PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
OverUnity Research > Electrical / Electronic Theory and Learning Center > Joule Thief Circuits > Lawrence Tseung sent a Prototype to test... any comments?
2024-11-26, 17:19:41
News: Check out the Benches; a place for people to moderate their own thread and document their builds and data.
If you would like your own Bench, please PM an Admin.
Most Benches are visible only to members.

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 [17] 18 19 20 21 22
Author Topic: Lawrence Tseung sent a Prototype to test... any comments?  (Read 342706 times)
Topic Options img
img  poynt99
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #400 on: 2011-01-24, 19:33:50 »

Group: Administrator
Hero Member
*****

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

Wait for 3 things:

1) Accurate numbers produced by taking the mean average of the wave form

2) Replication of the results by others

3) COP measurements from others using other valid methods

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   
img  aetherevarising
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #401 on: 2011-01-24, 19:50:40 »
Group: Guest
Quote from: ltseung888 on 2011-01-24, 18:44:21
True COP > 1 comparison

The following diagram uses the rough Area Comparison method.  The Input Power Curve is placed side-by-side with the Output Power Curve.  The reference zero (M) line is aligned.  The Output Power Curve is stretched 200% to account for the 10mVV scale as compared with the 5mVV on the Input side.

Without doing the actual area under the graph method or the exact integration method, it is obvious that the Area under the Graph at the Output side is much larger than that on the Input side.

Thus we can safely say with absolute confidence that:
According to the waveforms captured by the Tektronic Oscilloscope by PhysicsProf on Jan 22, 2011, the Coefficient of Performance (Output Power/Input Power) on Prototype A is greater than 1.

Poynt99, can I use COP > 1 now???


I agree with you Lawrence.  I think the proof is unequivocal.  And surely Professor has given us the equivalent of a 'replication'. 

I'd like to be the first to congratulate you Lawrence.  VERY WELL DONE INDEED.

Rosemary
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #402 on: 2011-01-24, 22:33:27 »
Group: Professor
Hero Member
*****

Posts: 3017
Quote from: poynt99 on 2011-01-24, 19:33:50
Lawrence,

Wait for 3 things:

1) Accurate numbers produced by taking the mean average of the wave form

2) Replication of the results by others

3) COP measurements from others using other valid methods

.99

We have shown that the DUT is intriguing -- and I agree that kudos are due to Lawrence for that.  At the same time, I agree with .99 that the three steps he indicates are needed.   I would add that I think it is crucial to have a few of the FLEET devices for testing that have a larger output, above 100 mW if possible, to make the tests easier and still more definitive.
   
img  MileHigh
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #403 on: 2011-01-24, 22:39:18 »
Group: Guest
Quote from: aetherevarising on 2011-01-24, 19:50:40
I agree with you Lawrence.  I think the proof is unequivocal.  And surely Professor has given us the equivalent of a 'replication'. 

I'd like to be the first to congratulate you Lawrence.  VERY WELL DONE INDEED.

Rosemary

The waveforms are drowning in noise when in fact you know ahead of time that they should be smooth curves.  When the LED diode in the secondary pick-up loop (the "output LED") is reverse biased there is no current flow and the power trace should show zero.  Instead you see all of this horrible random noise.  Therefore the data is not even remotely unequivocal.  With all due respects to PhysicsProf this was just a first run at his testing.

Lawrence and Rosemary, you have to be real here and look at the data in an unbiased fashion.  The data is no good, it's only a first attempt.  It's doesn't even remotely resemble what it should look like under proper conditions.

If the noise problem can't be overcome because of the very low power levels, it may be necessary to look at other strategies.  In my opinion the analog power measurement strategy on the supply side, and the thermal power measurement on the output side would be an attractive option if it proves to be too difficult for the DSO to make measurements in this case.

MileHigh
   
img  aetherevarising
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #404 on: 2011-01-24, 22:48:38 »
Group: Guest
Quote from: PhysicsProf on 2011-01-24, 22:33:27
We have shown that the DUT is intriguing -- and I agree that kudos are due to Lawrence for that.  At the same time, I agree with .99 that the three steps he indicates are needed.   I would add that I think it is crucial to have a few of the FLEET devices for testing that have a larger output, above 100 mW if possible, to make the tests easier and still more definitive.

Hello Professor,

I have also just lost a post - and - just out of interest - it's not the first time. 

I get it that the values are small.  But I was rather hoping that it could - nonetheless - be unequivocal.  This because a lot of the tests on these and other forums are at very low values.  I would have thought it's enough to - at its least - start the questions going.

In any event.  I have some data sheets that I'm hoping to send to you.  They're too big for the email and I'll have to find out how this is done.  It also gives a low wattage output from the supply - but that's the sum of voltages that are certainly not attributable to 'noise' levels.  I need you to evaluate this fact though - in advance - and, obviousy, if you're up for it.

Kindest regards,
Rosemary
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #405 on: 2011-01-24, 23:02:48 »
Group: Professor
Hero Member
*****

Posts: 3017
Quote from: PhysicsProf on 2011-01-24, 22:33:27
 I would add that I think it is crucial to have a few of the FLEET devices for testing that have a larger output, above 100 mW if possible, to make the tests easier and still more definitive.

Having an output power of 100 mW plus would solve the concern you raise, MH.   And in any case, I would like to prove/check that the device is scalable -- that is, that the output power can be raised significantly from the first test with a sophisticated oscilloscope.  That's what I will be working on now, getting to higher output power, and understanding what raises the output power.  (I would also like to try the JouleRinger ;)
« Last Edit: 2011-01-25, 01:36:37 by PhysicsProf »
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #406 on: 2011-01-24, 23:08:52 »
Group: Professor
Hero Member
*****

Posts: 3017
  @Rosemary -- sure, I'll take a look at what you send.  At the same time, I REALLY like to take data for myself, even when I'm learning how best to do this...  I'm an experimentalist.
   
img  aetherevarising
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #407 on: 2011-01-24, 23:15:48 »
Group: Guest
Quote from: PhysicsProf on 2011-01-24, 23:08:52
  @Rosemary -- sure, I'll take a look at what you send.  At the same time, I REALLY like to take data for myself, even when I'm learning how best to do this...  I'm an experimentalist.

I've already committed myself to sending you the apparatus.  I just want it first for a public demo.  I feel I owe this much - both to the university and to our academics.  But I really want your input.  It'll be the American 'arm' of this 'validation' exercise.  Intrigued to see where it will go.

Rosemary
   
img  chrisC
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #408 on: 2011-01-24, 23:23:31 »
Group: Guest
Quote from: aetherevarising on 2011-01-24, 19:50:40
I agree with you Lawrence.  I think the proof is unequivocal.  And surely Professor has given us the equivalent of a 'replication'. 

I'd like to be the first to congratulate you Lawrence.  VERY WELL DONE INDEED.

Rosemary

Seriously, Rosemarie. If the Input/Output were close sinusoidal, i would have to agree with you but look at the junk noise - it bears no resemblance, not even close. How can any 'educated' researcher reach such a conclusion? Seemed like people want to believe what they want to believe. The 'proof' is far from conclusive at this stage.

ChrisC
   
img  wattsup
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #409 on: 2011-01-25, 00:08:16 »
Hero Member
*****

Posts: 520
@ltseung888

I don't think those two waveforms are proof otherwise I myself and many others have done that much and more many many times in many builds. Unfortunately it does not mean OU and I know I was jumping up in the air the first time I saw it on my scope. I had 12 volts in and spikes of 400 volts, but it means nothing, unfortunately. I also did it with micro amps.

The easiest way to prove OU is to take two capacitors, one as feed, one as output tank. Take two identical capacitors and try it. Should be at least 1000uf to keep it running for a decent amount of time. Enough for it to cycle through many times until it stops. Then measure the voltage of each. Bingo. This will tell you if you are OU or not at very low cost.

Otherwise, on the output led put in parallel a germanium diode pointing to a capacitor and see how much energy you can make while lighting the led. Then check voltage on both caps. This is quick and easy.

Now you have to please explain to me one thing in what you sent the Prof. The question is why on an OU device are you putting an LED in front of the coil and a second one on the output. Usually an OU device will have a source (ok battery), device (ok toroid/transistor), but then why put that a load before the toroid. If the Toroid/Transistor is the OU device, why would you want to waste battery power by lighting that led. You can save that power right away. If you decide to do the cap/cap method, I would remove that first led and even the second led and just let the device run with some well sized capacitors. This will tell you right away if there is anything special in this DUT.

wattsup



---------------------------
http://etherimpress.com
   
img  MileHigh
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #410 on: 2011-01-25, 00:19:37 »
Group: Guest
Wattsup:

Just to review a concept that I stated before.  You can't necessarily use capacitors.  They are not a "one size fits all" solution.  When a device is some sort of a source of power, how much power gets transferred into the load depends on the nature of the power source and the nature of the load.  You simply can't substitute the load for a capacitor in every case and expect to get good results.  The load has to be tailored to the nature of the power source.  In many cases a capacitor looks like a short circuit and more often than not the power source does not want to drive a short circuit.

MileHigh
   
img  ion
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #411 on: 2011-01-25, 00:31:31 »
Group: Elite
Hero Member
******

Posts: 3537
It's turtles all the way down
http://www.overunityresearch.com/index.php?topic=602.0

I could post the Excel spreadsheet of efficiency data over 4 different voltages 1.5 V, 3.0 V , 4.5 V and 6.0 V or typically representing one to four zinc-carbon cells. This from actual bench tests of the replicated circuit (touted by L Tseung as his "invention")

No one would like or believe the numbers as they show disappointing efficiency.

So I won't as it may affect illusions.
« Last Edit: 2011-01-25, 03:35:43 by ION »


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #412 on: 2011-01-25, 01:42:11 »
Group: Professor
Hero Member
*****

Posts: 3017
Quote from: aetherevarising on 2011-01-24, 23:15:48
I've already committed myself to sending you the apparatus.  I just want it first for a public demo.  I feel I owe this much - both to the university and to our academics.  But I really want your input.  It'll be the American 'arm' of this 'validation' exercise.  Intrigued to see where it will go.

Rosemary

Alright -- I'd be happy to test it also.

I wrote about the need for "scalability" -- the ability to scale the output power up.  One example -- the captret is kind of interesting, I've been reading out it today, but it has VERY low power output and probably the explanation for self-charging capacitors (at low low power levels) is quite prosaic.  In any case, unless this output power CAN BE scaled UP significantly, the capret is absolutely hopeless as a power source.

We'll see whether the Tseung device can be scaled up from Prototype A (test results above, 18 January).  Lawrence says this is possible, and I have seen higher readings so I'm encouraged.
   
img  poynt99
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #413 on: 2011-01-25, 03:00:11 »

Group: Administrator
Hero Member
*****

Posts: 3217
It's not as complicated as it may seem...
Here is a perfect example where knowing the scope and its limits is important for critical measurements. Not your fault Professor, but please take careful note of the following:

Close examination of the picture will reveal that the CH2 blue current trace falls mostly below the zero mark, clearly showing about a -4mV offset in the scope. The signal acquisition is in error and will therefore produce erroneous results for the multiplication.

In all Tek scopes there is a Signal Path Compensation (SPC) which needs to be run prior to critical measurements because of this inherent "problem" in the scope channels. The offset is quite significant relative to the signal peak to peak level, and therefore will grossly skew the current measurement. It is wise to allow the scope to stabilize for 20 to 40 minutes, then run the SPC. This is mainly to compensate for DC offsets in the signal path so that 0=0. Info about how to run the SPC can be found in the TDS3032 manual on pages 1-4 and 3-77. keep in mind that the SPC should be run any time the ambient temperature varies more than 10ºC. I would try to take your input and output measurements within a few minutes of each other to ensure drift is not a significant factor.

Sorry Professor, but these measurements are not useful at all. Please for future measurements, run the SPC calibration prior to taking any critical measurements. And also of great importance; use the MEAN computation on the instantaneous power wave form.

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #414 on: 2011-01-25, 03:17:27 »
Group: Professor
Hero Member
*****

Posts: 3017
Quote from: poynt99 on 2011-01-25, 03:00:11
Here is a perfect example where knowing the scope and its limits is important for critical measurements. Not your fault Professor, but please take careful note of the following:

Close examination of the picture will reveal that the CH2 blue current trace falls mostly below the zero mark, clearly showing about a -4mV offset in the scope. The signal acquisition is in error and will therefore produce erroneous results for the multiplication.

In all Tek scopes there is a Signal Path Compensation (SPC) which needs to be run prior to critical measurements because of this inherent "problem" in the scope channels. The offset is quite significant relative to the signal peak to peak level, and therefore will grossly skew the current measurement. It is wise to allow the scope to stabilize for 20 to 40 minutes, then run the SPC. This is mainly to compensate for DC offsets in the signal path so that 0=0. Info about how to run the SPC can be found in the TDS3032 manual on pages 1-4 and 3-77. keep in mind that the SPC should be run any time the ambient temperature varies more than 10ºC. I would try to take your input and output measurements within a few minutes of each other to ensure drift is not a significant factor.

Sorry Professor, but these measurements are not useful at all. Please for future measurements, run the SPC calibration prior to taking any critical measurements. And also of great importance; use the MEAN computation on the instantaneous power wave form.

.99

In decades of taking data, my colleagues and I have often found that an instrument has shown a measurement bias such as this -4mV offset that you mention.  Certainly it is better to compensate for such offsets before data is taken -- and I will do this next time, .99 -- thank you for this poynter.  And we'll use the mean computation on the instantaneous power waveform.

AT the same time, my experience is that usually we can make the correction off-line, especially if the raw data are saved.  This is more difficult in this case, and probably not worth the effort to do in detail since we will be taking measurements again.  

But to say " but these measurements are not useful at all"  seems a bit overstated to me...  Let me ask, .99, does not this -4mV offset make the channel 2 blue trace conservative as a measure of the current in the output and input resistors?  And recognizing this correction, can we not say that our measurements are conservative?
   
img  poynt99
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #415 on: 2011-01-25, 03:50:48 »

Group: Administrator
Hero Member
*****

Posts: 3217
It's not as complicated as it may seem...
Quote from: PhysicsProf on 2011-01-25, 03:17:27
But to say " but these measurements are not useful at all"  seems a bit overstated to me...  Let me ask, .99, does not this -4mV offset make the channel 2 blue trace conservative as a measure of the current in the output and input resistors?  And recognizing this correction, can we not say that our measurements are conservative?

Actually, no.

As an example, for the input power measurement, you end up with a net negative power coming from the battery, that means power going INTO the battery.

However, we do not realize this by looking at your RMS computation on the power trace, because by taking the RMS computation of a wave form, the result is always a positive number!!! Again, this is why you must take the average, not RMS. If you take the average of your power trace shown in the picture, it will be a NEGATIVE number!!! You can also realize it by looking at the red trace, it's almost all below the zero mark.

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   
img  gridbias
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #416 on: 2011-01-25, 05:24:02 »
Group: Guest
This is getting interesting.

I have a question

What is the noise ?
It looks like pn junction noise

   
img  wattsup
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #417 on: 2011-01-25, 05:38:15 »
Hero Member
*****

Posts: 520
@MH

The present load is an LED.
What is an LED? It is a diode that consumes x to light b. A feed cap with dioded capacitor output (capdcap) is still a good enough analogy and will not effect the working of the DUT. It all depends on your choice of the cap value. A pf won't do.

The Prof showed frequency rises as feed voltage drops. That would be perfect for a capdcap.

Usually, you would think the voltage (LED brightness) would go down as the feed voltage drops, but in this case the Prof says the brightness was rather linear. This can be captured in a capdcap.

If you can find as feed cap at 24000uf 2 vdc capacitor (maybe 24 x 1000uf) and as output cap a 24000uf 6 vdc, both capacitors should have the same drop in that if both capacitors are loaded to 2vdc and then apply a 1 volt led on both at the same time, both should die off at the same time. Using the 6 vdc range for the output cap is just to provide the output with a higher captation potential while lowering any back effect while it gets filled.

My usual testing is with LEDs that light up at 4.5 volts and my cap tank is a huge 88uf 1200vdc. I get spikes in there up to 800 volts.

The major point is standardizing how you look at things which is very personal indeed. I always have my huge load cap there ready for testing. Anything out of the ordinary between my pulsing method and the output will be seen and noticed because I am always using the same output method. Working with a capdcap with your devices will still give you a quick means comparable in performance. You only measure the always stabilized cap voltages. No need for a scope.

Here is an example.

Feed cap is a 250kuf 2v loaded to 1.5x volts.
Run device,
Identical output cap goes from its internally floating .024 volts to 2.625 volts.
And the feed cap still has .456 volts in it.

Question 1: Is that OU or not?
Question 2: Can capacitors lie?
OK Forget Question 2. Ahh too late.

Anyways, I am looking for a way that does not involve so much technically biased measuring instruments. A method that is quickly repeatable and non-intrusive for low voltage DUTs. Once you have done 3-4 runs with totals for an average, then you are in a position to add a scope probe and see how that probe effects a next set of tests.

Also, if it can save 140 miles round trip per test intervals, this may be a good option.

wattsup



---------------------------
http://etherimpress.com
   
img  ltseung888
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #418 on: 2011-01-25, 08:13:23 »
Group: Guest
The scope shots

Let us forget the small Tseung FLEET Comparison Index Prototype A used by the PhysicsProf for a moment.

Let us focus on the better one obtained on July 14, 2010 in Hong Kong.

By comparing the two graphs, can we conclude that the RHS graph has much more area than that on the LHS?  I believe it is irrelevant to talk about average, mean or rms once we have the Instantaneous Power Curves.  The correct scientific value is in fact the integration or the area enclosed by the curve.

Let me ask the question – if we have similar screen shots from dozens of prototypes done by independent verifiers such as PhysicsProf at various Universities, will any one accept these scope shots as evidence of COP > 1 ???
   
img  MileHigh
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #419 on: 2011-01-25, 12:09:09 »
Group: Guest
Wattsup:

This circuit has diodes in it which are nonlinear.  So the lower the supply voltage the more proportional energy will be burned off in the diodes and the higher the supply voltage the less proportional energy will be burned off in the diodes.  Therefore the if you are trying to investigate the energy audit trail for the circuit you want to use a constant supply voltage.

The output load for this circuit is basically a resistor in series with a diode, and it is being driven by an AC voltage so current only flows when the diode is forward biased.  Again, this is a nonlinear load.  If you replace the resistor with a capacitor then you are not accounting for the power that is being transferred to the diode (the LED).  The lower the capacitor voltage the more power is burned off in the diode and less power is being put into the capacitor.  The higher the capacitor voltage the less power is being burned off in the diode and more power is being put into the capacitor.  Also, the higher the capacitor voltage gets, then the current only flows when the secondary winding is higher in potential than the capacitor voltage plus the forward drop of the diode.  In other words as the capacitor starts to get charged it starts to choke off the current flow and limit the conduction angle for the AC waveform.  It's a complete mess and simply will not work in this particular case.  You are trying to measure the energy being dissipated in a diode and a resistor over a certain period of time and swapping a capacitor into this circuit simply will not work.

Quote
Feed cap is a 250kuf 2v loaded to 1.5x volts.
Run device,
Identical output cap goes from its internally floating .024 volts to 2.625 volts.
And the feed cap still has .456 volts in it.

Question 1: Is that OU or not?
Question 2: Can capacitors lie?
OK Forget Question 2. Ahh too late.

My first guess is that your two caps are not identical in value and one of them may be outside of the tolerance specification.  Have you measured their capacitance values?

If you think you have OU then there is a logical next step to your testing.  Put a resistor on your output cap that just got charged to 2.625 volts and burn off energy until it is down to 1.5 volts.   Now swap caps around so the output cap becomes the feed cap and run the test again.  If you think you have OU you should be able to do this indefinitely.  You also can't forget the self-charging effect that can happen with supercaps and electrolytic caps but based on your reported voltages I'll cautiously assume that the resistor burn-off will be much larger than any potential self-charging effects.

MileHigh
   
img  MileHigh
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #420 on: 2011-01-25, 12:41:00 »
Group: Guest
Quote from: ltseung888 on 2011-01-25, 08:13:23
The scope shots

Let us forget the small Tseung FLEET Comparison Index Prototype A used by the PhysicsProf for a moment.

Let us focus on the better one obtained on July 14, 2010 in Hong Kong.

By comparing the two graphs, can we conclude that the RHS graph has much more area than that on the LHS?  I believe it is irrelevant to talk about average, mean or rms once we have the Instantaneous Power Curves.  The correct scientific value is in fact the integration or the area enclosed by the curve.

Let me ask the question – if we have similar screen shots from dozens of prototypes done by independent verifiers such as PhysicsProf at various Universities, will any one accept these scope shots as evidence of COP > 1 ???


Lawrence,

That data capture is also not acceptable.  There is a huge amount of ringing on the output power waveform implying negative power, and negative power doesn't even make sense in this case.  The output waveform time base is also way too slow, so you are seeing subsampling aliasing effects when the scope tries to map the high-frequency waveform into the limited number of pixels on the display screen.  You simply don't have enough data to make any area estimations.

You have to compare the average power in and compare it to the average power out and the July 14, 2010 graph is totally unacceptable for this.

MileHigh
   
img  wattsup
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #421 on: 2011-01-25, 13:14:30 »
Hero Member
*****

Posts: 520
@ltseung888

Again, such waveform comparisons have been shown on forums for years by all of us but it does not mean OU.

@MH

The LED also has its minimum voltage under which all energy is lost before the LED will light up. A germanium diode has only a .2 volts level, which is probably even less then the LED so even less chance of wasted energy for proper accounting of final energy transfers going on inside that toroid.

The first LED is receiving straight DC from the battery at its gradually decreasing voltage level. The second LED is getting pulsed DC with voltages that could be as high as 100 volts easy and the frequencies shown by the Prof. The second LED is basically acting like an RMS or means leveller so you see visually the light is still lit because anything above 50-52 cycles in our eyes will be like constant illumination. That output if 100 volts is way above the battery voltage and even higher above the 0.2 volts of the germanium diode and this should all pass through into the dcap side.

Look I am not saying a precise measurement with a scope is not desirable but it seems this will never be the case with low voltage DUTs. Getting to know how to use a capdcap by first doing preliminary testing to see the real dynamics of this is both low cost and in my view proof enough to show preliminary results without relying on scopes.

The only fail safe method is to treat such devices as joules-in versus joules-out and this can only be done with the right caps. I am sure if anyone in China is reading this, they are already working on it. lol

wattsup


---------------------------
http://etherimpress.com
   
img  MileHigh
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #422 on: 2011-01-25, 13:40:19 »
Group: Guest
Wattsup:

Just a few final comments and then I'm done.  I assume that by the "first LED" you mean the one across the transistor.  It's actually being pulsed when the transistor switches off, but the bulk of the pulse energy actually goes into the secondary winding.  At least that's what it looks like from my perspective.  One more time, I admit that to be sure I would have to check it on the bench.

If you were to change the secondary winding for a large capacitor and a germanium diode, then there is a good chance that when current starts to flow in the circuit for the first time, that most of the power will be dissipated in the resistance of the wire windings of the secondary winding itself.  That's because you have an EMF source pushing current through the resistance of the windings and the germanium diode.  When you first start the capacitor looks like a pure short circuit.  So you end up with a big chunk of the energy being dissipated in the wire windings themselves at the start of the testing, and not going into the capacitor.  There are just too many pitfalls for using a set of source and load capacitors for this particular circuit.

MileHigh
   
img  poynt99
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #423 on: 2011-01-25, 13:45:36 »

Group: Administrator
Hero Member
*****

Posts: 3217
It's not as complicated as it may seem...
Quote from: ltseung888 on 2011-01-25, 08:13:23
The scope shots
By comparing the two graphs, can we conclude that the RHS graph has much more area than that on the LHS?
It would be incorrect and unscientific to make any conclusions based on a visual comparison of those wave forms. High voltage spikes does not automatically indicate a large average power or energy.

Quote
I believe it is irrelevant to talk about average, mean or rms once we have the Instantaneous Power Curves.  The correct scientific value is in fact the integration or the area enclosed by the curve.
You noted previously that you accepted the fact that the average and integral ratios would be the same, and taking the integral is an unnecessary step. Now you disagree? Rose has even confirmed what I said, and admitted that I was correct about this.

So to summarize again:

1) calibrate the scope channel prior to measurements.
2) obtain the instantaneous power using the multiply function in the scope to multiply channel 1 voltage by channel 2 current.
3) use the "mean" or "average" function in the scope on that instantaneous power trace to obtain the running average power.
4) compare the average input and output power values.

Anything short of this procedure will produce erroneous results. Please don't attempt to make quantitative assessments based on a visual comparison of spiky wave forms, it's unscientific and prone to gross errors.

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   
img  humbugger
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #424 on: 2011-01-25, 14:56:10 »
Group: Guest
Quote from: aetherevarising on 2011-01-24, 19:50:40
I agree with you Lawrence.  I think the proof is unequivocal.  And surely Professor has given us the equivalent of a 'replication'.  

I'd like to be the first to congratulate you Lawrence.  VERY WELL DONE INDEED.

Rosemary

I think the discussion and the flaws that have been pointed out since the above proclamation of "success" and "unequivocal proof" should give all readers some insights:

1)  Presenters of claims of OU devices are typically very biased and all too eager to take the first shakey ill-founded hint of OU and proclaim it to be unequivocal proof.  This should put some of Rosemary's previous claims of "absolutely knowing for sure" that her device has COP>17 in the proper context.  What she and others like Lawrence rely on as being unequivocal evidence has inevitably been found to be grossly flawed in fact.

2)  The quality of the tools used to make measurements does not ensure quality measurements.  Many seem to believe that if they have a modern Tektronix DSO on the bench, their results and conclusions are somehow inherently unequivocal.  This thread shows the terrible flaws in this thinking.  It is the test methodolgy and the knowledge of how to use the features of the equipment and how to use probes correctly, etc.  that determine the validity and accuracy of the results.

3)  Were these folks operating in an environment without critical thinkers and people who know how to make proper measurements, you can see how easily an erroneous conclusion could be solidified into "granite fact" and take on its own momentum, becoming another episode in the mythology of "successful OU devices".

I still don't understand why the testers are ignoring suggestions such as using simple RC integrators on these tests (or at least as a parallel path).  In any case, we should all be grateful for those here who spend at least some of their time on critical thinking, instead of jumping to the first conclusion that agrees with their bias, no matter how thin the "evidence".  It's what makes OUR different from many "cultist" forums.
   
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 [17] 18 19 20 21 22
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-26, 17:19:41