifilarly wound coil instead of the bifilar one.  This needs to be tested of course if you or someone else have not done so.

EDIT:  I may have asked a silly question...  because the "upper" end of L2 should be dealt with? and it may modify circuit behaviour.

Thanks,
Gyula

Quote from: partzman on 2017.05.02, 00:42:18
Gyula,

Not a silly question at all. 

In all my testing I have used sandwiched pancake coils from 2 to 7 in number.  In a 3 coil arrangement for example, when the center coil is the primary, the distributed capacitance is double the 2 coil and the inductance varies depending on whether the secondaries are in series or parallel.  With 5 coils you can have 2 primaries with 3 secondaries or 3 primaries with 2 secondaries, etc.  These combinations allow for various ratios of L/C to be achieved depending on the frequencies one desires.  The secondaries can also be cross coupled, shorted, resonated, loaded, or in combination.

Pm

Quote from: TinselKoala on 2017.05.02, 01:58:54
However I don't even rely on the math trace of my scope any more,

Quote from: TinselKoala on 2017.05.02, 01:58:54
I perform the calculations of input power manually using V p-p x 0.3535 to give Vrms,

Quote from: TinselKoala on 2017.05.02, 05:34:20
...since cos(phase angle) tends rapidly towards zero as (phase angle) approaches + or - 90 degrees....

Quote from: TinselKoala on 2017.05.02, 03:01:07
Just to be sure I'm not making some silly error in connections, I've enlarged the connection area of my test setup. Could you all please take a close look and see if I've done anything wrong with the connections? Comments and questions appreciated.

Quote from: TinselKoala on 2017.05.02, 05:34:20
Now... since cos(phase angle) tends rapidly towards zero as (phase angle) approaches + or - 90 degrees.... and one can apparently "dial in" pretty much any phase angle by selecting frequency.... what does this mean? Can you arbitrarily set your COP simply by selecting a frequency that gives a phase angle approaching 90 degrees? What if I set a frequency that gives exactly -90 degrees of phase shift? Does this mean input power goes to zero and COP goes to infinity?

Furthermore, what accounts for my changed results now that I am using a lower resistance and non-inductive resistors for R1? Is the fact that it is much harder to get COP > 1 due to lower resistance, or lower inductance, or both? If stray or parasitic inductance is responsible for the COP > 1 effect.... well, assuming the OU is real, what's the problem? Maybe "realni OU" requires this added inductance, stray or parasitic or deliberately introduced. Or... assuming the OU is not real... then can _all_ of our COP>1 results be attributed one way or another to stray or parasitic inductances introduced by even the best probing methods, or coil layout, or other factors we haven't considered?


"They are just questions, Leon...."

http://www.youtube.com/watch?v=Umc9ezAyJv0

Quote from: partzman on 2017.05.02, 00:42:18
Gyula,

Not a silly question at all. 

In all my testing I have used sandwiched pancake coils from 2 to 7 in number.  In a 3 coil arrangement for example, when the center coil is the primary, the distributed capacitance is double the 2 coil and the inductance varies depending on whether the secondaries are in series or parallel.  With 5 coils you can have 2 primaries with 3 secondaries or 3 primaries with 2 secondaries, etc.  These combinations allow for various ratios of L/C to be achieved depending on the frequencies one desires.  The secondaries can also be cross coupled, shorted, resonated, loaded, or in combination.

Pm

Quote from: Itsu on 2017.05.02, 06:49:20
TK,

as far as i can see, i have the exact same setup, providing that the red winding is the most outer to start with (can't really see that).

As i have used 2 similar TBP coils, i could start using different resistors to see if there is a change there like you noticed.

Itsu

Quote from: gyula on 2017.05.02, 10:38:55
Hi Partzman,

Thanks for your additional info. 

I have devised a schematic to show how I mean to embed your bifilar coil circuit into a looped circuit, see attachment.  Of course, several solutions exist for the one transistor oscillator or to use an off the shelf micropower voltage regulator or a DC-DC switch mode converter.
In the oscillator I choose to work around 1 MHz, a 500 pF variable capacitor, labeled as C, can be used to tune the LC tank and find COP>1 frequencies.  AM pocket radio coil components like oscillator or IF stage coils can be adapted here.
In the lower left corner of the schematic I indicated to chain two bifilar coils in a hope it may be able to further increase output power, it is possible that a tuning capacitor would be needed across the second bifilar to help some matching or influence overall phase shift.

Maybe it is still too early to devise such looping circuit but there is a new replication at overunity.com
http://overunity.com/17186/the-bifilar-pancake-coil-at-its-resonant-frequency/msg505815/#msg505815
so things are promising...   :D

Thanks,
Gyula

Quote from: TinselKoala on 2017.05.02, 10:57:32
That's a good idea. I think the first thing to do would be to test whether the output will still work if the load resistor is removed and the voltage doubler part of the circuit is connected across L2 instead, still using the FG as input oscillator. That part can be tested pretty easily, probably. I've got plenty of 1n34a and 1n60 diodes hiding in my stash (also some small Schottkys in there too I think).

Quote from: ION on 2017.05.02, 01:54:15
To further complicate the number of variables, you could also treat the "pancakes" in the "coil sandwich" as elements of a capacitor compression type trimmer.....who knows how changing the spacing slightly between the sandwich layers will affect the output?

And while were at it, how do we know what type of spiral would give better results? Would varying the pitch of the spiral tune the layer to layer capacitance in a positive way? There are numerous spiral types.

https://en.wikipedia.org/wiki/Spiral

Regards

Quote from: TinselKoala on 2017.05.02, 03:01:07
Just to be sure I'm not making some silly error in connections, I've enlarged the connection area of my test setup. Could you all please take a close look and see if I've done anything wrong with the connections? Comments and questions appreciated.

Quote from: partzman on 2017.05.02, 12:20:10
ION,

When the flat pancake coils are separated by a dielectric such as polycarbonate (K~4) there will be characteristic changes to the overall performance that I have not totally understood yet.  I have assembled asymmetrical 3 coil devices wherein only one dielectric separator was inserted or two were inserted of different thicknesses.  Spirals would add yet another variable to the mix !!

Pm

Quote from: TinMan on 2017.05.02, 12:40:43
Because polycarb holds a static charge very well.


Brad

Quote from: TinselKoala on 2017.05.02, 05:34:20
Now... since cos(phase angle) tends rapidly towards zero as (phase angle) approaches + or - 90 degrees.... and one can apparently "dial in" pretty much any phase angle by selecting frequency.... what does this mean? Can you arbitrarily set your COP simply by selecting a frequency that gives a phase angle approaching 90 degrees? What if I set a frequency that gives exactly -90 degrees of phase shift? Does this mean input power goes to zero and COP goes to infinity?

Furthermore, what accounts for my changed results now that I am using a lower resistance and non-inductive resistors for R1? Is the fact that it is much harder to get COP > 1 due to lower resistance, or lower inductance, or both? If stray or parasitic inductance is responsible for the COP > 1 effect.... well, assuming the OU is real, what's the problem? Maybe "realni OU" requires this added inductance, stray or parasitic or deliberately introduced. Or... assuming the OU is not real... then can _all_ of our COP>1 results be attributed one way or another to stray or parasitic inductances introduced by even the best probing methods, or coil layout, or other factors we haven't considered?


"They are just questions, Leon...."

http://www.youtube.com/watch?v=Umc9ezAyJv0

Quote from: gyula on 2017.05.02, 10:38:55
Hi Partzman,

Thanks for your additional info. 

I have devised a schematic to show how I mean to embed your bifilar coil circuit into a looped circuit, see attachment.  Of course, several solutions exist for the one transistor oscillator or to use an off the shelf micropower voltage regulator or a DC-DC switch mode converter.
In the oscillator I choose to work around 1 MHz, a 500 pF variable capacitor, labeled as C, can be used to tune the LC tank and find COP>1 frequencies.  AM pocket radio coil components like oscillator or IF stage coils can be adapted here.
In the lower left corner of the schematic I indicated to chain two bifilar coils in a hope it may be able to further increase output power, it is possible that a tuning capacitor would be needed across the second bifilar to help some matching or influence overall phase shift.

Maybe it is still too early to devise such looping circuit but there is a new replication at overunity.com
http://overunity.com/17186/the-bifilar-pancake-coil-at-its-resonant-frequency/msg505815/#msg505815
so things are promising...   :D

Thanks,
Gyula

Quote from: partzman on 2017.05.02, 12:22:21
TK,

Your connections are correct as far as I can see.

Pm

Quote from: TinMan on 2017.05.02, 12:43:09
TK

Just out of curiosity ,can you re-do you P/in P/out measurements ,with  R2 in the position of the circuit  below.


Brad

Quote from: TinselKoala on 2017.05.02, 13:37:28
Good, thank you for checking.

Do you think it makes a difference if the "Red" and "Blue" windings swap roles? That is, I have been using the Red winding for input, one end open, and the Blue winding for the "load" and output. So will it make a difference if I now use the Blue winding for input, one end open, and put the "load" and output on the Red winding instead?

Quote from: TinselKoala on 2017.05.02, 13:53:23
Partzman, yes, I've seen negative phase angles greater than -90 degrees with the present configuration and these of course give "negative" power input figures.

Yes I get your point about using cursors, but unfortunately my scope's cursors do not report phase angle directly, they only give time and voltage values. I'll try to figure out how to derive phase angles from the time differences at zero-crossings and the frequency,  but it may be only as accurate as the scope's own computations anyway.

Oh, yes, I know about large reactive power ratios. Some of my power transmission circuits would blow the minds of the casual scoposcopists because of their high reactive power in the transmission loops.


Another thing I'll be trying later tonight is to sub in a 10 ohm non-non-inductive resistor for the 9.4 ohm NI presently installed. Will that restore the ease of obtaining COP>1? I wish I could find or make a 50 ohm NI, but no joy there.

Quote from: ION on 2017.05.02, 12:59:38
Gyula's approach looks promising, I like the idea of the tuned secondary, however I question the need for the linear regulator LM317, unless it is just for runaway protection. O0 Also a push pull Mazili type oscillator with resonant secondary may prove to be more efficient.
-------------------------------------------------------------------------------------------------------------
May I submit some ideas for looping the device that keeps it in the AC domain rather than all the losses of the AC to DC then DC to AC techniques. Phase correction network is to be determined but should not be difficult.

I have included only three elementary phase compensation schemes, however the number of possible arrangements of passive devices for the network is very large.

With such an approach, the COP need not be very high, as L's and C's of high quality can provide very low insertion loss for the phase correction.

Quote from: TinselKoala on 2017.05.02, 15:31:06
Whoopie! I just figured out how to get the Rigol to report phase angle using the cursors! Using Options deep within the Cursors menus, not documented in the User Manual as far as I can tell.
There are still more Cursor functions and options that I don't yet understand, but it seems that the scope's usefulness evaluation has just jumped up another increment! 

In the scopeshot below the Automatic Measurement is reporting -122.8 degrees and the manual Cursor measurement is reporting -119.6 degrees.

Quote from: partzman on 2017.05.02, 14:13:41
You might try both of your NI resistors in series to lighten the load on the secondary and see what change that gives you.
Pm

Quote from: TinselKoala on 2017.05.02, 16:22:37
So.. using the cursor value for phase angle...

F43 FG at 2.86 MHz....

IN: 2.77 x 0.583 x (cos-83.55) = 0.1814 W or 181.4 mW
OUT: 1.51²/9.5 = 0.2400 W or 240.0 mW
COP  1.32

Quote from: partzman on 2017.05.02, 16:24:31
TK,

That's great!  I'm sure other Rigol users would like to know about that.

Also, that phase angle is huge!  The max I've seen on my testing was ~110 degrees as I recall.  The Pin should well into the negative range.

Pm

Edit:  OK, I think this test was not on an operating coil, right?

Quote from: partzman on 2017.05.02, 16:32:59
Those are nice clean appearing measurements.  I like the way Rigol displays a snapshot of all 4 channels simultaneously whereas TEK shows more than you ever need to know on 1 channel at a time.

Pm

Quote from: partzman on 2017.05.02, 17:04:58
Below is a pix of some of the various coil arrangements I have used that all yield COP>1 except for one, that is the coil in the upper right hand corner.

Quote from: evolvingape on 2017.05.02, 17:28:03
What measurement protocol was used to attain COP>1 observations ?

Quote
Were the COP>1 claims based on standard mathematical calculations that suggest COP>1 performance ?

Quote
Is there a device you can demonstrate that shows a COP>1 result and can be easily replicated by the community ?

Quote
Good stimulation!  O0

Quote from: TinselKoala on 2017.05.02, 16:43:58
Yes, it was! I get those large angles at lower frequencies.

Quote from: ION on 2017.05.02, 12:59:38

Gyula's approach looks promising, I like the idea of the tuned secondary, however I question the need for the linear regulator LM317, unless it is just for runaway protection. O0 Also a push pull Mazili type oscillator with resonant secondary may prove to be more efficient.
...

Quote from: Grumage on 2017.05.02, 19:36:55
Is this any use?

http://www.ebay.co.uk/itm/10m-Professional-Quality-Pure-Copper-OFC-4-x-1mm-Blue-Bi-Wire-Loud-Speaker-Cable-/332043613736?hash=item4d4f577628:g:73IAAOSwImRYOCxf

Quote from: Itsu on 2017.05.02, 21:18:10
I toke the liberty to make a video of my coil and measurements, perhaps someone can spot what i am doing wrong here.
Sorry for the rather long video:  https://www.youtube.com/watch?v=OsvJNlYhNUM

As i have made measurements of 2 different coils with a set of 2 different resistors with similar (negative) outcome, i only can conclude for myself that
it behaves different as the other (already 3 i understand) replicators which all 3 report COP>1 results.

If no-one can point me to an obvious problem or error i make, i think it would be best that i step out of this thread so you guys can concentrate on exploiting
this effect by trying to further enhance it and even loop it.

Good luck.

By the way, i swapped over what TK calls his "red" and "blue" windings, and did not note a major change there.

Regards Itsu

Quote from: TinselKoala on 2017.05.03, 06:12:33
Another: F43 FG at 1.255 Mhz,  Scope auto phase angle -73.8, cursor phase angle -74.17, R1 = 18.8 ohm NI, R2 = 1.00 ohm NI

IN:
Math average 187 mW
Manual average 147.7 mW

OUT:
using (7.4V_p-p x 0.3535)²/19.8 = 0.3456 W

COP 2.35 (manual average In) or 1.84 (scope math average In)

Quote from: forest on 2017.05.03, 06:54:55
I'm not experienced in electronics like you are  TinselKoala, though it still surprise me why you all dismiss the thermal measurements on resistors (placed outside the RF from pancake coils) as the simplest way to calculate average power dissipated without not reliable scope measurements in RF field.Any reasons ?

Quote from: TinselKoala on 2017.05.03, 07:22:02
Yes, at such low power levels it is extremely difficult to obtain accurate and precise enough temperature measurements without doing some pretty elaborate calorimetry using fixed-loss-to-ambient techniques and very sensitive and precise thermometers.  In addition, this circuit is very sensitive to stray and parasitic inductance, so the lead wires to external resistors in a calorimeter will add significant inductance and alter the performance of the circuit in unpredictable and perhaps unquantifiable ways.

The output measurements here are probably fairly reliable anyway; where the difficulty lies is in the _input_ power measurements.  Since the input power is provided by a function generator it is more difficult to use, say, a capacitor charged with a known amount of energy to provide the input. It could be done though, for example with a preamp stage clocked by the FG and providing power to the circuit, through a bipolar transistor or a mosfet, from a charged supercap. 

Perhaps more importantly, we still have not resolved the issue of whether it is legitimate and proper to use the current through R2 as the correct input current measurement. We have several opinions that it is not proper to do so. Alternate means of measuring the input current should be tried and qualified.

Quote from: TinMan on 2017.05.02, 12:43:09
TK

Just out of curiosity ,can you re-do you P/in P/out measurements ,with  R2 in the position of the circuit  below.


Brad

Quote from: Itsu on 2017.05.02, 21:18:10
If no-one can point me to an obvious problem or error i make, i think it would be best that i step out of this thread

Quote from: TinselKoala on 2017.05.03, 09:03:07
Brad, in one test I have running right now, I put a 1.00 ohm NI resistor in your location and a CH4 probe on the coil side of this resistor. I still have the 1.00 ohm in the original position too, on CH2. Unfortunately I seem to have hit another Rigol bug because I can't get a good Math CH1-CH4 differential reading, but using cursors I see that CH1 p-p is at 22.6 Vp-p and CH4 is at 22.2 Vp-p for a difference of about 0.4 V which translates to about 0.4 A p-p through the new resistor. I can't get more precise than that because of the cursor readouts and the high voltage to begin with. The current through the "old" resistor using CH2 at the same time is almost exactly the same at 394 mA p-p using cursors or 388 mA p-p using the scope's auto measurements. As near as I can tell the phase angle of the CH1->(CH1-CH4 difference) is also very close to the phase angle of the CH1->Ch2 measurement.

Quote from: TinMan on 2017.05.03, 09:47:09
So you are seeing a 400mA difference at each end?
Im not sure where the original 1.00 ohm resistor was in your circuit.


What i was looking for,is there a difference in input current value,between having the CVR at the ground side of the circuit-as PM has it,and having it at the input end of the circuit-as MH said it should be.


Brad

Quote from: TinselKoala on 2017.05.03, 10:03:55
OK... Now I've checked the Math CH1-CH4 subtraction values against the input traces, and the Math value IS giving me the correct figures for the difference between the traces, just at a coarse resolution. 

So I get a RMS value of about 880 mA rms on the Math trace for the differential CH1-CH4 (current through R2b at TinMan's position) , against a CH2 measurement of 138 mA rms (current through R2a at original position) . Phase angle is about the same.

So if this is correct, this brings down the COP to well below 1.

Quote from: TinselKoala on 2017.05.03, 11:45:25
Actually, I think the problem is with the differential measurement across R2b. As you can see from my scopeshot above there is a tiny phase difference between the CH1 signal and the CH4 signal. This is what produces the phase shift and the voltage amplitude in the Math trace. I think the differential measurement across the CVR to give a voltage drop that can then be turned into a current... is only valid if there is _no_ phase shift between the signals at the two probes on either side of the resistor. I think.

SO here I've twiddled my scope to delay CH4 a bit so that the tiny phase shift between CH1 and CH4 is removed, see below. This brought the Math differential measurement CH1-CH4 down to reasonable values that are in the ballpark of the CH2 R2a measurement.  It also removed the phase shift from the Math trace as well. The 1.20 Vp-p reported is affected by the coarse resolution and should probably be more like 600 mV p-p or even less. Which brings it almost equal to the CH2 current from R2a. Is this cheating, or valid, or not? Let's hear from some more experienced scoposcopists on this one.

How to interpret all of this? Too far past my bedtime to be sure.    :D

Quote from: TinselKoala on 2017.05.03, 12:10:55
This is a really good illustration of how tiny inductances can really screw up measurements. I had my two probes CH1 and CH4 on either side of the R2b resistor, but the CH1 probe was separated from the resistor by about 2 inches of wiring and connectors. SO I moved both probes to be right up against the body of the resistor where the lead wires go in, and I made sure the probe ground clips were symmetrical and to the exact same ground point, and I got a lot better result. Only needed -4 ns on CH4 to align phase, and the Math trace went way down. So I'm saying at this point that I see actually no significant difference in the current between the values read at the two resistors.

That is, if I am correct about the effect of phase of the differential measurement across R2b.

Quote from: TinMan on 2017.05.03, 11:54:25
The way i checked,was simply to swap the polarities of the FG leads across the circuit.
I also seen no difference at all.

Brad

Quote from: ION on 2017.05.03, 01:22:11
Attached is another version using a current fed push pull oscillator similar to Mazzilli and some others. It uses Gyula's idea of the oscillator feed and the doubler.

Could possibly also use a bridge in place of doubler, if the output voltage is a bit higher than input.

I have included an LED surge load that becomes active should the voltage begin to soar much above the PP3 9 volt level. Also has a momentary push to start switch.

The next variation will have fail proof limiter and a crowbar shutdown circuit.

partzman and I were discussing the possibility of building the primary coil as part of the oscillator circuit by adding some taps, This would eliminate external transformer losses.

Values of components will be defined when other parameters are nailed down.

Comments / critiques welcome

Quote from: TinMan on 2017.05.03, 12:22:47
So,here is the next problem TK--you still have not found your !red herring!

No matter what you look for-no matter how hard i look,we both cannot find why the OU result exist's.

Could it be that we have truly found one of Tesla's secrets"

Is anyone going to have a go at picking the measurement error's ?,as we have just eliminated MHs theory.

Where is Poynt when you need him?-maybe even PW ?

I am 2/3rd's through winding a rather long tube bifilar coil,but it is time consuming getting the second layer wire to sit perfectly between two turns of the first layer--but worth the effort-i hope :)

This !effect! dosnt seem to be limited to the BPC,although it would seem that the BPC gives the best result's?

I think it is more due to the voltage difference between the windings of the two coils--but we shall see.


Brad

Quote from: TinMan on 2017.05.03, 12:55:07
OK,carried out the test as you stated TK,and once again,no difference in current value.

Took some pics,instead of doing a short video--hope there clear enough to see the value on the scope.


Brad

Quote from: TinselKoala on 2017.05.03, 12:49:24
Do the resistors heat up enough?

Quote from: Vasik041 on 2017.05.03, 13:00:09
When resistor heat up its resistance change and you can't use Vrms2/Rtotal anymore because you don't know R.

http://www.resistorguide.com/temperature-coefficient-of-resistance/ (http://www.resistorguide.com/temperature-coefficient-of-resistance/)

QuoteTemperature Coefficient (ppm/°C) ±90 for 0.100Ω-0.99Ω, ±50 for1.00Ω-10.00Ω, ±20 for >10.00Ω

Quote from: Itsu on 2017.05.03, 08:38:28
Thanks PM,   Thanks TK,   

I will take another thorough look at my resistors inductance and measurement methods (like TK mentioning using a RF probe tip, which i do not have on my Tek probes, but do on the Owon probes)
Mind you, concerning the use of my current probe/amp combo, i also did the same power calculations with same results using the R2 current, but only at 190Khz, so i will use that further up in frequency too


Itsu

Quote from: TinselKoala on 2017.05.03, 12:49:24
Well, we have apparently eliminated the objection about the location of the CVR at least. However I think we still don't know if the whole probing scheme is the proper way to test this circuit.  Does the Vrms²/R_total calculation accurately give a real output value? Do the resistors heat up enough?
Mine do feel slightly warm but remember I'm using the F43 at over 20 v p-p input.

It's a great puzzler all right. But so far in my testing with the Gyula Doubler on the output, the "ou" effect goes away when that is connected.

Quote from: Vasik041 on 2017.05.03, 13:00:09
When resistor heat up its resistance change and you can't use Vrms2/Rtotal anymore because you don't know R.

http://www.resistorguide.com/temperature-coefficient-of-resistance/ (http://www.resistorguide.com/temperature-coefficient-of-resistance/)

QuoteIs there a chance that the extra energy being seen,is in some way related to electron stripping,via dielectric stress.

Quote from: ION on 2017.05.03, 13:26:55
Hi Vasik
This is very true, however as TK pointed out,  don't think we have to worry about the level of heat that would skew measurements too far for a good quality Manganin or better low TC resistor. Thus far you would be lucky to get a few degrees change in temperature, better if insulated, but not enough to be outside the  tolerance versus temperature band, if e.g. a 1% resistor were used. You could also use a quality 0.1% resistor or better with good TC.

I have some ideas on my bench here:
http://www.overunityresearch.com/index.php?topic=2029.msg29983#msg29983

There are other problems making exacting fractional degree temperature near MHz operating devices that can skew measurements. Eddy current heating of thermocouple junctions and series and common mode filtering of measuring instrument may not be adequate to reject MHz coupling into instrument.

All of these problems can be worked around with a some skill in the art.

One important thing we should be looking very closely at is techniques for efficient looping because that would make very high output power due to power regeneration  become possible. e.g. even a COP of 1.1 or less when efficiently looped would yield enormous power build up due to regenerative effect. The "fire that feeds itself" per SM.

This would then get all power levels well out of the noise and make measurement easy. It would also be the ultimate proof of the device.

Thus far I think the direct feedback with phase control network holds the most promise as it presents a linear load on the output unlike rectification / conversion schemes.

Such a device would need a limiting scheme for the regenerative feedback to prevent thermal runaway and destruction. It could be as simple as a tungsten filament bulb limiter which has non-linear resistance characteristic in the feedback loop along with  crowbar circuit and fuse for ultimate safety.

Brad said:
Interesting idea, since the bifilar pancake is designed to have maximum dielectric stress due to maximum voltage created between turns. And there are other ways to do this without a pancake constructed coil. A cylindrical bifilar coil connected in the same way should have large voltage between turns.

Regards

Quote from: Magluvin on 2017.05.03, 14:25:30
Hey Ion

Been thinking about the looping...   

If we have more currents between the coils, what if we have a transformer primary of very low ohms and inductance and a step up secondary.  My thinking is if the primary is very low ohm and as low in inductance as one is comfortable, then a heavily loaded secondary should reduce the pri inductance even further.  Like when we plug a transformer into the wall outlet, it has min current due to inductance, but as we add and increase the load on the sec the primary is able to take on currents very easily. A problem we wish we didnt have when we think about it.  But here it may work to our advantage. The harder the load, the easier the currents can flow through the primary, of which powers the sec.  Just a thought.



Mags

Quote from: partzman on 2017.05.03, 13:17:16
Itsu,

Yes please re-check the results in and around 1-2MHz using R2 only as I really don't think you will see any gain at 190kHz.  Plus the deskew differential becomes more effective in producing errors as the frequency increases.

Pm

Quote from: Magluvin on 2017.05.03, 14:25:30
Hey Ion

Been thinking about the looping...   

If we have more currents between the coils, what if we have a transformer primary of very low ohms and inductance and a step up secondary.  My thinking is if the primary is very low ohm and as low in inductance as one is comfortable, then a heavily loaded secondary should reduce the pri inductance even further.  Like when we plug a transformer into the wall outlet, it has min current due to inductance, but as we add and increase the load on the sec the primary is able to take on currents very easily. A problem we wish we didnt have when we think about it.  But here it may work to our advantage. The harder the load, the easier the currents can flow through the primary, of which powers the sec.  Just a thought.

Also with making a cylinder style layered coil, for the same amount of wire you can get increased induction and capacitance.  Was wondering if say we used round wire like we do if we were to immerse the coil in a dielectric liquid that didnt affect the enamel/insulation to fill in the air cavities along the wires if we could increase the capacitance.

Mags

Quote from: Itsu on 2017.05.03, 14:32:11
Spot on PM, back in business.....?  COP=1.2 @ 1MHz   


Paralleling the both used 51 Ohm metal film and metal oxyde resistors with a 1nF silver mica cap shows resonances at 18MHz, see SA screenshot.
This equates to 0.078uH inductance for both 51 Ohm resistors used.


When omitting the current probe, and more important.... to ignore the -80° phase shift..... i redid my measurements using the blue R2 (1 Ohm) csr trace as current value.
I started at 1MHz and this caused the phase shift to lower to -64.5°, and calculating the input (both by the scope math function as by hand) shows a 143mW input.
Using the CH3 purple probe value in the output calculation i now get 3.005² / 52 = 173.6mW so a COP of 1.2
See scope screenshot which shows:

yellow; voltage across the FG
blue; voltage (= current) across the 1 Ohm csr
purple; voltage across the 51 plus 1 Ohm resistors
red; math of yellow x blue is averaged power input


At 500KHz i see a COP = 1.18      phase shift yellow/blue -75°
At 1MHz    i see a COP = 1.2       phase shift yellow/blue -64°
At 1.5MHz i see a COP = 1.125     phase shift yellow/blue -51°
At 2MHz    i see a COP = 1.04      phase shift yellow/blue -45° 

The current probe deskew function (thanks PM) is to limited (±10ns) to correct the skew, so cannot be used to double check.
I know these COP values are only marginal, but its a start.
Are we sure that what we are using is the correct way to calculate the output thus the COP?

Regards Itsu

Quote from: verpies on 2017.05.03, 09:13:55
You've got to be kidding!

In scientific research, negative data is as much valuable as positive.

You've got plenty of things to try.  Such as:
- Using your TG + spectrum analyzer to sweep your resistors (and capacitors - if you use any).
- Putting a 1:1 ferrite isolation transformer on the output of your FG and measuring the input power this way (http://www.overunityresearch.com/index.php?topic=3445.msg61467#msg61467).
- Calibrating the skew and amplitude difference of you current probe vs. CSRs. at various frequencies.
- Shielding /putting the device in a metal box (not the CSR though).

Quote from: gotoluc on 2017.05.03, 15:46:42
@partzman

Here is the link to woopy's test and video demo: http://overunity.com/17119/pulling-energy-from-the-ambient-energy-field-using-a-coil-capacitor/msg504400/#msg504400

Quote from: TinMan on 2017.05.03, 13:19:15
Well,it's one of two things,when measuring a voltage drop across a pure resistance.
1-ohms law stand's,and the results are correct
2-ohms law fails in this situation.

So,am i correct to say that -when the phase angle between input voltage,and input current are at 90*,the circuit is purely reactive,and no real power is being consumed?
So,i can change the frequency to a point where the voltage and current are 90* out of phase,and yet still dissipate 53.8mW from the R2 resistor in my circuit.

Changing the frequency up or down,to sweep across either side of the 90* phase angle,shows no drop in dissipated power from R2.

I still think there is something to this current transfer between coil's,where it is both inductive and capacitive.

Is there a chance that the extra energy being seen,is in some way related to electron stripping,via dielectric stress.


Brad

Quote from: Vasik041 on 2017.05.03, 16:48:27
Hi Partzman,

Could you please confirm - how you feeding power into the coils ?
Do you take any special measures to match impedance or just connect to signal generator ?

I think this is important part of circuit (power source) left without much attention.

Thanks,
Vasik

Quote from: partzman on 2017.05.03, 18:01:15
Vasik,

The generator is set for a 50 ohm output impedance and is connected directly to the primary of the coil assembly under test.  No attempt is made at impedance matching at the input at this point in time.  Do you have any suggestions for this?  Perhaps this is not what you are intending!

Pm

Quote from: partzman on 2017.05.03, 15:49:09
Itsu,

That's great O0!  Now you might give some thought on confirming the inductance of the resistors you are using. 

Regarding the output calculation, using Pout = Vout^2/(R1+R2) is not exactly correct but close enough for our current observations.  As we increase the power levels we will see that using this current formula understates the actual power output.

One other thing you might consider that should raise your COP levels is to shorten your leads wherever possible.  I use a chassis mount BNC male connector on the output of the coax from the signal generator and R2 connects directly to the shield with just enough wire to clip the scope probe leads to.  The coil leads are also as short as physically possible.

Pm

Quote from: Vasik041 on 2017.05.03, 16:48:27
Hi Partzman,

Could you please confirm - how you feeding power into the coils ?
Do you take any special measures to match impedance or just connect to signal generator ?

I think this is important part of circuit (power source) left without much attention.

Thanks,
Vasik

PS I also don't understand why not use setup like this ? (attached)

Quote1kg 3.15mm x 1.12mm

Quote from: partzman on 2017.05.03, 19:08:56
Vasik,

I apologize as I previously missed this post until now.  I have already answered your input question so I'll respond to your question regarding the sense resistor location.  With the optional position you show, the signal level across the sense resistor is low enough to create some difficulty in making accurate differential measurements with a scope.  The only reason to use this location as compared to the grounded connection I show, would be to include any current drawn from the generator as a result of coupling from the DUT to any nearby grounded object, probe cable shield, etc, that would otherwise not be seen by the grounded sense resistor.

Tests have shown this to be negligible.

Edit:  See my post #138 for results of this testing.

Pm   

Quote
When using these correction factors in my original example, the corrected input voltage of 6.492v rms = 6.487 and the corrected input current of 30.53ma = 31.84ma for a corrected Pin = .207*cosine(-81.8 degrees) = 29.46mw.  The original uncorrected and calculated Pin = (6.492*.03053)*cosine(-81.8 degrees) = 28.27mw for comparison.  The corrected COP = 43.7mw/29.46mw = 1.48.

Quote from: Peterae on 2017.05.03, 19:26:23
The rectangular wire that Gyula suggested using Is it worth trying this wire, if so i can order a couple of reels.

How are you guys determining the turns required in your coils?

QuotePeterae,

IMO there is no need to go to any expense at this point with the wire although I'm sure it would work.  The results are achievable with magnet wire and ribbon cable whether wound as individual coils or as a Tesla bifilar pancake coil.

Using two wires from a ribbon cable wound on a form (other than round) with 25-50 turns should give you a workable device.  More turns increases both the inductance and distributed capacitance which means the operating frequency range will be lower and perhaps more manageable.

Pm

Quote from: Vasik041 on 2017.05.03, 19:39:29
Partzman,

Thank you for clarification.
from your post #138

Not corrected COP = 43.7 / 28.27 = 1.54, so difference is 6%, not so small.

Quote
It would be easier make measurements on lower frequencies, if you have ferrite rod from MW radio you can try placing it into middle of the coils (just an idea)

Quote
Regards,
Vasik

Quote from: Grumage on 2017.05.03, 14:33:08
Dear Gyula.

Many thanks for your reply.

I was suggesting that pure Copper cable as a quadfilar arrangement, as PM seems to have used what we call " bell wire " over here, you call it " hookup wire " ?

Slightly changing the subject what's needed to wind a coil that can resonate in the 10/100 of kHz range? I would like to join in but my gear is out of puff at 1 MHz.

Kind regards, Graham.

Quote from: gotoluc on 2017.05.03, 21:21:31
Hi guys,

I tested partzman's bifilar coil as per TK's circuit below.

I do get OU from 450kHz to 2.7MHz with all probes set to 1x. However, if I change the input voltage probe (only) to 10x the OU is mostly gone as demonstrated in video.
The higher the frequency (450kHz and above) the more the output. However, the input  goes up as well as I raise the frequency. Haven't took the time to find the ideal frequency.

Link to video demo: https://www.youtube.com/watch?v=B0V18JAafhc

The signal generator is a 20MHz BK Precision with output at max.
The BPC is 16 turns (each coil) of  0.9mm wire with 0.07 Ohms DC resistance per coil.
The Inductance is 16mH and the capacitance is 1500pf
The load and current sensing resistors are selected (tested) precision metal film resistors.

The video is kind of long and I had to cut off the end as the camera created a second file and it wasn't worth the time to edit them together. It should be good enough to get the idea of what it does.

Luc

Quote from: gotoluc on 2017.05.03, 21:21:31
Hi guys,

I tested partzman's bifilar coil as per TK's circuit below.

I do get OU from 450kHz to 2.7MHz with all probes set to 1x. However, if I change the input voltage probe (only) to 10x the OU is mostly gone as demonstrated in video.
The higher the frequency (450kHz and above) the more the output. However, the input  goes up as well as I raise the frequency. Haven't took the time to find the ideal frequency.

Link to video demo: https://www.youtube.com/watch?v=B0V18JAafhc

The signal generator is a 20MHz BK Precision with output at max.
The BPC is 16 turns (each coil) of  0.9mm wire with 0.07 Ohms DC resistance per coil.
The Inductance is 16mH and the capacitance is 1500pf
The load and current sensing resistors are selected (tested) precision metal film resistors.

The video is kind of long and I had to cut off the end as the camera created a second file and it wasn't worth the time to edit them together. It should be good enough to get the idea of what it does.

Luc

Quote from: partzman on 2017.05.03, 20:01:44
That is true and I agree.  We are assuming that this change is due to slight displacement and induction currents which may or may not be the case.  Prudent shielding would be in order to determine for sure.

I agree that lower frequencies would better but my experience has shown that as the frequency is lowered by purposeful coil design, the the COP<1.  When one adds adds a ferrite core in the center of the coil arrangement as you suggest, there is no change on the circuit measurements or parameters as one would expect.  These are clues to the device's operation and no, I do not have the answers  :-[!

Pm

Quote from: partzman on 2017.05.03, 19:30:42
All,

For anyone trying to replicate the MEI device, I would recommend you do a simple test when the device is in operation: Move the scope probe cables one at a time to see if your measurements are affected in any way to see if they are acting as antennas at your operating frequency.  You might want to reduce your sample averaging to 4 if it is higher so there is little delay in response.

I would also recommend a scope re-calibration after a suitable warmup just prior to taking measurements.  Perhaps even several times a day if you are on the bench that long.

Pm

Quote from: picowatt on 2017.05.03, 19:01:11
Greetings all...

Well, I must say this is all a bit intriguing to say the least.  If I can clear up my schedule a bit, i may have a go at a replication.

Might I suggest that those replicators that are using an FG with a known output impedance, such as 50R, measure the FG's open circuit voltage and compare that to the FG's output when connected to the circuit.  The difference between those two numbers divided by the 50R output Z will give you another Pin number to compare to the Pin calculated using R2 (or R2A).  Although it does not rule out the possibility of phase shift within the FG's 50R, it will give you a another Pin for comparison. 

TK, when measuring across R2A where you say you observed a phase shift, was The FG out, R2A, and the CH1 probe point in close proximity (i.e., limited wiring inductance)?

Also, if one were to use a 1/4 watt 1% (or .1%) low TC MF resistors at the output and another identical resistor in series with the FG out and the L1 input, at some of the power levels observed (particularly with TK's FG), the Trise of those small wattage resistors should make Pin/Pout comparisons a bit more obvious to the touch (or a thermister/Tsensor).

Only somewhat tongue in cheek, and with a bit of a grin, it should even be possible to do a "which resistor smokes first" test using 1/8 watt resistors (or even 0805's or smaller).  Just watch the TC (and inductance) of the resistors used. 

Just a few quick thoughts...

PW

Quote from: TinselKoala on 2017.05.03, 23:06:33
How in the world do you get such high inductance with only 16+16 turns in that small coil? I have far more turns in my coils and my inductances are in the 600-700 microHenry range. You say 16 milliHenry for a small coreless coil with 32 turns? Are you quite certain of that?

Quote from: TinselKoala on 2017.05.03, 23:06:33
Also, it would be very helpful if you could have your scope show the Phase Angle between CH1 and CH2, because that would allow manual calculation to check the scope's Math result.

Vrms (CH1) x Irms (CH2) x cos(Phase Angle 1-->2) = average input power

Quote from: TinselKoala on 2017.05.03, 23:06:33
The fact that you need to use 1x probes is a huge red flag, because your probes are becoming part of the circuit at that level of no attenuation. 

Quote from: TinselKoala on 2017.05.03, 23:06:33
Is your Function Generator's Black output lead isolated from ground? Many such FGs have their BNC shields connected to chassis ground and back through the mains line cord to the ground pin, and hence to all other instruments that are on the same mains circuit. This may cause complications that need to be taken into account when probing various parts of the circuit.

Quote from: TinselKoala on 2017.05.03, 23:06:33
When I changed from metalfilm resistors and long wiring like yours to non-inductive resistors and minimal wiring length it became harder to get "OU" results at low frequencies.

Quote from: TinselKoala on 2017.05.03, 23:27:07
Yes, the slight phase shift observed when trying to do the CH1-CH4 differential measurement across R2b was due to wiring inductance, and I was able to reduce it (but not eliminate it) with more careful probe and reference lead placement.  Let us please be sure we are talking about the same thing here.... the CH1 probe point is connected to R2b, not R2a, and the R2a phase shift CH1-->CH2 is an entirely different matter and does not require a differential measurement. But yes, the FG Black output lead and CH2 reference are connected very closely to one side of the R2a resistor and the CH2 probe is connected right at the other side of this resistor.  See the closeup of my connections posted earlier in the thread. (But note that I now am using non-inductive load resistors, 4ea. 4.7 ohm 1% in TO-220 packages for R1, connected in series with as short leads as possible.) And yes, the CH1 probe and CH4 probe are connected directly against the body of the R2b resistor when trying to do the differential measurement across this resistor (which strategy I have now abandoned as being irrelevant and hard to do properly.)

However all problems trying to read the Vdrop across R2b are eliminated simply by removing all probes and their reference clips from the circuit and measuring at R2a and R2b with the same probe and reference lead sequentially. (Obviously requires an isolated FG so as not to create a groundloop.) Doing this shows that the currents flowing in R2a and R2b are identical in amplitude (see scopeshot below comparing a stored reference Vdrop trace from R2a with a live trace from R2b.)

Quote from: picowatt on 2017.05.04, 02:26:22
TK,

Sorry for the R2a/R2b confusion.  I thought R2 remained just plain old R2 and that the added CVR at the FG out was referred to as R2a.  But yes, I was discussing R2b, the CVR at the FG out.

However, I am now confused as to why you are disconnecting all other probes when measuring the Vdrop across R2b.  What "problems" are eliminated by disconnecting the other probes?


PW

Quote from: TinselKoala on 2017.05.04, 07:44:34
To completely delete or "remove" a post.... click on the "remove" button next to the "quote" and "modify" buttons at lower right.

QuoteThere is no such button  :-\

Quote from: TinselKoala on 2017.05.04, 09:12:15
You only get to "remove" or "modify" your own posts, and you have to be logged in!

Quote from: poynt99 on 2017.05.04, 13:21:23
Some may find this document interesting and/or helpful regarding these investigations. Talks mainly about coil self-resonance and self-capacitance.

It contains tests/experimental setups, and is not overly technical.

Quote from: TinMan on 2017.05.04, 13:38:31
Ok,below is the new coil i just finished.

There are two layers on it,the first is on winding,and the second layer is the second winding.

This coil has a low resonant frequency for an air core coil,at just 71KHz exactly.

Inductance value is-- 2.54mH
Capacitance value is-- 4.8nF-->with the two coils open.
Resistance of each coil is 3.8 ohm's,and 7.5 in series.

So,here is the bomb shell.

With this coil,i can get an exact 90* phase shift between voltage and current--pure reactive circuit,and still get 450mV across the R2 resistor.
But my scope is playing funny buggers,and CH2 is very noisy-until i give the scope a good thump,then it clears up.
Im thinking a dry joint on the PCB inside the scope some where.

I can also make the coil either more inductive or capacitive(voltage leads or lags current),as well as a ratio of the two,so the coil is being seen as a pure resistance-voltage and current in phase.

This coil is much easier to tune in many way's,but dose not give the OU results that the BPC dose.
Maybe just the wrong geometry ?


Brad

Quote from: TinselKoala on 2017.05.04, 07:44:34
Since my oscilloscope does not have isolated channel references (like Russ Grier's TDS2024b does), it is necessary to disconnect all the probe reference leads in order not to create a groundloop when measuring across the two resistors separately.  So to measure the current through R2a, I connect the probe tip close to one side of the resistor and the probe reference to the other side. Then I move the probe and its reference to either side of the R2b resistor. If some other probe reference remains connected to the "ground" side of R2a, the common ground point, then this point is shorted to the side of the R2B resistor which has its probe reference connected. 

My FG's "black" output (bnc shield) can be isolated by a switch. It is pretty neat the way the F43 achieves this; its metal box is grounded to the mains cord ground pin, but the circuit board is isolated and all the BNC connectors (it has 5 on the front panel) are insulated bulkhead feedthrus. The circuit board is only grounded to the case at a single point and this point is connected through the switch (which is normally on the back panel but I have relocated it to the front panel for convenience.) Of course as soon as one connects a BNC patch cord from the F43 to another, grounded, instrument this isolation goes away, but in this case I have not done that.

My ElCheepo DDSFG is powered by a 2-pin 5 VDC wallwart power supply so its outputs are also isolated from the mains ground, unless patched to another grounded instrument.

You are of course correct in that this results in some capacitive coupling to true Earth ground, but this doesn't seem to be a problem with measurements of this circuit. The important feature of the "isolation" is that the Black (bnc shield) FG lead is not directly connected to the mains cord ground pin and thus to the chassis ground and probe ground references of the oscilloscope or other mains-powered and grounded instruments.


To completely delete or "remove" a post.... click on the "remove" button next to the "quote" and "modify" buttons at lower right.

Quote from: Smudge on 2017.05.04, 14:13:06
It is known that that electric forces and magnetic forces between moving charges are two manifestations of the same effect, magnetic forces being v/c times the electric ones.  It is the historical evolution of electromagnetic knowledge that has labelled them as separate forces.  So there is some characteristic of space that creates or carries this velocity dependent force.

Quote
Now imagine you are an observer within your laboratory reference frame where you observe a ball coming towards you at a certain velocity.  The ball is spinning and it has some observable characteristic that tells you the spin axis.  You can discern the angle between the arriving velocity direction and the spin axis.  Now imagine that you are moving relative to the reference frame.  That particular ball will not now appear as travelling towards you it would pass you by.  However another identical ball travelling at the same velocity and on a path parallel to it and having the same spin orientation would appear as travelling towards you.  But now the angle between the spin axis and the arrival direction would be seen to be different.  I believe that this simple velocity dependent spin angle is the basis for the link between electric and magnetic forces.
 
My space is inhabited by particles that resemble neutrinos whizzing through it at light velocity coming from all directions and at a huge number-density.  They have momentum so on impact with a matter particle there is a force impulse.  They carry spin that is an independent vector which can have any orientation with respect to the velocity vector.  Absorption of these space particles by matter particles and subsequent emission obey rules that use the velocity dependent spin angle to determine the resultant force endured by the matter particle.  It is the pattern of those spin angles that is imprinted onto the vast collection of arriving space particles that are the photons that drive our conduction electrons.  Or rather the pattern as seen by each moving electron.  We have the ability to create patterns that produce measurable effects.  Note that there are two patterns involved, that which drives the output electrons and that which drives the input electrons, and normally the measured effects yield COP<1.  We have this two way communication between output and input, but the forces involved are always due to interaction between the electrons and the space particles.

Now in the experiments being performed here the electrons in the output are being driven by a combination of both magnetic and electric forces.  That combination is being produced in an unusual manner.  IOW the pattern imprinted onto the space particles that will arrive at the output electrons is unusual.  And the pattern imprinted onto the space particles that will arrive at  the input electrons is unusual.  I see no reason why those patterns should not produce COP>1 or even COP=infinity.

Quote
Smudge

Quote from: TinselKoala on 2017.05.04, 11:03:09
Another pair of 30 second sweeps, this time logarithmic from 50 kHz to 25 MHz . The big spike in CH4 is at about 1.7 MHz.

Quote from: Itsu on 2017.05.04, 09:48:44
It shows how pure (or not) the input signal is.

Quote from: Smudge on 2017.05.04, 14:13:06
It is known that that electric forces and magnetic forces between moving charges are two manifestations of the same effect, magnetic forces being v/c times the electric ones.   

Quote from: verpies on 2017.05.04, 15:08:44
Notice, that the cos(phi) power factor is applicable only to the 1MHz peak.
All others are unaccounted for.

Quote from: Itsu on 2017.05.04, 15:23:53
Does that mean that these unaccounted peaks are NOT properly being reflected into the input power?  But still are there causing this COP>1 output?

Quote from: verpies on 2017.05.04, 15:22:31
...and gravitational forces being v/c times the magnetic ones.

Aren't such musings better suitable for this thread (http://www.overunityresearch.com/index.php?topic=3447.msg61320#msg61320)?

Quote from: partzman on 2017.05.04, 14:34:45
Smudge,

Thanks for your input!  So I can grasp what you are saying and meaning above in the highlighted area, could you elaborate on this please.  For example, is the "v" above the velocity of the electromagnetic longitudinal wave in the transmission  line equivalent or......?   

Quote from: TinselKoala on 2017.05.03, 12:49:24
Well, we have apparently eliminated the objection about the location of the CVR at least. However I think we still don't know if the whole probing scheme is the proper way to test this circuit.  Does the Vrms²/R_total calculation accurately give a real output value? Do the resistors heat up enough?
Mine do feel slightly warm but remember I'm using the F43 at over 20 v p-p input.

It's a great puzzler all right. But so far in my testing with the Gyula Doubler on the output, the "ou" effect goes away when that is connected.

Quote from: Itsu on 2017.05.03, 21:11:01
Leads shortend where possible yields a slightly higher COP @ 1MHz (1.22 instead of 1.2).

Linking my both identical TBP coils together (both equiped with a 51 Ohm load resistor and 1 Ohm csr) drops the output of the first to half while input stays the same.
COP now way below 1 (@ 1MHz)

Input of 2th TBP coil was from top of 51 Ohm load resistor of 1st TBP coil (CH3 purple probe point), ground of 2th TBP coil was via its 1 Ohm csr to the common ground point.

I also tried Picowatts suggestion for an alternative Pin method, but that did not match up (using Vrms values). FG open circuit value 7.127V, when connected to circuit 5.082V.
Difference is  2.045V, divided by 50 gives 40.9mW while the measured input power was 147mW


Itsu

Quote from: picowatt on 2017.05.04, 14:27:07
TK,

The schematic you provided showed four probes connected to the circuit, with two of those probes across R2b.  I thought you were just measuring the difference between those two probes for the R2b Vdrop.  Did you confirm that removing the other probes from the circuit had no effect on the R2b Vdrop or phase angle?  (Luc's measurements changed just from switching a probe's attenuation factor)

Quote

Was R2b positioned close to the FG's output or the coil?  What does the Vdrop across R2b look like with R2b positioned right at the FG output

Quote

I would be very cautious about connecting a shield to various measurement points and just assuming it is having no effect elsewhere on this circuit.  Floated grounds should only be considered as isolating the signal and chassis/AC ground via a capacitance in parallel with a large resistance.  That capacitance and resistance can be measured.

I could not find a remove or delete button, apparently newbies don't have full driving privileges.

No worries, some kind soul cleaned up my mess...

PW

Quote from: partzman on 2017.05.04, 14:53:51
TK,

This is very interesting!  Is this the same coil you started with that you previously posted the specs for?  Also having the dimensions we could calculate the MLT to determine the number of turns if you don't have that info.  I'm curious as to what your interpretation is of this peaked induction?

Pm

Quote from: gyula on 2017.05.04, 16:26:08
Hi TinselKoala,

Would you mind to share some details on this?
What I am mainly curious about is
-phase shift at input goes away and retuning is needed to get the -80 to -82 degree phase shift when the resistor R1 removed from L2
-any other detail you have done on this test when you have a few minutes

Thanks
Gyula

Quote from: verpies on 2017.05.04, 15:39:47
I do not have data on reflections in this circuit, but the cos(phi) power factor can be used for the phase correction of the V_RMS * I_RMS product only for purely sinusoidal waveforms.  Any peak besides the first one indicates a departure from a pure sine wave, and the energy content of these peaks is not properly accounted for by the single cos(phi) power factor.

P.S.
^{I'm posting from a mobile phone now and I don't remember whether your spectrogram represented input current or input voltage.}

Quote from: TinselKoala on 2017.05.04, 19:03:50
I don't have a Spectrum Analyser but the Rigol z-box can do a crude FFT display so might be able to say something about the spectral purity of the signal input I'm using. More work ahead!

Quote from: TinselKoala on 2017.05.04, 19:03:50
Do we have some kind of idea just how far off the basic formula will be, for some degree of distortion in the sine wave?

Quote from: partzman on 2017.05.04, 21:12:47
...it would be wise to isolate the signal generator ground from the test set as ION suggested. 

Quote from: verpies on 2017.05.03, 09:13:55
- Putting a 1:1 ferrite isolation transformer on the output of your FG...

Quote from: verpies on 2017.05.04, 23:22:51
It was also my suggestion and I think it predated ION's

Quote from: TinselKoala on 2017.05.04, 19:03:50
Do we have some kind of idea just how far off the basic formula will be, for some degree of distortion in the sine wave?

Quote
Do you think something like this might account for the fact that my scope's automatic measurement of average power is consistently about 20 percent higher than the manual calculation indicates?

Quote from: Itsu on 2017.05.05, 08:41:28
I started to use a 1:1 ferrite isolation transformer yesterday, (see picture) and it lowered the COP @ 1Mhz from 1.2 to 1.03.
No more COP > 1 on other frequencies.

A spectrogram with or without the 1:1 ferrite isolation transformer however, does not show any significant difference.

I will be waiting for my caddock non inductive resistors to arrive for doing further measurements, but its not looking good.


Itsu

Quote from: poynt99 on 2017.05.05, 03:45:46
This should be relatively easy to test, and I would suggest using "kinder" frequencies such as the audio range where your scope is sure to perform well. One way would be to put a 1Vp sine signal into the attached circuit with the diode(s) removed and measure the baseline distortion of your generator/op-amp. Then insert a single diode for even-harmonic (2nd) distortion, or add a second diode in reverse parallel with the one shown for odd-harmonic distortion, and vary the amount of distortion by varying the pot. You may have to go up to a 500k or 1M pot to get more subtle amounts of limiting and distortion. Perhaps you can now compare the two measurement methods by driving a small load with the op-amp.
That is troubling.

Quote from: TinselKoala on 2017.05.05, 10:32:35
I've spent all night making a detailed video demonstrating the setup and the phenomenon and another phenomenon we haven't yet discussed here. The video is about 15 minutes long. It will take some time (hours) to render and upload.

I intend to post the link here privately at first (via PM to the experimenters and other interested members) , unless Partzman gives permission to post it publicly. I'm sure it will cause a storm of controversy.

By the way, that reposting of the Transmission Line video to Hackaday has caused an incredible inflation of the view count over the past couple of days. It went from about 500 views yesterday morning, which is about what I usually get for these things, to nearly 4,500 views by yesterday evening. It seem to have topped out for now at about 4800 views. And not even one single "thumbs down" ! That is extremely unusual, my trolls must be asleep.

Front title slide and schematic for the upcoming video:

Quote from: TinMan on 2017.05.05, 09:12:24
Itsu

Are you measuring the P/in after the 1:1 transformer,or at the input of the 1:1 transformer.
In other words,is the power the transformer is consuming,being included in the P/in measurements ?

BTW--it is great to see you working on this  O0


Brad

Quote from: TinselKoala on 2017.05.05, 10:32:35
Front title slide and schematic for the upcoming video:

Quote from: Itsu on 2017.05.05, 11:00:38
Hi Brad,

its after the 1:1 xformer, like depicted by TK's diagram here above.

.....


Itsu

QuoteThe caddocks are in, so time to fire up the solder iron

Quote from: TinselKoala on 2017.05.05, 12:12:09
Don't fool yourself... current is the flow of charge, from higher potential to lower potential. I know some people who "go both ways" at once,  but current doesn't.

Quote from: TinselKoala on 2017.05.05, 12:12:09
Don't fool yourself... current is the flow of charge, from higher potential to lower potential. I know some people who "go both ways" at once,  but current doesn't.

Quote from: verpies on 2017.05.04, 23:22:51
It was also my suggestion and I think it predated ION's

Quote from: TinMan on 2017.05.05, 11:58:34
O0

Well,it seems that the old BPC out-performs the new solenoid type i made.

Using Partzmans circuit,i can get constant COP>172-176%

Seems the P/in measurements should be correct,as the voltage and current are in phase,meaning a power factor of 1. RMS value's are now used to calculate P/in,and P/out is the voltage across the resistors-ohms law.

So,it would seem that at least 3 here are seeing constant COP> result's--but dose anyone know why yet?,or where the red herring is hiding?

It has been great research at the very least--who'd of ever thought that current can actually travel in both directions,at the same time,through a coil  :o

Seem to remember some one once saying that,and most of us laughed at him.
Maybe time !for me anyway! to go and eat some humble pie.


Brad

Quote from: TinMan on 2017.05.05, 12:21:29
Not sure if there was any type of message in the !!I know some people who "go both ways" at once!! ,but are you saying the phase inversion we see across R1 and R2 is not the current starting to flow in the opposite direction?

As far as i know,when two current flows are 180* out of phase with each other,then they are flowing in opposite directions--of have i missed something some where?


Brad

Quote from: partzman on 2017.05.05, 12:54:28
It becomes easier to understand and accept current flow into or out of a given coil simultaneously if consideration is given to the transmission line nature of these special circuits.  With induction being created by displacement currents, it is possible under the right configurations to view this phenomenon both in simulation and on the bench.  In my experience the nexus appears to be in the center of the winding.  It is most enlightening in a simulation with the transmission line modeled as lumped L/C segments to view this phenomenon by measuring the currents and voltages up and down the line at any given point in time.

Pm

Quote from: partzman on 2017.05.05, 12:30:17
Verpies,

Yes I apologize for not recognizing you for that suggestion.  In my early testing of the MEI I questioned the measurements and did apply isolation transformers as well as an isolated TDS 3034 scope use via battery operation.  I still have the test data somewhere in my files if anyone is interested.

Pm

QuoteI'm sure partzman (or someone) has probably tried this and I missed it. If of no value, disregard.

Quote from: TinselKoala on 2017.05.05, 13:08:07
So when you connect a bit of test gear to such a place where "current is flowing in both directions simultaneously"... how does the test kit know which direction to measure?
Sorry, but in no model of electrical reality that I can consider does current flow both ways simultaneously.  I also don't understand how such a measurement could even be made properly.

Quote from: TinMan on 2017.05.05, 13:20:47
Perhaps a better explanation.

What i should have said was- at the junction of the end of one coil,and the beginning of the second coil(where R2 is on my schematic),it is possible to see an inversion of 180* to that of R1 at the input.

If both phases are reading the voltage drop across R1 and R2,and these values become inverted across 1 of the resistors,but not the other,dose that not mean that the current is now flowing in one direction at a point in time,and in the other direction through the other resistor,at that same point in time?.


Brad

Quote from: ION on 2017.05.05, 13:50:33
TK

Try moving your CSR to the (normally grounded) output leg of the ferrite, and repeat the readings per the attached schematic or Vasik's recent schematic.

If anyone else already suggested this, be it known they did.

Regards

Quote from: TinselKoala on 2017.05.05, 14:04:40
I have no problem with that. Two different measurement points separated by an active circuit node or reactive components .... fine. Especially when capacitors and inductors are involved.

Where I draw the line is at the claim that current can flow through the _same point_ in both directions simultaneously. How would you even measure that?

Quote from: TinselKoala on 2017.05.05, 14:04:40
I have no problem with that. Two different measurement points separated by an active circuit node or reactive components .... fine. Especially when capacitors and inductors are involved.

Where I draw the line is at the claim that current can flow through the _same point_ in both directions simultaneously. How would you even measure that?

Quote from: TinselKoala on 2017.05.05, 14:19:08
I would also like to know what Partzman found when he tried it that way (as I'm sure he must have done.)

Quote from: TinselKoala on 2017.05.05, 10:32:35
I've spent all night making a detailed video demonstrating the setup and the phenomenon and another phenomenon we haven't yet discussed here. The video is about 15 minutes long. It will take some time (hours) to render and upload.

I intend to post the link here privately at first (via PM to the experimenters and other interested members) , unless Partzman gives permission to post it publicly. I'm sure it will cause a storm of controversy.

By the way, that reposting of the Transmission Line video to Hackaday has caused an incredible inflation of the view count over the past couple of days. It went from about 500 views yesterday morning, which is about what I usually get for these things, to nearly 4,500 views by yesterday evening. It seem to have topped out for now at about 4800 views. And not even one single "thumbs down" ! That is extremely unusual, my trolls must be asleep.

Front title slide and schematic for the upcoming video:

Quote from: TinMan on 2017.05.05, 14:55:49
What can we do,or how would we go about bumping up the power into the watts range.

Is there an amplifier we can use with our SGs,that could achieve this ?.


Brad

Quote from: TinMan on 2017.05.05, 14:55:49
What can we do,or how would we go about bumping up the power into the watts range.

Is there an amplifier we can use with our SGs,that could achieve this ?.
Brad

Quote from: partzman on 2017.05.05, 15:23:51
TK,

Kudos to you for all the long hours you have spent on this plus your demonstration of an external coil with load and the effect it has on the inducing MEI source coil.  I do appreciate your consideration of keeping the video private out of respect for my MEI efforts over the years.  For those who may not like or agree with this and the attendant "behind the scenes" discussions regarding such a matter, place yourself in the position of holding potential intellectual property that could be game changing.  Would you in all honesty be willing to "give it away"?

Here is the link to TK's video-

https://www.youtube.com/watch?v=MNzbc-N-e9c&feature=youtu.be

I agree with you TK, this will create controversy but will open doors for more research.

Pm

Quote from: TinMan on 2017.05.05, 14:55:49
What can we do,or how would we go about bumping up the power into the watts range.

Is there an amplifier we can use with our SGs,that could achieve this ?.


Brad

Quote from: TinselKoala on 2017.05.05, 16:15:26
Well, now anyone who finds this forum can see this link, right?

So is it OK with you if I list it publicly on my channel, so my subscribers and my trolls will see it? Or would you rather I didn't... it's up to you.    ;)

Quote from: partzman on 2017.05.05, 16:22:54
Yes, and it's OK with me to make it public on your channel.

Pm

Quote from: TinselKoala on 2017.05.05, 16:15:26
Well, now anyone who finds this forum can see this link, right?

So is it OK with you if I list it publicly on my channel, so my subscribers and my trolls will see it? Or would you rather I didn't... it's up to you.    ;)

Quote from: ION on 2017.05.05, 15:35:35
The power amplification (if it is real and being measured correctly) is inherent in the device. All you need to do is loop it correctly and a tiny COP improvement over COP=1 (enough to cover losses in looping components) will yield enormous and dangerous results via regeneration which in this case is continuous power multiplication until the device melts down and stops or otherwise self destructs.

The problem then becomes controlling the reaction, "the fire that feeds itself"

I will post some simple limiter circuits to eliminate the "dangerous" part.

Regards

P.S. read up on SM's stuff, he covers all this. And BTW, I believe someone? said that in a transmission line a current pulse can travel in both directions e.g. the forward pulse passes through the reflected pulse of an open or shorted line. There are simulators on line demonstrating something like this.

Quote from: TinselKoala on 2017.05.05, 10:32:35
I've spent all night making a detailed video demonstrating the setup and the phenomenon and another phenomenon we haven't yet discussed here.

Quote from: partzman on 2017.05.05, 16:38:37
ION,

I agree with you here.  Perhaps more power is not needed for successful looping, but maybe it would be useful to have higher power levels to help solidify accurate measurements for easier confirmation the device has a COP>1. 

What is really interesting to me at this point is TK's preliminary proof of loaded external induction not affecting the input power in a negative way abut rather positively.  IOW, does this mean that induction by displacement current is basically one-way, no lenz effect?  I've wondered this for a long time and TK has shown an easy way to test this.  This may also be a path for successful looping.

Is this all beginning to somewhat resemble the FTPU to you?

The regenerative issue is a big thing and certainly deserves discussion.  I hope we reach these kinds of problems :)!

Pm

Quote from: verpies on 2017.05.05, 17:01:04
The calculations are fine in your video.
The multiplication of two ADC quantization errors might introduce additional inaccuracies, especially if the ADCs are severely undersaturated.

Most of all, I still have doubts about the entire input current being "caught" by the R2a, especially the current due to the parasitic capacitance, that I've depicted in red below.  It would be more convincing if your measurement setup was more like on the attached schematic.

Quote

Also, at the end where you loaded the bifilar transformer with a spaced monofilar pancake coil and an LED, I was expecting the calculation of the output power (at least out of the BPT) ....alas it never came :(

Quote
P.S.
^{IIWY I would wind the isolation transformer to minimize the inter-winding capacitance (i.e. two separate windings, non-interlaced like in a CM Choke.).  I realize that the coupling coefficient is lower with that type of winding, but the inter-winding capacitance is much smaller and that is important in an isolation transformer.}

Quote from: ION on 2017.05.05, 17:05:37
Agreed on all points, especially scaling this up for higher output should be part of the discussion.

Just driving it harder may just give the same COP, but at higher power levels. Part of the investigation needs to be alternate topologies that would increase the COP. Nevertheless the results look very good and are adequate for looping.

My reservations are with measurements, but outside of the two schematics presented way back, I can't come up with anything that would help at this point. (disclaimer: if anyone also did, I missed it)

The displacement current Lenzeless idea is interesting.

An LED by itself half wave loads the coil which can be good or bad, TBD.

I would like to see TK replace the LED with a "grain of wheat" lamp which he has eyeball pre-calibrated for approximate known brightness vs mW. Then down the road he could use the temp or brightness measuring methods we all know and enjoy.

Regards

Quote from: verpies on 2017.05.05, 17:01:04
The calculations are fine in your video.
The multiplication of two ADC quantization errors might introduce additional inaccuracies, especially if the ADCs are severely undersaturated.

Most of all, I still have doubts about the entire input current being "caught" by the R2a, especially the current due to the parasitic capacitance, that I've depicted in red below.  It would be more convincing if your measurement setup was more like on the attached schematic.

Also, at the end where you loaded the bifilar transformer with a spaced monofilar pancake coil and an LED, I was expecting the calculation of the output power (at least out of the BPT) ....alas it never came :(

P.S.
^{IIWY I would wind the isolation transformer to minimize the inter-winding capacitance (i.e. two separate windings, non-interlaced like in a CM Choke.).  I realize that the coupling coefficient is lower with that type of winding, but the inter-winding capacitance is much smaller and that is important in an isolation transformer.}

Quote from: ION on 2017.05.05, 13:14:48
P.S. Smudge, Verpies: what do you think of the concept of morphogenesis and the 100th monkey effect. I have some ideas, but that should be on another thread.
How does your knowledge of physics explain it.

Quote from: Smudge on 2017.05.05, 18:03:50
The number of monkeys being studied was 59, so the 100th monkey effect didn't exist, and the spread of information through the colony was quite normal.  So it's a man made legend.  Don't need physics to explain this.

Smudge

QuoteMy order of "gow" bulbs just arrived in today's mail. They are 12 volt bulbs though, not like the really low voltage ones I remember from childhood. And I don't need to "eyeball" calibrate, I have a good lightmeter so I can actually do a lux vs. DC power kind of graph.

Quote from: ION on 2017.05.05, 18:34:57
So to clarify Drawing "*annot3newCVRvideo1" was used in the video and "*annot4_1" is currently being tested?

QuoteProbably i need to redo my 1:1 toroid as the present one has interwoven windings which need to be seperated each on its own half as i understand.

Quote from: Grumage on 2017.05.05, 15:30:28


Hi Brad.

The coil L1 is open ended, I assume that it's capacitively connected to the other side.

I'm a novice here but do we know what the actual capacitance is? How would a variable air gap trimmer in the pF range look?

Isn't this entering the RF category at 1 MHz frequency?

Cheers Graham.

Quote from: Itsu on 2017.05.05, 21:13:10
Hmmm,  i have received my Caddock non inductive 1% resistors (50 and 1 Ohm) and installed them, first making a bnc plug connection via a short (30cm) coax cable directly to my FG (no
toroid).

The COP > 1 dissappeared (well, @ 1Mhz i am left with cop = 1.03, rest of frequencies all below 1).


Then i reinserted the 1:1 toroid, isolating the FG ground and using TK's latest scheme (common ground point between R1 and R2, R2 signal reversed)

The COP went up right away reaching a COP = 4  @ 1Mhz.
When tightly inductively coupling (so back to back) a 2th TBP coil (in series connected with a LED) i get a COP = 10 when at 400KHz (2th TBP coil resonance) and with the led full on.


This cannot be right, so i will run these tests again tomorrow.

Probably i need to redo my 1:1 toroid as the present one has interwoven windings which need to be seperated each on its own half as i understand.

 
Itsu

Quote from: poynt99 on 2017.05.06, 00:26:30
Pardon my intrusions, but TK I did have a few questions regarding your video if you wouldn't mind indulging me:

Quote
1) Is there still L1/L2 coupling action (is there still some output power) if R2a is made infinite, i.e. an open?

Quote
2) What are the power calculations with 15-20 cycles on the screen (and appropriate memory depth/record length) and using "MEAN" and "RMS" rather than "Period" measurements?

Quote
3) What is the capacitance between L1 and L2?

Quote from: TinselKoala on 2017.05.06, 01:31:41
Just so I  know exactly what you mean, can you please indicate this on a schematic?

Quote
The phase angle measurement is less accurate with that many cycles on screen, so here are 3 scopeshots.

QuoteFirst, with a small number of cycles so you can see the phase angle more accurately, then the large number of cycles, and the complete data dump for that last one. You can "do the math" yourself....    ;)

Quote
(Rigol uses the word "average" instead of "mean" but they are the same thing numerically.)

Quote
I assume you mean when everything is disconnected and both windings are left floating at the non-dot ends, with capacitance meter connected to the dot ends.
The capacitance measured this way is 2.19 nF.

Quote from: poynt99 on 2017.05.06, 02:30:04
(snip)
What happened to the 4.6mW? It is now indicating 22mW. Is the Pout still 12mW?

Quote
(snip)
You might wish to try replacing the PBF with a 2.2nF cap and repeating the tests.

Quote from: poynt99 on 2017.05.06, 02:40:11
TK,

You may also wish to try this configuration...and repeat the tests (22m should give about 2.2nF).

Quote from: TinselKoala on 2017.05.06, 06:08:53
Another:

Quote from: TinselKoala on 2017.05.06, 04:59:40
So I think what the bulb test is telling me is that there is something seriously wrong somewhere.

Quote from: TinMan on 2017.05.06, 06:56:47
Would you not just measure the RMS value across the globe,when being powered by the BPC circuit,and then drop the equivalent DC voltage across the glode,and see what the DC power calculation comes out at?.


Brad

Quote from: TinselKoala on 2017.05.06, 08:02:09
Now you are getting ahead of the game plan!   ^-^

Right now I'm checking the sanity of the scope. I'm feeding a DC voltage into one channel to act as a reference. This voltage is monitored by my Fluke 87-III. Then I set the FG to make a sine wave that comes right up to the line of the DC voltage. Then I check the scope's automatic measurements against the Fluke's voltage reading, the frequency counter against the FG's setting, the cursor's math, and whatever else I can find in the full data dump. Checked both high range (5.04 volts) and low range (512 mV). Not surprisingly it is more accurate at the high end but even at the low end in the 200 mV/div range anyway it is still acceptably accurate, or at least it agrees with the Fluke to within one or two percent. (The Fluke still has a NIST-traceable calibration sticker on it, too.... )

So I think that my scope can probably be trusted, at least for p-p voltages and frequency.

Quote from: verpies on 2017.05.06, 07:57:57
Different load - different numbers.
What's the problem you see with this ?  Filament inductance, skin effect, GOW bulbs' parameter variations... ?

P.S.
^{What is the RMS value of the HF voltage across the GOW, as calculated by your scope vs. a DC voltage across the same GOW shining with the same brightness verified by your luxmeter ?}

Quote from: TinMan on 2017.05.06, 08:15:23
TK

Have you been getting any sleep?

I know this is exciting and all,but a TK collapsed in a big heap on the floor from fatigue overdose ,is no good to anyone.

I only ask,as my day time over here,is your night over there,and i see you here every time i log in.

It's now 4.15pm her,and so must be close to 4.15-5.15 am over there.

Look after yourself  O0


Brad