PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2021-01-27, 23:25:37
News: Registration with the OUR forum is now by invitation only.

Pages: 1 ... 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 [30] 31
Author Topic: Itsu's workbench / placeholder.  (Read 32075 times)

Group: Moderator
Hero Member
*****

Posts: 2501

Quote
Hi Itsu,

Very good and informative tests, thanks for performing them and presenting the results.

All this means that a "very carefully selected 0.1 Ohm, 1% Tolerance, Metal Strip Through Hole Resistor" cannot give correct measurement results from the some ten kHz or in the some hundred kHz frequency range and certainly not at the 800-900 kHz frequencies.
 
Flat wires, metal strips or any piece of wire that are about 3 cm long or longer with their connecting wires and mounted on a PCB board are inherently inductive and their inductive reactance gradually adds to the initial 0.1 Ohm DC resistance as the frequency increases.

This is valid for the Measuring Board Captainloz (or anyone else) used and got a COP > 1 result. All these results should be revisited by using a different 0.1 Ohm shunt resistor.

Gyula


Thanks Gyula,

i agree and i hope the Capatian will step up the plate and join me in finding out what that means for his COP = 2

Itsu
   

Group: Moderator
Hero Member
*****

Posts: 2501
bump
   
Newbie
*

Posts: 23
   Guys:

   It's sad, that it's taken this long to prove what is visually obvious, to me.
   
   So, where do we go from here???  If our input to output tests are proving nothing...
   
   Self running, becomes the only way to go. Or, should we still want to test for and have faith in, input to output readings???

   In any case, thanks itsu for your time and dedication to this effort. It means a lot, to me.
 
    NickZ
   

Group: Moderator
Hero Member
*****

Posts: 2501

Hi Nick,

"where do we go from here",  well, it seems Captainloz is back to asymmetrical transformer circuit testing which hopefully means he continues to work on his COP = 2 asymmetrical regauging experiment with me as promised.

Chris has released the pcb gerber files for the used csr pcb, so i ordered some as to have the same 0.1 Ohm csr measuringboard the captain uses.

If it really turns out he still has the COP = 2, then indeed the next step would be to try to loop it as a final test.

Regards itsu
   

Group: Moderator
Hero Member
*****

Posts: 2501
Like Chris, i am getting tired of explaining, in this case  what reactance does with a csr.

I have no problems with the "measurement protocol" used overthere itself, they seem sound.

Chris is right about using a good csr in a DC environment.
There, a 0.1 Ohm (100 mOhm) 1% csr is very accurate in measuring DC currents.

But he conveniently does not mention what it does when dealing with AC signals.
The captainLoz device runs at 830KHz, so surely no DC.

Lets see what this means for a high quality Johanson 100 mOhm 1% 5W csr like the RMCJ3U000R1FS from Mouser recommended by Chris:
https://nl.mouser.com/ProductDetail/Johanson-Dielectrics/RMCJ3U000R1FS/?qs=ofF%252BRbqKDcYsoVgaSsLAJA%3D%3D

The specs ( https://nl.mouser.com/datasheet/2/611/rmc-series-1074317.pdf ) say:

• Resistances from 0.005 to 0.100 Ohms
Low Inductance (<10nH)
• Tolerances to ± 1%
• Resistance Wire TCR: ± 20ppm/ºC
• For Current Sensing and Shunt Applications
• All Welded Construction
• Economical Bare Metal Element


So inductance = <10nH, lets say 9nH (it won't be much less)

According to this inductive reactance calculator:
https://www.66pacific.com/calculators/inductive-reactance-calculator.aspx
9nH @ 830KHz = 0.05 Ohm = 50 mOhm

So at 830KHz (working frequency of CaptainLoz his device), the ADDED reactance of this csr is 50 mOhm, so the total resistance (Impedance) is 150 mOhm, which is an 50% increase.

If i would use a 1 Ohm (1000 mOhm) 1% csr, this same 50 mOhm reactance would only increase the total impedance by 5%.

But... even more important then this 50% increase in impedance of the 0.1 Ohm csr @ 830KHz is how it is measured.

When using a pcb with connections and a voltage probe like the Captain does, the reactance (and resistance) at 830KHz causes an even worse situation (900% worse) like shown earlier in this thread where i used my VNA to characterize that situation:
https://www.overunityresearch.com/index.php?topic=3691.msg85851#msg85851

Waiting for the csr pcb's the make additional measurements with my VNA.

Itsu
« Last Edit: 2020-12-17, 12:46:08 by Itsu »
   
Group: Experimentalist
Hero Member
*****

Posts: 1221
Like Chris, i am getting tired of explaining, in this case  what reactance does with a csr.

I have no problems with the "measurement protocol" used overthere itself, they seem sound.

Chris is right about using a good csr in a DC environment.
There, a 0.1 Ohm (100 mOhm) 1% csr is very accurate in measuring DC currents.

But he conveniently does not mention what it does when dealing with AC signals.
The captainLoz device runs at 830KHz, so surely no DC.

Lets see what this means for a high quality Johanson 100 mOhm 1% 5W csr like the RMCJ3U000R1FS from Mouser recommended by Chris:
https://nl.mouser.com/ProductDetail/Johanson-Dielectrics/RMCJ3U000R1FS/?qs=ofF%252BRbqKDcYsoVgaSsLAJA%3D%3D

The specs ( https://nl.mouser.com/datasheet/2/611/rmc-series-1074317.pdf ) say:

• Resistances from 0.005 to 0.100 Ohms
Low Inductance (<10nH)
• Tolerances to ± 1%
• Resistance Wire TCR: ± 20ppm/ºC
• For Current Sensing and Shunt Applications
• All Welded Construction
• Economical Bare Metal Element


So inductance = <10nH, lets say 9nH (it won't be much less)

According to this inductive reactance calculator:
https://www.66pacific.com/calculators/inductive-reactance-calculator.aspx
9nH @ 830KHz = 0.05 Ohm = 50 mOhm

So at 830KHz (working frequency of CaptainLoz his device), the ADDED reactance of this csr is 50 mOhm, so the total resistance (Impedance) is 150 mOhm, which is an 50% increase.

If i would use a 1 Ohm (1000 mOhm) 1% csr, this same 50 mOhm reactance would only increase the total impedance by 5%.

But... even more important then this 50% increase in impedance of the 0.1 Ohm csr @ 830KHz is how it is measured.

When using a pcb with connections and a voltage probe like the Captain does, the reactance (and resistance) at 830KHz causes an even worse situation (900% worse) like shown earlier in this thread where i used my VNA to characterize that situation:
https://www.overunityresearch.com/index.php?topic=3691.msg85851#msg85851

Waiting for the csr pcb's the make additional measurements with my VNA.

Itsu

Itsu,

Good work and keep it up!

I have never understood why Chris requires the sensing resistor to be .1 ohm with all of it's attendant problems as you have attempted to point out? 

In general terms, it is better to have as low a sense resistor value as possible for accuracy in certain circuits but in his POC circuit, the power loss in a 1 ohm sense resistor in the input and/or output would simply be added to the output or subtracted from the input which wouldn't affect the COP. 

In fact, a non-inductive high quality load resistor equal to the operating resistance of the bulb could be used to eliminate the output sense resistor altogether and provide an extremely accurate output measurement.

Regards,
Pm
   
Group: Tech Wizard
Hero Member
*****

Posts: 661
Hi Itsu and Partzman,

I agree with both of you.

Itsu, your measurement results are rock-solid and clearly show the inductive behaviour of such metal strip resistors http://www.farnell.com/datasheets/1802151.pdf and Captainloz used such (either  the type this link shows or the one you just referred to) in his video 9 tests. 

Even a 1 cm long piece of wire with an OD of 0.5mm has about 10 nH inductance (empirical rule of thumb in microwave engineering) and both the 0.1 Ohm CSRs (PWR4412-2S Series Bare Metal Element Resistor) from Bourns and the RMCJ3U000R1FS type from Johanson are nothing else but a rectangularly bent piece of wire, their total length being in the range from 20 mm to 49mm (decimals are neglected) as their power rating varies between 1 W to 5 W.

So do not worry whatever is said to the contrary or even against you.

Hopefully Captainloz will return and repeat his tests with a CSR which really has maximum of 3 - 4 nH inductance.  4 nH adds 0.02 Ohm reactance in series with a 0.1 Ohm csr.

Gyula
   

Group: Moderator
Hero Member
*****

Posts: 2501

Thanks guys,  lets see what is going to happen.

Itsu
   
Newbie
*

Posts: 23
Itsu,

Glad to see someone doing something energy related on this forum!  I really liked your CSR sweeps.  Excellent demonstration of the effect inductance has on low resistance sense resistors. 

Keep in mind that another trap many fall into is with regard to a waveform's frequency content.  I've stated this ad nauseum in the past, but unless a waveform is a perfect sine wave, it contains harmonics.  The waveforms you posted of your OZ replication are a good case in point. A scope's trigger rate might read 830KHz, but with all the non-sinusoidal peaks and squiggles, there is obviously a large harmonic content.  Just the second and third harmonics alone are going to get you up over 2.4Mc and there looks to be some fifth and a bit of seventh in there as well.  Use the FFT on your scope to see this.

When the 0.1R resistor is used, the higher frequency harmonics see a significantly reduced load because the CSR's impedance is greater at those higher frequencies, hence the increase in the amplitude of higher frequency components, or "hash".

The DDS's are fast enough that they are calculating the contribution of these higher frequencies using a CSR value that can be off by a very significant amount at the higher frequencies those harmonics present.

PW
« Last Edit: 2020-12-18, 21:22:50 by picowatt »
   

Group: Moderator
Hero Member
*****

Posts: 2501

Hi Picowatt,

good to see your here.

Yes, this forum can hardly be considered as a technical one nowadays  :(

Not sure what you mean by "OZ replication", but its a good reminder to keep in mind when dealing with a large harmonic content.

Regards Itsu
   

Group: Moderator
Hero Member
*****

Posts: 2501
I received my "measuring block" pcb and was able to make some measurements using my VNA.

Below picture shows how i measured it which is how CaptainLoz also measured it using a voltage probe with ground lead to measure the voltage across the 0.1 Ohm csr @ 830KHz.

The red line shows the total measurement path taken before presenting this voltage to the scope to do some power calculations using this voltage as current.

The below graphs (screenshot 1) show the severall measured value's of this measurement path like inductance (Seriel L), Impedance (Z) and Resistance (R) (sweep 10KHz to 10MHz).

At 809KHz (the working frequency of the CaptainLoz device) see the red marker, we have:

Inductance L = 174nH
Resistance R = 232 mOhm
Impedance Z = 915 mOhm

So the voltage measured across the csr is not only from the 100 mOhm resistor, but 9 times higher from a 915 mOhm one        <<<<<<< This is wrong, we cannot add this extra 700 mOhm to the csr!!

Therefor the compensation for the scope to do its power calculations should not be 1:10 , but only 1:1.1 instead.                  <<<<<<< This is wrong, we cannot add this extra 700 mOhm to the csr!!         



To compare, screenshot 2 shows the graphs of the 100 mOhm csr ALONE, so without pcb, connection leads and scope voltage probe leads.

Again at 809KHz (the working frequency of the CaptainLoz device) see the red marker, we have:

Inductance L = 35nH
Resistance R = 101 mOhm
Impedance Z = 210 mOhm


So the values at the 800KHz working frequency shows an already 2 times (210 mOhm) higher impedance then expected from this 100 mOhm csr at DC.

So when using an RF probe tip directly across the csr resistor this might be the better method needed to do proper power measurement, but even then the impedance at 800Khz is already
twice the value of the 100 mOhm csr at DC (210 mOhm), so not a 1:10 compensation but a 1:5 compensation for power calculations is needed.

Itsu
« Last Edit: 2021-01-09, 11:12:15 by Itsu »
   
Group: Experimentalist
Hero Member
*****

Posts: 1221
Itsu,

Again, excellent work!  The results speak for themselves as they explain the "gain" mechanism of Captain Loz's device.

I know Chris speaks very highly of their measurement protocols on the AU forum but he would be wise to heed what you have published.

Regards,

Pm
   

Group: Tinkerer
Hero Member
*****

Posts: 2507
Commendations Itsu!  Excellent analysis.  Brilliant detail.

I agree with Partzman.  This documented discovery of yours is worthy of widespread proliferation.


---------------------------
"The best teachers are those who show you where to look, but don't tell you what to see." - Alexandra k. Trenfor.
   
Group: Tech Wizard
Hero Member
*****

Posts: 661
Hi Itsu,

Great tests, thanks for sharing.  I would not be surprised if the silence continues on the other side on your recent measurement results with the metal strip resistor and the Measurement Board.

Greetings
Gyula
   

Group: Moderator
Hero Member
*****

Posts: 2501

Thanks guys,   i won't be surpised myself Gyula.

I PMed CaptainLoz to take a look here,  will see what happens.

Itsu
   

Group: Moderator
Hero Member
*****

Posts: 2501

I got some responses on the AU forum on my reactance measurements.

The responses made, about when in resonance loosing reactance etc., are irrelevant imho as the csr is not in resonance so won't loose its reactances.

Again, i am replicating / commenting ONLY on the CaptainLoz device which runs at 830KHz, so therefor i am using this frequency for my measurement tests.

I pointed out where CapatainLoz might went wrong and am waiting for him to show up and do some tests with me.

Itsu 
   
Newbie
*

Posts: 23
Commendations Itsu!  Excellent analysis.  Brilliant detail.

I agree with Partzman.  This documented discovery of yours is worthy of widespread proliferation.


   Agreed,  and I would not be surprised either, if Captainloz does not reply, or confirm the measurement error.

   NickZ
   
Group: Experimentalist
Hero Member
*****

Posts: 1221
Itsu,

Chris is not correct in his network analysis when an inductive CSR is used to measure a resonant circuit while in resonance.  The attached simulation proves the point.

Here we see a .10 ohm CSR with 35nH of inductance as you have measured.  With a frequency of 800kHz, the current in Lres is 383.52ma rms which should produce a voltage of 38.352mv across .10 ohms.  However, we see the the voltage across the CSR is 776.11/10 = 77.611mv rms (the plot voltage is 10x for clarity). 

Thus, the error just from the inductive .100 ohm CSR is 77.611/38.352 = 2.02 times the real value.  This doesn't include any other inductive pickup errors from long circuit leads such as the scope probe ground lead as you pointed out.

Regards,
Pm
   
Group: Tech Wizard
Hero Member
*****

Posts: 661

...

The responses made, about when in resonance loosing reactance etc., are irrelevant imho as the csr is not in resonance so won't loose its reactances.
...



Hi Itsu,

I agree with you and Partzman, the inductive nature of a csr remains valid and this manifests in the increasing AC voltage drop across its terminals A and B as the frequency increases, see the attached circuit schema below.

It is okay that the total circuit can be tuned to its new resonant frequency when we apply an inductive csr in the circuit.  The new frequency can be very close to the previous one.
This is because the metal strip resistor has 35 nH as you found while the L coil (labeled as L2 in Captainloz's video 9 may have at least some ten uH inductance.

 At this new resonant frequency the generator current will be in phase with the generator voltage BUT the voltage across points A and B will not be in phase with the generator current at all.

How much phase offset is created depends on the value of Ls, the inductive part of the csr, of course. And how much the voltage increase would amount to can be calculated by considering the inductive reactance of Ls at the resonant frequency and then one can use the impedance formula for a series Rs Ls circuit applied for the csr.
 
So the combined L+Ls and C reactances do disappear BUT ONLY FROM the GENERATOR POINT OF VIEW, the inductive reactance of Ls does remain between points A and B. 

This is what increases the voltage drop across the csr and then this increased voltage (combined with the voltage drop of the real part of the csr) is measured by the probe of an oscilloscope.
The inductive part, Ls of the csr creates a VLs = I x XLs voltage amplitude where XLs = 2Pi x f x Ls and  I is the generator current at resonance.

All this boils down to this: your measurements shown and deductions you wrote in your Reply #735 https://www.overunityresearch.com/index.php?topic=3691.msg86553;topicseen#msg86553  above remains fully valid and Captainloz should revisit his COP 2 claim which he got by using an inductive csr (in his Measuring Board).

Gyula
   

Group: Moderator
Hero Member
*****

Posts: 2501
Hi Guys,


PM,   thanks, this nicely confirms my real life measurement that my 100 mOhm csr at 800KHz has an impedance of 210 mOhm (so twice its value at DC).


Gyula, thanks, i know Chris is wrong with his comment about the csr loosing its reactance when measuring current (voltage) in a series LC at resonance.

Its true for the whole LC, the reactances cancel out each other, but like you said, the reactance of the individual components L, C  and our R do NOT go away magically at resonance, see picture below where XC and XL keep on decreasing / increasing with frequency during and after resonance.

So the impedance due to the inductive reactance of our csr still continues to grow with frequency despite its in a resonante LC circuit.

But it obviously is useless to further point this out to him, so i won't.


Regards Itsu
« Last Edit: 2020-12-31, 10:32:47 by Itsu »
   

Group: Moderator
Hero Member
*****

Posts: 2501
I was trying to make a video showing how the extra inductance/reactance of a voltage probe with ground lead measuring a csr adds impedance to this csr, but i cannot as the extra amplitude from this extra impedance did NOT show up on the scope.

My vna showed that a 100 mOhm csr (at DC) measured on its own, increases its impedance (resistance plus reactance) at 830KHz to more then double to 232 mOhm.

This still holds.


My VNA also showed that this same 100 mOhm csr mounted on a pcb now measured using a voltage probe with ground lead etc. adds another 700 mOhm of impedance to this, totalling 915 mOhm of impedance at 830KHz of this 100 mOhm csr.

This partially holds, but the part of "adding this extra impedance to the csr" is WRONG.......

we CANNOT add this extra 700 mOhm to the csr total as it is not carrying the current going through the csr.


We should ONLY add the extra impedance created by the inductance/reactance of the current carrying csr, meaning a total of 232 mOhm.

This is still more then twice the original 100 mOhm at DC, so something to take into account.

To get to compensate for this for using power measurements, the scope needs to translate this 232 mOhm by a factor of 1:4.3  (so not 1:10 for a 100 mOhm csr).

The below picture in which i showed in red the path of the extra inductance is still there, but will NOT influence the current measurement hence the red cross.
 
Apologies to anyone who might have been confused.

Regards Itsu
« Last Edit: 2021-01-09, 11:35:38 by Itsu »
   
Group: Experimentalist
Hero Member
*****

Posts: 1221

The below picture in which i showed in red the path of the extra inductance is still there, but will NOT influence the current measurement hence the red cross.
 
Apologies to anyone who might have been confused.

Regards Itsu

Itsu,

I agree with you in regards to the extra inductance of the scope probe ground lead however, this same ground lead is an antenna at 800kHz!  This results in the virtual ground as seen by the probe, to be above or below the actual ground on the CSR.  This too can lead to increased measurement errors.  That is why Tek includes the spring ground clip to measure HF circuitry as you well know.

I had to chuckle at a comment left on AU by one of the members that expensive scopes should surely compensate for parasitic inductance and capacitance.  They do but only for the particular probes used.  The circuit they are connected to is the responsibility of the user regarding these parasitics.  It doesn't matter if the scope costs $500 or $5000, they are all the same in this regard!

Regards,
Pm
   
Group: Tech Wizard
Hero Member
*****

Posts: 661
Hi Itsu,

Thanks for pointing out this mistake I did not notice it either, my bad also.

However, this mistake DOES not undermine the fact that the 0.1 metal strip csr has the 35 nH inductance (practically a piece of wire) and at the 830 kHz test frequency may cause the COP=2 result Captainloz reported.


Partzman did not consider in his Spice simulation the scope probe ground lead inductance so his simulation remains also fully valid as do your all measurement done on several 1 Ohm and 0,1 Ohm resistors,  including where you used your current probe.   

In fact, I find it a much bigger mistake to attempt to explain away the inductive behaviour of the 0.1 Ohm csr by claiming that it disappears at resonance!

Correct measurements and simulation show of course that is not the case! And all the inductive and capacitive reactances of the individual components in the circuit "disappear" only from the AC source point of view: the AC current taken out from the source is in phase with the AC voltage the source provides  as I already mentioned earlier in Reply 743 https://www.overunityresearch.com/index.php?topic=3691.msg86555#msg86555 

Gyula

PS edited for clarity
« Last Edit: 2021-01-09, 16:45:05 by gyula »
   

Group: Moderator
Hero Member
*****

Posts: 2501
Thanks Guys.


What i did find out is that the relative high inductance of this specific 0.1 Ohm csr (@ 800KHz 35nH) is causing a phase shift of up to 40° at 800KHz making it unreliable for power calculations at that frequency.

For testing this phase shift offset i set up a simple circuit consisting of my battery operated FG, a 100 Ohm 1% inductionfree resistor as load and the both csr's (1 and 0.1 Ohm) see diagram below.

As the point of these csr's is to do current measurement for power calculation of the load any (false) phase offset will lead to erroneous power measurements


The below screenshot shows:

green:  current probe signal
yellow: voltage across load (100 Ohm)
blue:   voltage across 1 Ohm csr
purple: voltage across 0.1 Ohm csr   (40° phase offset leading @ 1MHz)
red:    calculated power across load

Video here:  https://youtu.be/y0jtW-tJgYw

It shows that with increasing frequency from 10KHz to 1MHz, the current probe signal (green), the voltage signal across the load (yellow) and the voltage signal across the 1 Ohm csr stays "in phase".

The voltage signal across the 0.1 Ohm csr (purple) however slowly offset till about a 40° phase difference leading compared to the other signals.

This 40° phase shift due to the inductive behaviour of this specific 0.1 Ohm csr will render any power calculation at this frequency useless.

Regards itsu
   

Group: Moderator
Hero Member
*****

Posts: 2501
While waiting for CaptainLoz to respond i found another thread on AU.com in which member John made an seemingly successful replication of a circuit called "Non-Inductive Coil Experiment" which seems to be the basic circuit for creating "above unity".

His thread is called  "John's Non-Inductive Coil Experiment" and is linked here:
https://www.aboveunity.com/thread/john-s-non-inductive-coil-experiment/?order=all#comment-1e61d9ea-991e-4062-ab21-ac990164a1d5

I say successfull as the first comment on his result started with:


Quote
Hi John,

Thank You for Sharing! You have achieved the Goal! Congratulations! You have everything right!
 


So i replicated what John has done, see HIS diagram and what i came up with (see picture):

Amorphous transformer core AMCC200
L1 15 turns 1mm over L2
L2 225 turns 0.8mm
L3 225 turns 0.8mm

PS input 12V
FG 6KHz square wave 10% duty cycle
L2 bulb is 12V/5W and bright.

The below screenshot shows the current in the L3 circuit (green) and shows the so called "sawtooth waveform" which seems to be an indication for above unity as the current continues after the MOSFET goes off until the next cycle (current rushing in from the ambient!?).
Purple signal is the MOSFET gate signal.


Itsu


EDIT  I added an additional picture of this setup, see last picture
« Last Edit: 2021-01-14, 20:39:51 by Itsu »
   
Pages: 1 ... 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 [30] 31
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2021-01-27, 23:25:37