Pics of the assembled unit that arrived yesterday.
I added battery holder leads roughly the same length as what Lawrence had in his photo.
There was one cold solder joint that was loose and unconnected. I did a quick job of getting the solder to flow on the mag wire, careful not to disturb the joint too much. Another cold joint was left alone, as it appears there must be some "good" contact there, good enough for the base drive to get through anyway.
.99
Some test results of this unit:
TEK00017.PNG indicates an average INPUT power of 63.31mW.
TEK00019.PNG indicates an average OUTPUT power of 237.1mW/10 = 23.7mW.
The light output from these tiny super-brights is impressive.
Lawrence, let me know if you feel I am doing anything wrong. The Po/Pi ratio is quite low compared to your findings.
.99
Quote from: poynt99 on 2011.02.11, 04:39:20
Some test results of this unit:
TEK00017.PNG indicates an average INPUT power of 63.31mW.
TEK00019.PNG indicates an average OUTPUT power of 237.1mW/10 = 23.7mW.
The light output from these tiny super-brights is impressive.
Lawrence, let me know if you feel I am doing anything wrong. The Po/Pi ratio is quite low compared to your findings.
.99
Lawrence's claim that this pre-assembled LTJT demonstrates over unity in it's original configuration has been refuted.
There are important lessons to learn here with respect to how to take proper measurements, to understanding how electronic circuitry works, and to understanding the energy dynamics of electronic circuits.
MileHigh
Quote from: poynt99 on 2011.02.11, 04:39:20
Some test results of this unit:
TEK00017.PNG indicates an average INPUT power of 63.31mW.
TEK00019.PNG indicates an average OUTPUT power of 237.1mW/10 = 23.7mW.
The light output from these tiny super-brights is impressive.
Lawrence, let me know if you feel I am doing anything wrong. The Po/Pi ratio is quite low compared to your findings.
.99
Dear Poynt99,
Now try the following:
1. Use a DC Power Supply*** and lower the voltage. In our test, the Vrms was lowered to 80mV. The Vpp was 600 mV. Please compare your results with the ones I supplied.
2. Please take the Vpp readings as well. That will give grounds for clearer comparison. When the Input Vpp was 600mV, the Output Vpp was 10.4V.
3. The dip into the negative region for our Power Curve was much higher. If the correct explanation for negative area were – more feedback to the source, your feedback to the source is much lower than ours. May be the lowering of the input power will explain that.
Continue to have fun.
You can start to assemble the other LTJT and vary the parameters much more.
*** You can get a really cheap DC Power Supply for less than the Federal Express Mail Price. I paid US$45 for shipping. I believe the cheap DC Power supply from China cost less than that.
Hi Guys,
Just wanted to point out the non linearity of both the phase angle and the Voltage, Current curves (and thus the resulting power).
Q. Can you identify where the load LED turns on and where it turns off?
Q. Is it possible to have a negative power where both the Voltage and Current are positive? (Hint, notice slope of power trace compared to Voltage and Current)
Q. How do we know if there is capacitance involved? (Hint "ELI the ICE Man") Where is this capacitance if it exists?
Q. What would happen to the PO / PI ratio if we used a One Megohm output CSR? How about 100K? 100 Ohms?
Q. Would using a 0.1 Ohm output CSR allow more current to flow in the output stage?
Q. Does a single test in a qualified lab substantiate or refute a claim?
Q. Is it ok to ground the output section to the input section through the scope grounds?
8)
Quote from: Harvey on 2011.02.11, 09:07:36
Hi Guys,
Just wanted to point out the non linearity of both the phase angle and the Voltage, Current curves (and thus the resulting power).
Q. Can you identify where the load LED turns on and where it turns off?
Positive current indicates when the LED is forward-biased, and the negative current when reversed-biased. These LED's are quite "leaky" in the reverse bias condition, which is why there is power dissipation even when the LED is "OFF". So in terms of the question, the load is always ON in this case, it's just a matter of degree, depending on the bias.
Quote
Q. Is it possible to have a negative power where both the Voltage and Current are positive? (Hint, notice slope of power trace compared to Voltage and Current)
No. p(t) is simply the product of v(t) and i(t), and the basic math rules apply.
Quote
Q. How do we know if there is capacitance involved? (Hint "ELI the ICE Man") Where is this capacitance if it exists?
It can be argued that capacitance is always involved. There is inter-winding capacitance, capacitance in the junctions and leads of the transistor, capacitance across the LED junction, capacitance to the bench and environment, etc.
Quote
Q. What would happen to the PO / PI ratio if we used a One Megohm output CSR? How about 100K? 100 Ohms?
There should be better power transfer as the impedance gets closer to that of the input side, i.e. about 1 Ohm or so in this case.
Quote
Q. Would using a 0.1 Ohm output CSR allow more current to flow in the output stage?
In theory, yes.
Quote
Q. Does a single test in a qualified lab substantiate or refute a claim?
In a perfect world, yes. However, someone should always double-check the test setup, the methods, and the results.
Quote
Q. Is it ok to ground the output section to the input section through the scope grounds?
In theory, yes.
In both the simulation and the actual test, it made no difference to the wave forms or measured power levels.
.99
@poynt99
Maybe some dumb questions but here I go anyways.
I see in the photos that you have three probes on the device at the same time. On a low voltage pulsing device such as this any probe may effect the overall function simply by extending the conductive paths.
With the three probes (1,2 and 3) on and while looking at the three waveforms....
1) what happens to waveform 2 and 3 if you removed probe 1
2) what happens to waveform 1 and 3 if you removed probe 2
3) what happens to waveform 1 and 2 if you removed probe 3
4) Under the three above conditions, can you notice a change in the brightness of the LEDS.
If the answer is yes to 1, 2 or 3, then the probes are effecting the circuit.
5) Then, is it possible to use probe 1, grab the screen, remove probe 1 and use probe 2, grab the screen, then remove probe 2 and use probe 3 and grab the screen again. This way only one probe is on the device at any given time.
6) Is it possible to probe without the ground leads of the probes connected?
7) Also, are you probing before or after the feed LED. Seems to me anything before the feed led or on the feed led is irrelevant since it is only the energy after the feed led or across the drive coil that should be considered and compared to the output.
8) Maybe one last thing or definitely a repetition. If you have enough juice in there to light up LEDs, then you have more then enough to pass a germanium diode and into a capacitor tank of low voltage but high enough uF. I would really consider running this on a capdcap method because it seems those LEDs will skew the results. Or at best remove the feed LED and replace it with a germanium diode going in the same direction as the LED was.
wattsup
I know these are many points so there is no real need to answer them all.
wattsup,
There are 4 probes used at all times for these measurements. Two for the INPUT power (scope 1) and two for the OUTPUT power (scope 2). The probes are placed as per this drawing (http://www.overunityresearch.com/index.php?action=dlattach;topic=717.0;attach=3603;image), and they are indicated by P1T/P1G, P2T/P2G (T being TIP, and G being GND) and so on.
The tests are being performed as per the parameters measured by Lawrence and his team of experimenters. The device should not be modified from the original in order to properly compare results.
.99
DC power supply used in Hong Kong
It is likely that the results of this particular prototype were obtained with the DC Power Supply as shown in the diagram.
We used both AA battery and the DC Power Supply for our testing. When the Input Voltage was way below 1.0V, there was a good chance that the DC Power Supply was used.
However, we do not necessarilyy need to use the DC Power Supply to get very high COP values as shown in the Hong Kong University testing in the second photograph. If Poynt99 can get a DC power supply that can drop to 600mV peak-to-peak, the chance of hitting pseudo resonance is much higher than using an AA battery.
If not, Poynt99 can build a prototype from the components supplied and try to produce the same result as that at Hong Kong University. Tuning can be fun. FLEET is not the very tolerant Joule Thief!
Continue to have fun.
I have a variable DC power supply.
I will connect it up and see how low a supply voltage the assembled unit will still operate. Then I'll take some measurements again.
.99
Quote from: poynt99 on 2011.02.11, 16:23:15
I have a variable DC power supply.
I will connect it up and see how low a supply voltage the assembled unit will still operate. Then I'll take some measurements again.
.99
Great. Looks like you are the only one in OUR forum who has the necessary equipment to have fun. 8) ;D
Just for reference, I am reposting Lawrence's data for the fully assembled prototype that Poynt is testing:
QuoteDear Poynt99,
This particular prototype showed Tseung FLEET Comparison Index of 64 rms when built on Oct 10, 2010 in Hong Kong. It is soldered and not much tuning can be done.
The details are as follows:
Windings:
11 turns Joule Thief type
22 turns Transformer type
0.5 mm diameter wire
Frequency = 147 KHz
Input
Channel 1 (Instantaneous Voltage)
Vpp = 600 mV
Vrms = 80 mV
Channel 2 (Instantaneous Current, 1 ohm resistor)
Vpp = 188 mV
Vrms = 60 mV
Output
Channel 1 (Instantaneous Voltage)
Vpp = 10.4 V
Vrms = 3.4 V
Channel 2 (Instantaneous Current 10 Ohm resistor)
Vpp = 324 mV
Vrms = 92 mV
Calculations:
Input Power pp = 0.6 x 0.188 = 0.1128 watt
Input Power rms = 0.08 x 0.06 = 0.0048 watt
Output Power pp = (10.4 x 0.324)/10 = 3.369 watt
Output Power rms = (3.4 x0.092)/10 = 0.0306 watt
COPpp = 30
COPrms = 64
The verification is simply to display the waveforms and check whether the true COP is greater than 1. The actual numbers are unlikely to come out identical as in all pseudo resonance experiments.
*** Please do not modify this particular prototype as it may be shipped to other sites for additional verification later. Use the components package to build another one. Thank you.
Lawrence
Lawrence,
Assuming that the copy/paste above is for the LTJT that Poynt is working with, I have some questions about this data:
Who generated this data? Was it you or Aaron Quant? Perhaps somebody else?
QuoteInput
Channel 1 (Instantaneous Voltage)
Vpp = 600 mV
Vrms = 80 mV
Whether you used a 1.5-volt AA battery or a power supply set to 600 mV as the power source, quoting peak-to-peak or RMS values for a DC power source doesn't really make sense. As you can see in Poyn't waveforms, the output voltage for the battery is more or less a constant DC value. The voltage drops slightly when Joule Thief is drawing current from the battery. This is all normal.
So the question for you is why are you quoting peak-to-peak and RMS values for what is supposed to be a DC voltage source that powers the Joule Thief?
Thanks,
MileHigh
I went to the waveforms posted by .99 in order to check things another way.
Both the Input and Output power waveforms display a characteristic triangular pattern, for which it is rather easy to calculate the Energy per cycle. The energy can be calculated simply from 1/2*base*height, where the base is the length of the pulse in microseconds (Time) and the height is the Power, in mV*V or V*V. For the input, this is approximately:
Einput ~ 1/2 * 4 uS * 150 mV*V ~ 300 mV*V * uS
The output power waveform also shows a small amount of power over about 4us, and the total Energy per cycle is roughly:
Eoutput ~ 1/2 * 1.2 uS * 1800 mV*V/10 + 4uS*15mV*V/10 ~ 114mV*V * uS
The ratio Eoutput / Einput by this method for one cycle is approximately 114/300 = 0.38 .
Using the average Power input and output method, .99's method, we get PoutAve/PinAve = 23.7/63.1 = 0.38 .
I conclude that the numbers check out and I do not challenge .99's method.
However, his coefficient will improve if he does as I suggested, for instance-- simply cutting one wire to the LED connected to the transistor, which LED is quite superfluous IMO. However again, I understand that Lawrence has requested that nothing be changed in this prototype so that it can be sent to others for further testing.
OK,
I've tested this unit on my other bench using my 1GHz scope and variable power supply.
The device would operate, i.e. "oscillate" only down to a (RMS or MEAN) voltage of right around 0.6VDC. Below about 0.58V, the unit would cease oscillation.
Further to MH's comments on the given INPUT data, I concur. Vpp measurements should not be used in any of these measurements, as the true COP will be skewed significantly as a result.
For instance, While I had about 600mV as the supply voltage, and the scope was displaying the supply voltage (using either RMS or MEAN), changing the measurement to a "Vpp" setting changed the display to about "200mVpp". It would appear that when the scope is told to measure P-P, it uses an "AC" coupling, and discards the DC voltage component in the process (and this makes perfect sense). So this is only now measuring "noise" on the DC supply and does not represent an accurate value to use for any calculations.
I have to conclude Lawrence, that I can not verify your results or claims in this case, using this particular unit. It appears that some incorrect assumptions may have been made also in terms of using Vpp and RMS in the calculation process, and as we've been through the debate about correctly processing the p(t) wave form, similar rules apply to the use of Vpp and RMS where it is not warranted.
It is best advised, that the individual voltage and current "measurements" displayed by the scope, be ignored or not used. The focus must remain on only taking the MEAN of the p(t) wave form, for that is the only computation that provides a true indication of power.
.99
Thank you professor for verifying my data using the good old fashioned method you described. O0
I will be building one or two more of these devices, and I shall leave this assembled unit as is.
I also hope to have a little surprise that some may find quite interesting, exciting, and even sobering. ;)
Regards,
.99
Thank you, .99, for your measurements and your willingness to build/test further devices along this line.
Thank you, Lawrence, for your willingness to send prototypes to people for testing, and your willingness to debate regarding the proper methods for measurements. This shows your sincere desire to learn the truth about what is happening in these circuits.
Quote from: poynt99 on 2011.02.11, 18:03:30
OK,
I've tested this unit on my other bench using my 1GHz scope and variable power supply.
The device would operate, i.e. "oscillate" only down to a (RMS or MEAN) voltage of right around 0.6VDC. Below about 0.58V, the unit would cease oscillation.
Further to MH's comments on the given INPUT data, I concur. Vpp measurements should not be used in any of these measurements, as the true COP will be skewed significantly as a result.
For instance, While I had about 600mV as the supply voltage, and the scope was displaying the supply voltage (using either RMS or MEAN), changing the measurement to a "Vpp" setting changed the display to about "200mVpp". It would appear that when the scope is told to measure P-P, it uses an "AC" coupling, and discards the DC voltage component in the process (and this makes perfect sense). So this is only now measuring "noise" on the DC supply and does not represent an accurate value to use for any calculations.
I have to conclude Lawrence, that I can not verify your results or claims in this case, using this particular unit. It appears that some incorrect assumptions may have been made also in terms of using Vpp and RMS in the calculation process, and as we've been through the debate about correctly processing the p(t) wave form, similar rules apply to the use of Vpp and RMS where it is not warranted.
It is best advised, that the individual voltage and current "measurements" displayed by the scope, be ignored or not used. The focus must remain on only taking the MEAN of the p(t) wave form, for that is the only computation that provides a true indication of power.
.99
Dear Poynt99,
Please display the waveform and the mean value when the DC voltage is dropped to 600mV. The final COP should change significantly.
Let us focus on the waveform first. That reveals most detail. Please keep the voltage and current waveforms. Their shape and values are significant in our analysis.
Thank you in advance.
@PhysicsProf, thank you for your support and great work. At this moment in time, I would like to "hit on" a pseudo resonance condition where the apparent COP is greater than 1. There may be disputes. But if we have the waveforms, we have something more solid to discuss.
Poynt99 now has the same components as you. It will be interesting to see the results from your two different builds. Beijing also has the same components so we have more checks. I still have my best prototype with me and that prototype has been double, triple checked.
Hitting pseudo resonance is still an art.
May God Guide us all.
Lawrence,
You can ignore my questions if you choose to, but note for certain that the analysis of any circuit has to be based on understanding the concepts and making proper measurements and understanding and interpreting the evidence.
QuoteHitting pseudo resonance is still an art.
There is no "pseudo resonance" going on with a Joule Thief and there is no "art." The circuit operates based on well understood principles which have already been discussed. You can expect that the analysis that you see done by Poynt and hopefully others will elaborate on what we have seen so far.
This can be a beneficial learning experience for all involved, and that should include you. I am not getting a sense that you want to learn and understand. Please open your mind and try to absorb the information. Sometimes it may not be what you were expecting or hoping for, but you should still use the experience as an opportunity to learn and grow.
MileHigh
Lawrence,
The wave forms are very much the same as when the supply is around 1VDC. There is no significant amount of power going back to the source.
I did not propose we not show the voltage and current wave forms. What I suggest is that we not use the "Measure" feature to display what the computed RMS or Vpp values are, because they are of no real value and they can lead to erroneous assumptions.
The only computation that needs to be displayed, is the MEAN of the p(t) wave form.
.99
Here are the scope shots of the low voltage run. The "in_mean.bmp" file is black and white because the floppy would not hold all 4 bmp files. I had to use a smaller file format (.tif) for this pic.
in_mean.bmp indicates an average INPUT power of 11.2mW. Note that the supply voltage is 576mV.
out_mean.bmp indicates an average OUTPUT power of 32.4mW/10 = 3.24mW.
.99
Quote from: MileHigh on 2011.02.11, 22:47:19
Lawrence,
You can ignore my questions if you choose to, but note for certain that the analysis of any circuit has to be based on understanding the concepts and making proper measurements and understanding and interpreting the evidence.
There is no "pseudo resonance" going on with a Joule Thief and there is no "art." The circuit operates based on well understood principles which have already been discussed. You can expect that the analysis that you see done by Poynt and hopefully others will elaborate on what we have seen so far.
This can be a beneficial learning experience for all involved, and that should include you. I am not getting a sense that you want to learn and understand. Please open your mind and try to absorb the information. Sometimes it may not be what you were expecting or hoping for, but you should still use the experience as an opportunity to learn and grow.
MileHigh
Dear MH,
We are looking for pseudo resonance behavior. It is like the two tuning fork experiment. If the two tuning forks are 5% different, no sympathetic vibration will occur. When your experiments are not tuned to correct resonance, you may conclude that sympathetic vibration is false.
I know that you said that you did not have oscilloscopes. There is no possibility that you could have ever observed the pseudo resonance effect.
Please give time for Poynt99, the PhysicsProf or the Physics Society to reproduce the same result as those at Hong Kong University and the many teams in Hong Kong and China.
The pseudo resonance behavior normally shows itself with a significant dip in the input power waveform to the negative area. The Voltage waveform will not be smooth. There will be significant pulsing. The Current waveform will show large negative values. So far, the waveforms from Poynt99 have not shown these signs yet.
Another hint is the very large increase in Vpp in the output (10 times or more). Poynt99 will not display those values because he believes that it will mislead the researchers. My opinion is different. When it is hunting for pseudo resonance, every small hint will help.
Wait for Poynt99 to build his own LTJT similar to the PhysicsProf on a breadboard. He can then observe the change in waveform (similar to the above described) with small changes in hole positioning, wire length etc. TK already showed a sample. The moment that Poynt99 could show vast differences in waveforms himself; the analysis will take a completely different turn.
When I have seen dozens of working FLEETs and have one good one with me, I do not care about blind guesses??? I ran two successful workshops at Hong Kong University using my tested equipment. The trouble with resonance is that a small change will shift the condition. I shall just wait for Poynt99 or others to hit a pseudo resonance condition. (My first hit with the Tong Wheel took a few weeks in 2009.)
Once I get back to Hong Kong in a few months, I shall be able to lay hands on the tested and trusted equipment again. The waveforms do not lie. The 108 LED Christmas Tree does not lie. (I am sure the Joule Ringer does not lie also.)
Lawrence,
What you are calling "pseudo resonance" is almost certainly something else. Sometimes circuits can exhibit what's called "metastability" where you typically see unexpected high frequency oscillations. Sometimes the cause can be a cold solder joint which makes an intermittent connection. Note there were two cold solder joints on the sample that you sent to Poynt. Even if you did indeed observe an effect related to metastability, there still won't be any over unity to be found in that situation.
If and when more data comes in, and if it is quality data, you will see that there is no over unity associated with the Joule Thief circuit.
Also there is a very simple explanation for how a Joule Thief can light up 108 LEDs. Inductors have an innate capability to do this. One more time this is not an over unity phenomenon. Certainly the Joule Ringer does not lie. The Joule Ringer always runs out of energy and has to be recharged.
This is all about understanding and the pursuit of the truth. Sometimes the truth is not what we were hoping for but it doesn't matter, it's still the truth.
MileHigh
Quote from: ltseung888 on 2011.02.12, 01:23:22
The pseudo resonance behavior normally shows itself with a significant dip in the input power waveform to the negative area. The Voltage waveform will not be smooth. There will be significant pulsing. The Current waveform will show large negative values. So far, the waveforms from Poynt99 have not shown these signs yet.
Lawrence,
Do you perhaps mean something like this?
.99
Poynt99,
I was hoping to see waveforms closer to the one below which I posted a few times.
The characteristics are:
1. The Input Voltage Waveform showed sharp spikes or pulses.
2. There is a significant negative area in the Current Waveform.
3. The Vpp Voltage Output is more than 10 times the Vpp Input.
4. The Area covered by the Output Power Waveform is much larger than that of the Input Area.
When I have such prototypes, I can use them to light up > 100 LEDs; and produce long lasting devices; recharge the Input batteries etc.
When you have built the new LT-JT from components on the breadboard, let me know. I shall walk you through the various possible tuning steps. You can skype me. My skype name is Lawrence_Tseung. You can also try some of the suggestions from PhysicsProf first – such as removing JT LED; changing the 100-10 ohm combination to 1 ohm etc. Please leave the existing prototype as is. Make your changes on the new prototype. Thank you.
@PhysicsProf, do you mind sharing your private email to me related to the prototype that has reasonable Index but may suddenly fall below 1. On restart, rises to reasonable value again?
Quote from: poynt99 on 2011.02.12, 02:30:37
Lawrence,
Do you perhaps mean something like this?
.99
That's a fascinating one Poynt. It looks like when you lower the supply voltage below a certain threshold you get a different oscillation mode at about 330 KHz, which is very fast. I can see that when the transistor switches off, that some of the energy stored in the core gets pushed back into the supply rail. Did you increase the output impedance of the voltage source also to simulate a nearly dead battery? That's one part of explaining the spikes.
I think can see how it is shutting itself off too soon to run at 330 KHz. As the current increases the 1-ohm resistor raises the potential threshold for the base input to conduct and that chokes off the base current, causing a premature shut-off.
What I can't figure out is how the EMF or current related to the discharge goes back to the voltage source. If L2 pulses high on the "dot" then L1 pulses high on the "dot" also and that's connected to the voltage source. However, there is no reference between ground and the "dot" of L1 to give L1 the ability to kick-back to the supply. The discharge glitch is about 200 nanoseconds long. The only path that I can see would be if the base-emitter junction briefly conducted backwards, and that would let L1 kick back some energy into the supply. Perhaps that's possible because the diffusion layer of the diode has a population of electrons and holes in play and there is a short window of opportunity for reverse conduction? (It's been a long time since I delved into the semiconductor physics for an NPN transistor)
I am really going out on a limb here. It takes some serious analysis work on the bench to figure out the cause of a glitch.
MileHigh
Lawrence,
The basis for your conclusions about that graphic is very weak.
Look at the voltage and current; they are both mostly in the negative area....doesn't that seem strange to you?
How can you have a negative supply voltage???????????????????????????????
It would appear that the scope probes are connected - to + rather than + to -, OR the scope channels are inverted.
Something does not look right at all. At one part of the trace I've drawn a vertical line through, you have a negative voltage x a negative current, and the power trace shows that as negative power. How can that be?
I'm afraid I do not trust those measurements Lawrence. I hope you can agree that there is something quite out of sorts there.
.99
Quote from: MileHigh on 2011.02.12, 04:39:05
That's a fascinating one Poynt. It looks like when you lower the supply voltage below a certain threshold you get a different oscillation mode at about 330 KHz, which is very fast.
Yes, as the supply voltage decreases Fo increases. We touched on this the other day.
Quote
I can see that when the transistor switches off, that some of the energy stored in the core gets pushed back into the supply rail. Did you increase the output impedance of the voltage source also to simulate a nearly dead battery? That's one part of explaining the spikes.
I did increase the impedance of the source, but not as you may think. The voltage measurement also may not be what you think. It is easy to make a simple change that can render the measurements misleading. In this case it looks like there is energy going back to the source, but there isn't. It also may look like the voltage output is spiky, but it isn't.
Quote
I am really going out on a limb here. It takes some serious analysis work on the bench to figure out the cause of a glitch.
MileHigh
Don't risk falling off that limb. What you see is not what is truly happening. I wanted to see if this is what Lawrence was referring to, and also to show how easy it is to become fooled by your measurements if you are not aware of some simple but critical factors.
.99
Got it. Yes I am pretty sure I know what it is then. It's a theme that has come up relatively recently in the past, i.e.; the TK video.
MileHigh
I was thinking the wire between the positive voltage source and the positive power input of the LTJT was doing a little inductive discharge when the transistor switched off. So the positive lead, not the ground lead.
Update:
O0 I will bank the Brownie Points for a future fail. lol
Lawrence,
The power calculations shown on your graph are incorrect.
I don't know if I should even bother, but multiplying a voltage time average with a current time average, whether RMS, Mean, or whatever, is missing the time relationship between voltage and current.
By all means keep researching free energy, it's out there.
EM
MH, O0
Quote from: EMdevices on 2011.02.12, 05:20:29
By all means keep researching free energy, it's out there.
Oh yeah! More than many will ever know! It's literally everywhere!
Quote from: Grumpy on 2011.02.12, 05:38:32
Oh yeah! More than many will ever know! It's literally everywhere!
Aye!!
Often difficult to capture but it is definitely there.
Radiantly...
Quote from: poynt99 on 2011.02.12, 04:45:19
Lawrence,
The basis for your conclusions about that graphic is very weak.
Look at the voltage and current; they are both mostly in the negative area....doesn't that seem strange to you?
How can you have a negative supply voltage???????????????????????????????
It would appear that the scope probes are connected - to + rather than + to -, OR the scope channels are inverted.
Something does not look right at all. At one part of the trace I've drawn a vertical line through, you have a negative voltage x a negative current, and the power trace shows that as negative power. How can that be?
I'm afraid I do not trust those measurements Lawrence. I hope you can agree that there is something quite out of sorts there.
.99
Let us re-examine the scope shots by the PhysicsProf I am reproducing the scope shot on reply 350 from PhysicsProf here.
It has the characteristics I mentioned.
1. The Input Voltage had spikes.
2. The Input Current showed significant large negative area.
3. The Output Voltage is much higher than the Input Voltage
Some of you dismissed it as experimental error.
You will see many more of such errors. Pseudo resonance will have strange effects.
Quote from: poynt99 on 2011.02.12, 04:45:19
Lawrence,
The basis for your conclusions about that graphic is very weak.
Look at the voltage and current; they are both mostly in the negative area....doesn't that seem strange to you?
How can you have a negative supply voltage???????????????????????????????
It would appear that the scope probes are connected - to + rather than + to -, OR the scope channels are inverted.
Something does not look right at all. At one part of the trace I've drawn a vertical line through, you have a negative voltage x a negative current, and the power trace shows that as negative power. How can that be?
I'm afraid I do not trust those measurements Lawrence. I hope you can agree that there is something quite out of sorts there.
.99
Look at the voltage and current; they are both mostly in the negative area....doesn't that seem strange to you?
At the beginning, I thought that it was strange. Then we discussed the possibility of back emf. The particular screen shot was using air cores. Once we accepted that the back emf was giving a significant feedback to the source, we accepted the screen shots as captured. I can show at least a dozen more with other prototypes.
Something does not look right at all. At one part of the trace I've drawn a vertical line through, you have a negative voltage x a negative current, and the power trace shows that as negative power***. How can that be? In the debate, you explained to me that the negative power represents energy fed back to the source. You won the debate based on that. Have you forgotten your own arguments?????
***Edit: Possibly the voltage waveform was inverted.
I'm afraid I do not trust those measurements.At present, you are the only one with two DSOs. The PhysicsProf needs to travel 140 miles to get a reading. I can delay all posting until I have two good DSOs next to me or wait until the Physics Society has its postings. A simple statement that you do not trust someone else hard work and ignore all evidence is very upsetting.
I can understand why the PhysicsProf was upset and temporarily left for awhile.
Or perhaps FuzzyTomCat could get the 4 channel scope again and run some tests on a single machine. At least the DSO he used had parallel data sampling so there is no skewing between sampling. :) I haven't looked at the A to D section in Darren's scopes to see if they are parallel latched inputs or not. Of course even the parallel latched inputs have a margin of error on fast rising or falling signals but I think that is far outside the realm of what you guys are doing. But then that's probably why Tektronix spent so much money marketing the Deskew kit. :D
Negative current flow through an LED . . . Hmmm. Not much of a diode then 'eh?
Lawrence your power traces show Math = RMS, so unfortunately they cannot be useful apart from similarity comparisons between indexes. Trying to compare your work with Darren's is pointless until new mean readings are taken ;)
Darren, your Power trace has a saddle back, so something else is happening there that shouldn't be, both the voltage and current are flat along their slopes.
Another real culprit here is the nonlinear phase angle, inductive on the rise and capacitive on the fall. This indicates a path dependence that raises issues with Kirchhoff's rules, because Kirchhoff does not handle path dependent scenarios. The fact that the current is dropping while the voltage is still present indicates a variable resistance and we are supposed to 'assume' the resistor is not the culprit, be we really don't know for sure. Remember when I suggested "Pure Resistance" for the debate? Nevertheless, you may need to resort to using Faraday's laws to do a proper analysis.
http://www.youtube.com/watch?v=eqjl-qRy71w (http://www.youtube.com/watch?v=eqjl-qRy71w)
http://www.youtube.com/watch?v=1bUWcy8HwpM&feature=related (http://www.youtube.com/watch?v=1bUWcy8HwpM&feature=related)
PhysicsProf - Where waveform is periodic you can also use the formula Amplitude / Sqareroot(3) on your sawtooth wave to double check the RMS value. (see http://en.wikipedia.org/wiki/Root_mean_square#RMS_of_common_waveforms (http://en.wikipedia.org/wiki/Root_mean_square#RMS_of_common_waveforms))
Rosemary, I'm sorry you didn't like the gift - it was sent as a gesture of kindness and nothing more.
To everyone else: the Iraqi Dinar will probably not be revalued anytime soon C.C
8)
Quote from: Harvey on 2011.02.12, 10:15:13
Or perhaps FuzzyTomCat could get the 4 channel scope again and run some tests on a single machine. At least the DSO he used had parallel data sampling so there is no skewing between sampling. :) I haven't looked at the A to D section in Darren's scopes to see if they are parallel latched inputs or not. Of course even the parallel latched inputs have a margin of error on fast rising or falling signals but I think that is far outside the realm of what you guys are doing. But then that's probably why Tektronix spent so much money marketing the Deskew kit. :D
My TDS784A uses multiplexed sampling at 4G/s. Deskew is used for matching probe delays. The scope inputs should have negligible skew between channels, and would not be a significant factor here. I was using 2 channels only, so there was 2G/s available for each channel; hardly pushing the envelope at all. I suspect that the TDS3012B's (they are an "older" DPO scope) I'm using also have multiplexed sampling.
Quote
Negative current flow through an LED . . . Hmmm. Not much of a diode then 'eh?
Indeed. Where a diode is required to resist reverse bias of more than about 5V or so, it seems an LED is a poor choice. Germaniums are quite leaky as far as diodes go too, but they are still considered "diodes". ;)
Quote
Lawrence your power traces show Math = RMS, so unfortunately they cannot be useful apart from similarity comparisons between indexes. Trying to compare your work with Darren's is pointless until new mean readings are taken ;)
This and all other scope shots from Lawrence and the professor were taken before the debate.
Quote
Darren, your Power trace has a saddle back, so something else is happening there that shouldn't be, both the voltage and current are flat along their slopes.
Harvey, could you please specify which scope shot you are referring to?
.99
Quote from: ltseung888 on 2011.02.12, 08:49:21
Let us re-examine the scope shots by the PhysicsProf
I am reproducing the scope shot on reply 350 from PhysicsProf here.
It has the characteristics I mentioned.
1. The Input Voltage had spikes.
2. The Input Current showed significant large negative area.
3. The Output Voltage is much higher than the Input Voltage
Some of you dismissed it as experimental error.
You will see many more of such errors. Pseudo resonance will have strange effects.
Although I have explained the "problem" here:
http://www.overunityresearch.com/index.php?topic=538.msg9850#msg9850
and here:
http://www.overunityresearch.com/index.php?topic=538.msg10307#msg10307
Let us review again why these particular scope shots are "suspect".
As it is not a perfect world, every digital oscilloscope has a small amount of DC offset at its input. Tektronix (and most other good brands) provide a means to "compensate" (i.e. zero-out) these offsets in the scope channels. This is a routine accessible within the menus which should be run after scope warmup of 20 minutes or so, and any time the ambient changes more than 5ºC or so since the last "calibration". An offset of 4mV is not uncommon, and I have seen this also with the TDS3012B scopes I am using for these tests.
When the signals being measured are a number of magnitudes larger than the residual offset in the scope channel, the offset will most likely not be a significant factor in the measurement. However, when the signal being measured is the same magnitude as the residual channel offset, you can see why the measurement will be "offset" by up to 100% of the scope reading.
In the case of that scope shot, there is quite clearly a channel offset of about -4mV in both channels, and since the signal amplitude is on the order of 4mVpp, the resulting p(t) product is also significantly offset in the negative direction.
.99
Quote from: ltseung888 on 2011.02.12, 09:06:34
In the debate, you explained to me that the negative power represents energy fed back to the source. You won the debate based on that. Have you forgotten your own arguments?????
Of course I have not.
Quote
***Edit: Possibly the voltage waveform was inverted.
Indeed, and I trust you can see how this would completely skew the entire measurement.
Quote
I'm afraid I do not trust those measurements.
I can not trust those measurements.
Quote
At present, you are the only one with two DSOs. The PhysicsProf needs to travel 140 miles to get a reading. I can delay all posting until I have two good DSOs next to me or wait until the Physics Society has its postings. A simple statement that you do not trust someone else hard work and ignore all evidence is very upsetting.
I can understand why the PhysicsProf was upset and temporarily left for awhile.
Sorry that you are becoming upset Lawrence.
As best I can, I am stating the facts and I am being scientific. This has always been our agreement, yes?
I have tested your assembled unit, and other than crudely repairing a loose solder joint and soldering on some battery leads, I made no modifications to your unit. I essentially tested it right out of the box.
The tests revealed that the unit was clearly underunity, and I believe I have not made any errors with my tests. You stated that this unit would perform and measure COP>>1, right out of the box with no tuning, tweaking, or modifications, yet it has only proven to be COP<<1.
May I ask you at this time to address this outcome?Regards,
.99
Quote from: MileHigh on 2011.02.12, 04:39:05
What I can't figure out is how the EMF or current related to the discharge goes back to the voltage source. If L2 pulses high on the "dot" then L1 pulses high on the "dot" also and that's connected to the voltage source. However, there is no reference between ground and the "dot" of L1 to give L1 the ability to kick-back to the supply. The discharge glitch is about 200 nanoseconds long. The only path that I can see would be if the base-emitter junction briefly conducted backwards, and that would let L1 kick back some energy into the supply. Perhaps that's possible because the diffusion layer of the diode has a population of electrons and holes in play and there is a short window of opportunity for reverse conduction? (It's been a long time since I delved into the semiconductor physics for an NPN transistor)
I am really going out on a limb here. It takes some serious analysis work on the bench to figure out the cause of a glitch.
MileHigh
That's what I was trying to explain before. The transistors diode can reverse at certain frequencies and with the right available potentials. They are usually mentioned in the specs I think as "reversing threshold".
@poynt99
The main problem here that I can see in these exercises is what I know from pulsing many coils and many frequencies and ranges to get resonances and/or maximum outputs. The right or the best resonance frequency is usually a thin band of remarkable results and almost nothing when out of the range. So here we have fixed a transistor, fixed LEDS, fixed resistors, everything is all prefixed. But is it? The coil wind could have one more turn and the fixed values could fall out of range. You have no way of varying the frequency. So.... is there a way to add a small POT to try and adjust or pin down the right frequency in the known range of the fixed components?
Otherwise, if only the coil was on your bench and fed by your frequency generator at 1.5vdc, you will quickly find the best resonance frequency by scoping the output and then you could simply calculated the required resistors and possibly the right POT value to obtain that same frequency range in an isolated device. I think this is why this is not working the way it is supposed to.
If you cannot make any additions to this device, then just forget it. You could be 5kz away from tremendous results and you will never know it and just waste time pushing a dead horse.
wattsup
In order to help solidify the important issue of the impact scope channel offsets can have on low signal level measurements, please see the attached scope shots.
In both cases, the probe tip is connected to the probe ground lead, shorting out the probe. The channel vertical gains are set to maximum of 10mV per division. No signal (AC or DC) should be present. What IS seen on the traces is residual AC internal circuit noise (the hash), and any DC offset present in the signal chain of each channel.
Scope 1 is indicating a DC offset of -1.25mV in CH1, and -3.11mV in CH2.
Scope 2 is indicating a DC offset of +1.04mV in CH1, and -2.99mV in CH2.
I will show the results after the scopes warm up and after running the SPC on each.
.99
From page "Glossary-8" of the TDS3000 series manual, a blurb on SPC (emphasis added):
QuoteSignal Path Compensation (SPC)
The ability of the oscilloscope to minimize the electrical offsets
in the vertical, horizontal, and trigger amplifiers caused by
ambient temperature changes and component aging. You should
run SPC when the ambient temperature varies more than 5º C
from the last SPC or before performing critical measurements.
and this from page 3-77:
QuoteSignal Path Compensation. For maximum accuracy at any time, run the
signal path compensation routine just before taking critical
measurements. To meet accuracy specifications, run the routine if
the ambient temperature changes by 10° C or more.
.99
Quote from: wattsup on 2011.02.12, 16:15:27
That's what I was trying to explain before. The transistors diode can reverse at certain frequencies and with the right available potentials. They are usually mentioned in the specs I think as "reversing threshold".
watts, I challenge you to show me any transistor data sheet that shows a "reversing threshold" and especially one that occurs "at certain frequencies". If you are talking about the base-emitter junction "diode", it is likely that any circuit which repeatedly reverse-biased (base far more negative than emitter for an NPN) that diode until it conducted would just break the part. Game over. Your theory is wrong.
See the very last sentence here: http://en.wikipedia.org/wiki/Bipolar_junction_transistor#Vulnerabilities (http://en.wikipedia.org/wiki/Bipolar_junction_transistor#Vulnerabilities)
Reverse breakdown of the B-E junction on a repeated basis just destroys the device and it is not a "certain frequencies" phenomena at all.
Quote
@poynt99
The main problem here that I can see in these exercises is what I know from pulsing many coils and many frequencies and ranges to get resonances and/or maximum outputs. The right or the best resonance frequency is usually a thin band of remarkable results and almost nothing when out of the range. So here we have fixed a transistor, fixed LEDS, fixed resistors, everything is all prefixed. But is it? The coil wind could have one more turn and the fixed values could fall out of range. You have no way of varying the frequency. So.... is there a way to add a small POT to try and adjust or pin down the right frequency in the known range of the fixed components?
Otherwise, if only the coil was on your bench and fed by your frequency generator at 1.5vdc, you will quickly find the best resonance frequency by scoping the output and then you could simply calculated the required resistors and possibly the right POT value to obtain that same frequency range in an isolated device. I think this is why this is not working the way it is supposed to.
If you cannot make any additions to this device, then just forget it. You could be 5kz away from tremendous results and you will never know it and just waste time pushing a dead horse.
wattsup
First of all, great pains were taken to assure that the circuitry Poynt is testing is exactly the very same one Lawrence tested. Secondly, I challenge your assertion that varying the frequency of oscillation by a few "kz" would or could, in this circuit, reveal some magic "resonance" that would significantly change the COP.
It's just more silliness. There are no hi-Q resonant circuits involved here. Period. Sorry...theories rejected.
Humbugger
Quote from: poynt99 on 2011.02.12, 04:45:19
It would appear that the scope probes are connected - to + rather than + to -, OR the scope channels are inverted.
.99
Once again you have impressed me with your knowledge and experience. ;)
Thanks Gibbs. :)
Here are the scope shots after 1 Hour and after running the SPC.
| | Initial Offset | | After 1 Hour | After SPC |
| Scope 1 CH1 | | -1.25mV | | -1.61mV | 0.294mV |
| Scope 1 CH2 | | -3.11mV | | -3.23mV | 0.839mV |
| Scope 2 CH1 | | +1.04mV | | +1.25mV | -0.396mV |
| Scope 2 CH2 | | -2.99mV | | -2.45mV | -0.064mV |
A pretty good improvement over all. On better scopes, such as my TDS784A, there is a calibration for the probes as well, and that eliminates the remaining offset seen in the far right column. The 0.839mV is still a little high. One can run the SPC again to see if there might be more improvement.
.99
Quote from: wattsup on 2011.02.12, 16:15:27
That's what I was trying to explain before. The transistors diode can reverse at certain frequencies and with the right available potentials. They are usually mentioned in the specs I think as "reversing threshold".
@poynt99
The main problem here that I can see in these exercises is what I know from pulsing many coils and many frequencies and ranges to get resonances and/or maximum outputs. The right or the best resonance frequency is usually a thin band of remarkable results and almost nothing when out of the range. So here we have fixed a transistor, fixed LEDS, fixed resistors, everything is all prefixed. But is it? The coil wind could have one more turn and the fixed values could fall out of range. You have no way of varying the frequency. So.... is there a way to add a small POT to try and adjust or pin down the right frequency in the known range of the fixed components?
Otherwise, if only the coil was on your bench and fed by your frequency generator at 1.5vdc, you will quickly find the best resonance frequency by scoping the output and then you could simply calculated the required resistors and possibly the right POT value to obtain that same frequency range in an isolated device. I think this is why this is not working the way it is supposed to.
If you cannot make any additions to this device, then just forget it. You could be 5kz away from tremendous results and you will never know it and just waste time pushing a dead horse.
wattsup
This was all a whackadoo theory om my part that was wrong. I was too lazy and tired to go look-up if this was even possible and was drawing on my vague memory about the properties of transistor junctions from 30 years ago. My confidence in what I was saying was quite low.
A few postings later Poynt hinted at the true cause for the positive glitch on the voltage reading and I got it right away. It was due to the inductance in the length of wire between the power source and the JT and the fact that the probe for reading the voltage was on the JT power input when it should have been placed on the power source output itself, i.e.; the battery terminal for example. When the JT transistor shuts off the inductance in the wire discharges it's stored energy and makes the tiny voltage spike that's seen by the probe placed on the JT power input.
Like Humbugger says, there is no resonance point for the JT circuit. When you power it up, it's runs at it's natural resonance frequency. There is no need to inject a signal generator into it and hunt for a resonance frequency and then try to get it to operate at that frequency. The frequency it runs at is its natural resonance frequency.
It's worth it to state again that "resonance" is being used very loosely here because it's not true resonance. It's a pulse circuit that has a charging time and a discharging time. The charging time and the discharging time are variable and are not necessarily directly related to each other. When you add the charging time to the discharging time you get the period, then you invert that to get the operating frequency.
It's much more accurate a term to say "operating frequency" as opposed to "resonant frequency" because nothing in the JT circuit is resonating. By extension, when Lawrence says, "pseudo resonance" it's a meaningless term and it should be avoided. The last thing you want to do when you talk about electronics is use meaningless terms. It does not add to the debate, it just creates confusion.
MileHigh
A bit of good news:
I've just tried running the SPC on both scopes with the scope probes attached (but shorted), and the results appear to be quite good. Better at least than the previous run with the scope channels "open" as they suggest you do.
All offsets are no worse than about 0.200mV. O0
.99
Quote from: wattsup on 2011.02.12, 16:15:27
...
The main problem here that I can see in these exercises is what I know from pulsing many coils and many frequencies and ranges to get resonances and/or maximum outputs. The right or the best resonance frequency is usually a thin band of remarkable results and almost nothing when out of the range. So here we have fixed a transistor, fixed LEDS, fixed resistors, everything is all prefixed. But is it? The coil wind could have one more turn and the fixed values could fall out of range. You have no way of varying the frequency. So.... is there a way to add a small POT to try and adjust or pin down the right frequency in the known range of the fixed components?
...
If you cannot make any additions to this device, then just forget it. You could be 5kz away from tremendous results and you will never know it and just waste time pushing a dead horse.
wattsup
Quote from: humbugger on 2011.02.12, 17:05:19
...
First of all, great pains were taken to assure that the circuitry Poynt is testing is exactly the very same one Lawrence tested. Secondly, I challenge your assertion that varying the frequency of oscillation by a few "kz" would or could, in this circuit, reveal some magic "resonance" that would significantly change the COP.
It's just more silliness. There are no hi-Q resonant circuits involved here. Period. Sorry...theories rejected.
Humbugger
Any who are expecting instantaneous results will be
sorely disappointed. The mode of operation which
accounts for the mysterious increase in output is
one which must be "tickled" from the circuit under
test. It can be a painstaking procedure.
There are many variables which are capable of
contributing to enhanced performance and/or
detracting from circuit performance.
The methods of "tuning" and "tweaking" will seem
exceedingly unorthodox to uninitiated observers.
Finding the "zone" can be a time consuming and
rather frustrating ordeal. Then, once found, trying
to determine why it occurs can be a daunting task.
A scope is a necessity. Paying attention to small
changes in waveforms as "tuning" is performed
is essential. Eventually a "signature" will be
discovered.
Most will give up before success is realized.
Extreme patience and perseverance are key.
You will not at first believe what you are seeing
when it happens.
You will not find many who will believe what you've
experienced, should you choose to disclose.
Silence can be golden.
Quote from: poynt99 on 2011.02.12, 20:09:22
A bit of good news:
I've just tried running the SPC on both scopes with the scope probes attached (but shorted), and the results appear to be quite good. Better at least than the previous run with the scope channels "open" as they suggest you do.
All offsets are no worse than about 0.200mV. O0
.99
I've used a lot of those older TDS500-700 series scopes and bought/sold tons of them on eBay. Onre thing I lerarned right away is that almost every one of them, if stored away unused for a while, develop Aquisition errors that show up as FAIL on the self-test on startup but still appear to work quite well.
In most cases, a good cleaning of the aquisition front end PCB with methyl alcohol will fix the problem. I think if the SPC can't quite compensate it, the self-test fails. I used to buy them with Aquisition FAILs real cheap, clean the board well and suddeenly have a scope that passed all self-tests and came right on home well under 0.1mV after SPC.
The scary thing about them is that the CRT and frame-sequential LCD color filter they use are getting real hard to find anymore without paying huge bucks. They are great scopes but I'm scared of them anymore as far as investments go. I guess you can buy LCD TFT display retrofit kits for them but those cost a fortune, too.
Humbugger
Quote from: Dumped on 2011.02.12, 20:31:04
Any who are expecting instantaneous results will be
sorely disappointed. The mode of operation which
accounts for the mysterious increase in output is
one which must be "tickled" from the circuit under
test. It can be a painstaking procedure.
There are many variables which are capable of
contributing to enhanced performance and/or
detracting from circuit performance.
The methods of "tuning" and "tweaking" will seem
exceedingly unorthodox to uninitiated observers.
Finding the "zone" can be a time consuming and
rather frustrating ordeal. Then, once found, trying
to determine why it occurs can be a daunting task.
A scope is a necessity. Paying attention to small
changes in waveforms as "tuning" is performed
is essential. Eventually a "signature" will be
discovered.
Most will give up before success is realized.
Extreme patience and perseverance are key.
You will not at first believe what you are seeing
when it happens.
You will not find many who will believe what you've
experienced, should you choose to disclose.
Silence can be golden.
I think you may have missed the point here that Poynt received the already-tweaked exact same unit that Lawrence had gotten his results with. So what does that leave as variables left to tweak? Just the test setup and the understanding of how to interpret the results and use correct math.
Or maybe the aether in Hong Kong is thicker than in Canada. Unless you are just babbling, please tell us exactly what variables there are here to tweak? I'd be interested to know.
I get the distinct impression that if the scope traces were exactly identical, Lawrence would find a way to interpret them as COP>>>>>1 while Poynt found COP<<1 and that, rather than using logic and experience to analyze why the discrepancy existed, the believers would insist that "more tweaking" was required...ad infinitum, ad nauseum.
Humbugger
Quote from: humbugger on 2011.02.12, 20:38:42
...
Or maybe the aether in Hong Kong is thicker than in Canada. Unless you are just babbling, please tell us exactly what variables there are here to tweak? I'd be interested to know.
...
Humbugger
Lawrence has already provided many clues.
It's rather like building a super sensitive
regenerative receiver for high frequency
operation.
Construction techniques which are thought
to be counter productive can yield extraordinary
results.
Or the fringe area TV antenna. Often very unusual
dimensions and shapes produce incredible
reception.
Compound resonances or interference patterns
are capable of mysterious phenomena - even on
the small and low powered scale.
Outside the box.
Dumped, Humbugger, Wattsup, Lawrence:
I suppose that we are in the classic "secret sauce" debate here. It's the belief that tweaking some parameter or parameters will unleash (pick your favourite word) something special and the output-to-input efficiency will go way up, and possibly achieve over unity.
The other side of the debate says let's look at this circuit and understand how it works. Once you understand how it works then you will understand how parameter changes affect the circuit and you will understand that there is no "secret sauce" to be found.
Besides the "secret sauce" debate, what is a bit of a pet peeve for me is how people almost never discuss the losses in the circuits they test. For the JT there are losses in the interconnect wires and the coils due to resistance. There are losses in the core due to hysteresis. There are losses due to coupling inefficiencies. Ultimately, that's why when you look at the numbers for the LTJT you get an efficiency of about 30%. That due to the losses and the fact that a lot of power is being dissipated in the collector LED. That's reality staring you in the face where your measurements are in line with the expected losses. You can also measure the losses individually and all the numbers will add up.
Another pet peeve is that if you believe that there is over unity in your electronic circuit, why not try to propose an explanation for it? That's never done. And just saying it comes from the "aether" or it is "vacuum energy" is a cop out. If you can explain the loss mechanisms, why can't you explain the alleged over unity mechanism? Just as importantly if you think that you have over unity, then show exactly where and when it occurs in your circuit. Just blindly looking at the output is a cop out one more time.
Finally, another big factor at play is the phenomenon of misinterpreting what you are seeing on the bench because you don't have the proper education, training, and experience to understand what you are looking at. You see it in YouTube clips all the time. So if you sweep a signal through your circuit or change a capacitor and all off a sudden see a jump in a voltage somewhere and get all excited and think that you have discovered something. The truth is that it is almost metaphysically certain that you have discovered nothing. The only way to get over that hump is through education, training, and experience. However, experience without the education and training and with a pre-disposed bias to be looking for phenomena that are "outside of the ordinary" and phenomena that are allegedly over unity is a recipe for minor disaster. You see this dynamic all the time. For example, look at the "captret" threads on the other forums that are just starting to die down. The whole thing is nothing, a tempest in a teapot.
I suppose that was a bit of a rant too but what the hell!?
MileHigh
I think Dumped is trying to intimate that he has one of these units on his bench exhibiting COP>1.
Looking forward to seeing your build and the awesome scope measurements of MEAN PI and PO Dumped. O0
You're welcome.
.99
Notice how I have been nipped by Occam's Razor twice? The voltage spike was simply due to the length of the wires powering the JT and the positioning of the probe. The offset in the DSO was not because of some esoteric problem with the internal guts, it was just that the coupling was AC instead of DC.
No doubt Occam's Razor is lurking just around the corner with respect to a lot of "miraculous" phenomena that experimenters might observe on the bench.
Here again are the measurements of the assembled unit Lawrence sent me. Corrections to the OUTPUT scope and probes have been made, so it is only proper to do the measurements again.
assembled_input_mean_2nd.PNG indicates an average INPUT power of 64.37mW.
assembled_output_mean_2nd.PNG indicates an average OUTPUT power of 236.6mW/10 = 23.7mW.
The results are very similar to the original test performed here (http://www.overunityresearch.com/index.php?topic=723.msg10637#msg10637).
.99
Looks great. You can see in the screen capture "assembled_output_viewing_2nd.PNG" exactly when the core has completely been exhausted itself of stored magnetic energy. The output voltage and current are on two linear downward slopes during the energy discharge. When the current hits zero there is no more energy left in the core and the resulting voltage goes into "free fall" until it hits zero.
For Lawrence: These are the types of scope waveform captures that I was expecting you to present to us. I knew ahead of time what the waveforms should look like and Poynt has presented us with data that does conform with the expected waveforms. In contrast, the first set of scope captures that you showed us several weeks ago were filled with noise and simply did not make any sense. You would be wise to refrain from making any more references to those waveforms because the data is invalid. The best course of action for your team in Hong Kong would be to do a new round of testing. Also, your "FLEET Comparison Index" is irrelevant data and will be ignored. The only data that is relevant is the average power out and the average power in. Now that Poynt has real samples to work with the analysis will be done using standard scientific measurement techniques.
MileHigh
Still enjoying variations on the LTJT circuit. It occurred to me that the output shows alternating currently clearly, and so I put TWO LED's on the output circuit, in parallel, biases reversed. As I expected, they both light up -- using red LED's. In fact, subjectively I can say that the two LED's glow roughly the same.
Then, using my ATTEN 60MHz DSO to look at the INPUT power (using the math function *), I find that the Pin waveform goes from saw-tooth to approximately square wave (with two LED's in the output circuit). Calculating the input energy as I described before, the input energy (and power) approximately doubles.
When I remove the added LED from the breadboard-circuit (easy to do), I get the sawtooth on the input Power we are familiar with. When I remove the OTHER LED, the pattern is quite different, more of a skewed rectangular tooth. (I still don't have the ability to bring the waveforms here to the forum from this new DSO, sorry.)
Note: one must exercise caution when installing the output-circuit-LED if one uses only one. Clearly for a white LED, one direction is much brighter than the other bias-direction.
It's all very interesting and instructive. I'm using a toroidal inductor I purchased from Jameco, 100 uH with 34 windings I counted, this for the secondary, then I added thirteen (paired) windings for the JT-windings, wound by hand.
Yes, I have done a rough calculation Ein and Eout from the Power functions, in and out, integrating crudely over time (by hand). This DSO displays the Power waveform but does not calculate the average of it for you, so I do it by hand. A bit tedious, but I get a close look at the waveforms this way. I also double check the frequency. The best COP I've obtained for this particular set-up (Eout/Ein for one cycle) is about 0.85. Of course, I removed the transistor-LED out of the circuit long ago.
MileHigh -- would you expect the frequency to go up or down when I add a second LED to the output-circuit as described?
Thanks for your detailed measurements 99. Well, Lawrence's trying to operate at a non ideal 600mV is surely asking for trouble where noise levels around 200mV. Then, trying to equate wild Vpp swings to real power is just silly. I hope this puts this nonsense of 'lead-in' or 'bring-in' energy crap to rest. But thank you for spending the time. TK already showed what bad understanding and bad measurements can lead to and EMD's comments are worth listening to.
chrisC
Quote from: PhysicsProf on 2011.02.13, 23:00:08
MileHigh -- would you expect the frequency to go up or down when I add a second LED to the output-circuit as described?
If you add a second LED to the output circuit it means that when the transistor is switched on the battery drives two loads so the average impedance is lower. The loads are the L2-core and the second LED. I am assuming this increased load will result in the current going into L2 leveling off sooner and switching off the transistor sooner. So that would make the operating frequency of the Joule Thief increase.
That's my reasoning but I could be missing something here.
MileHigh
Quote from: MileHigh on 2011.02.14, 02:10:22
If you add a second LED to the output circuit it means that when the transistor is switched on the battery drives two loads so the average impedance is lower. The loads are the L2-core and the second LED. I am assuming this increased load will result in the current going into L2 leveling off sooner and switching off the transistor sooner. So that would make the operating frequency of the Joule Thief increase.
That's my reasoning but I could be missing something here.
MileHigh
Nope. In fact, I observe the operating frequency to decrease when the second LED is added.
In the case I describe above (where the toroid I started with was from Jameco), the frequency decreased from 266 down to 254 KHz when I added the second LED.
Similarly, with another toroid, the operating frequency dropped from 417 KHz down to 334 KHz when I added the second LED as described.
Again, this little LT-JT device surprises us -- it surprised me, too.
Now, about two weeks ago (before the debate, when I was still taking RMS measurements), I wrote the following:
QuoteI modified the JT circuit by removing the LED (across the transistor), so as to lower the input power requirement. The JT still rings as seen with the DSO.
I also worked with and prepared the yellow toroid Lawrence sent me -- found that this one now has the following characteristics:
Just 8uH on each of the JT windings, and 0.023 ohms each.
35uH on the transformer winding and 0.054 ohms.
From there, we set the device up on a breadboard with a one ohm resistor + green LED on the output circuit.
This LT-JT behaves in an interesting fashion, in that it sometimes oscillates at an unusually high frequency, 571KHz, with lots of spikes observed. The LT FLEET Index is seen at approximately seven, which was very encouraging to us.
But sometimes, it does not hit the resonance, I think, for then the Index is less than one. Then I just disconnect the battery and re-start the device and then it will run at high Index for minutes without problem, until I disconnect the battery.
I can hardly wait for the good oscilloscope to arrive.
Now the new DSO has arrived, and it permits me to take a product and display the V*I waveform. This is very helpful.
I NO LONGER calculate the "FLEET Index" which was based on an erroneous approach. Rather, I used the V*I waveform, and soon I expect to have a DSO which will calculate the mean values from the Power waveform, so that I no longer need to these calculations (with estimates based on observing the waveforms closely) by hand.So I wish to clarify that the statement above, "The LT FLEET Index is seen at approximately seven" was based on erroneous calculations, yet the odd behavior of the device remains curious. That is why I am seeking the best measurement methods possible, including in particular an evaluation of MEAN (or average) Pin and average Pout on a device,
and why I intend to keep "tweaking" the device.I don't think we fully understand the LT-JT or variations, and here I would include the "Joule Ringer" discussed on other threads and forums.
I hope that we can continue the measurements -- and here I thank .99 for his measurements -- and proceed with
RIGOR WITHOUT RANCOR. That is what I seek.
Thanks Professor.
Looking forward to your upcoming measurements. :)
I am surprised at the low inductance of those two coils, especially since they are on a core. The air-core coils I made for the P9901 have an inductance of about 3uH each. I suspect those cores are iron-powder with the inherent distributed gapping. I have yet to measure the AL of the 3 cores I have, but I will. ;)
Also, I wonder if your JT is perhaps jumping into the spurious mode of operation ION talked about here (http://www.overunityresearch.com/index.php?topic=602.0)? (highly recommended reading btw, thanks ION).
Also thank you to Humbugger for his excellent work here (http://www.overunityresearch.com/index.php?topic=580.msg9211#msg9211) on alternative testing methods for these switching circuits.
Sounds like we're all on the same page Professor.... O0
.99
PhysicsProf:
I'm not surprised that I was wrong. I am getting more and more tempted to run pSpice on my computer. Of course having a real setup helps tremendously also!
If you really wanted to explore this further, you can easily separate the problem into two parts. The two parts would essentially represent pulling the Joule Thief apart into two separate test circuits that are not actual Joule Thief circuits anymore.
The first part would be to investigate how much energy the core can store with respect to the amount of current you put through the coil of wire you have wrapped around the core. If you plotted stored energy vs. coil current you would clearly see the slope of the curve level off as you approached core saturation.
The second part would be to investigate the power/energy and timing of the discharge of the stored core energy for varying loads with respect to a fixed reference amount of energy stored in the core.
When you deconstruct a Joule Thief and forget about the timing mechanism for a moment, that's all that it really is. I am repeating myself but it's worth repeating: It's just a device that stores energy in a toroidal ferrite core and then discharges that stored energy into a load. So running the two separate tests described above would allow you to investigate those two processes independent of the timing mechanism.
MileHigh
Quote from: PhysicsProf on 2011.02.14, 04:15:15
This LT-JT behaves in an interesting fashion, in that it sometimes oscillates at an unusually high frequency, 571KHz, with lots of spikes observed. The LT FLEET Index is seen at approximately seven, which was very encouraging to us.
But sometimes, it does not hit the resonance, I think, for then the Index is less than one. Then I just disconnect the battery and re-start the device and then it will run at high Index for minutes without problem, until I disconnect the battery.
...the odd behavior of the device remains curious. That is why I am seeking the best measurement methods possible, including in particular an evaluation of MEAN (or average) Pin and average Pout on a device, and why I intend to keep "tweaking" the device.
I don't think we fully understand the LT-JT or variations, and here I would include the "Joule Ringer" discussed on other threads and forums.
I hope that we can continue the measurements -- and here I thank .99 for his measurements -- and proceed with
RIGOR WITHOUT RANCOR. That is what I seek.
The behavior of the device changing the apparent index and waveform for no apparent change or explanation was
observed many times in Hong Kong. We attributed it to shifting of the pseudo resonance condition.
The prototype shipped to Poynt99 was damaged in shipping and he had to re-solder one of the joints. That could easily "shift the frequency" away from the pseudo resonance condition.
I am glad that both PhysicsProf and Poynt99 are continuing the experiments in a scientific way. Sooner or later, they will hit on the pseudo resonance condition again.
The best chance of hitting on the pseudo resonance condition is to use two DSOs and have the LT-JT on a breadboard. Then play with the various changes – different hole positions, different values of electronic components, different power sources, different wire lengths, etc. The most obvious hint is the moving of the hand close to the LT-JT and there are significant changes in waveform (under some configurations). That would be a good hint that a pseudo resonance condition is near.
The TK experiment is instructive – long leads contributing to changes in capacitance and inductance. In our case, we use such changes to effectively change the resonance condition. It is difficult to keep changing inductors or capacitors. But it is relatively easy to move to different holes on the breadboard and to tweak the leads.
God is again bringing us closer to the truth. May He continue to guide us along the right path! Amen.
Standardize a FLEET prototype???
Many have asked me the question – why are your FLEET prototypes have different turns? Would it be simpler to have one fixed configuration – such as 13 turns JT winding and 17 turns transformer winding?
At present, every FLEET prototype was produced by hand. When I used 13 turns JT and 17 turns transformer winding, I did not always get Index greater than 1. I had to change the number of turns to get the Index greater than 1. I did not always succeed.
I accept that I have not mastered the science of pseudo resonance tuning. I do not know the exact parameters and their relationships yet. Much more research needs to be done.
I view the Joule Ringer as much more advanced than FLEET. It has successfully demonstrated that the battery can be removed and the LEDs and CFLs can still be lighted via capacitors for many minutes. The number of LEDs and CFLs it can light is much more than FLEET.
If I were to give an honest opinion to a new comer and comment on which device to replicate, the answer is the Joule Ringer. Both devices need to "hunt for the sweet spot". More researchers are working on the Joule Ringer and providing more information.
If the new comer wants to leapfrog, I would advise him to go for large air toroids. The research hopefully will lead to a Steven Mark type TPU. I believe in some type of resonance effect to bring-in the electron motion energy. Exactly how that can be achieved is still an open question and a challenging area to researchers.
God helps those who help themselves. Those who never hunt for resonance will never find resonance. Amen.
I'm glad you're back on your feet mr. Tseung. You have faith and that's all it matters. The obvious is that the JT surprises everyone showing maybe we don't fully understand the process of induction. Keep up this paced and we'll figure it out soon enough.
I like the joule ringer because it might seems to have a high COP, but I like your joule thief because I believe there is something better ;D
Another debate topic – anyone interested?
While Poynt99, PhysicsProf and others are working on the experimental side, some of us without the apparatus can discuss the theoretical side.
I would like a debate or discussion on the following topic:
If an invention uses X units of supplied energy and it can lead-out or bring-in Y units of existing energy from the surrounding, the effective total input energy to the system is X+Y units. If there were no loss, X+Y units will appear in the output.
If the invention can loopback X units of energy back to input, that X unit can again lead-out or bring-in Y units of energy from the surrounding, the invention can keep itself running and have Y units of energy continuously to use.
Such a machine is NOT the impossible perpetual motion machine. It does not violate any Laws of Physics.
Any one interested??? O0 8) :)
God showed us the Truth. We should thoroughly understand and apply it. Amen.
Lawrence, I have a couple questions:
1) Who tested the assembled unit prior to you sending it to me?
2) Do you have test results and scope shots you can post of it? (if you have already, feel free to post it again here).
.99
Quote from: ltseung888 on 2011.02.14, 19:00:07
Another debate topic – anyone interested?
While Poynt99, PhysicsProf and others are working on the experimental side, some of us without the apparatus can discuss the theoretical side.
I would like a debate or discussion on the following topic:
If an invention uses X units of supplied energy and it can lead-out or bring-in Y units of existing energy from the surrounding, the effective total input energy to the system is X+Y units. If there were no loss, X+Y units will appear in the output.
If the invention can loopback X units of energy back to input, that X unit can again lead-out or bring-in Y units of energy from the surrounding, the invention can keep itself running and have Y units of energy continuously to use.
Such a machine is NOT the impossible perpetual motion machine. It does not violate any Laws of Physics.
Any one interested??? O0 8) :)
God showed us the Truth. We should thoroughly understand and apply it. Amen.
For this to work, you have to "lead-out or bring-in Y units of existing energy from the surrounding".
How do you propose to do this?
Quote from: poynt99 on 2011.02.14, 19:06:12
Lawrence, I have a couple questions:
1) Who tested the assembled unit prior to you sending it to me?
2) Do you have test results and scope shots you can post of it? (if you have already, feel free to post it again here).
.99
The assembled unit was tested in Hong Kong. I do not have Oscilloscopes here to check before sending. I did not bring the test results from Hong Kong - on the false assumption that it is easy to find two good oscilloscopes in Irvine. I believe the archived CD may still be available. I shall try to dig it out when I get back to Hong kong next time. :'(
Quote from: Grumpy on 2011.02.14, 21:13:39
For this to work, you have to "lead-out or bring-in Y units of existing energy from the surrounding".
How do you propose to do this?
That will be the substance of the debate :) :).
You can go to my bench ltseung888 and check out the locked teaching thread for some preliminary information.
We can discuss the details in the debate. And learn together with the World. Interested? ^-^
Quote from: ltseung888 on 2011.02.14, 22:26:13
The assembled unit was tested in Hong Kong. I do not have Oscilloscopes here to check before sending. I did not bring the test results from Hong Kong - on the false assumption that it is easy to find two good oscilloscopes in Irvine. I believe the archived CD may still be available. I shall try to dig it out when I get back to Hong kong next time. :'(
OK, Lawrence, that is the
where. Can you tell us
who tested this particular unit?
Thanks,
.99
Quote from: poynt99 on 2011.02.14, 22:55:10
OK, Lawrence, that is the where. Can you tell us who tested this particular unit?
Thanks,
.99
I believe this particular unit was tested by myself and Mr. Aaron Quant in Hong Kong.
Quote from: ltseung888 on 2011.02.14, 23:26:57
I believe this particular unit was tested by myself and Mr. Aaron Quant in Hong Kong.
OK Lawrence,
I am going to return the assembled unit to you so you can test it again, and post your results. Or, I can send it to Aaron in China.
You do not require two identical oscilloscopes to perform the COP testing or tuning.
.99
Quote from: poynt99 on 2011.02.14, 23:50:15
OK Lawrence,
I am going to return the assembled unit to you so you can test it again, and post your results. Or, I can send it to Aaron in China.
You do not require two identical oscilloscopes to perform the COP testing or tuning.
.99
Wait until I provide the return information. I shall send you PM. Thank you.
Lawrence:
I remember you claimed that the fully assembled unit that you sent to Poynt was an over unity device. It was tested and it was found to be an under unity device.
You made reference to the cold solder joint possibly having something to do with this discrepancy and I don't think that's possible. In addition, "pseudo resonance" is not a valid concept. If you believe it is a valid concept then you will have to demonstrate that with scope shots and proper average power-out vs. average power-in data. I will also remind you again that your "FLEET Comparison Index" concept is invalid.
I apologize if I am sounding tough, but this is a true reflection on where we stand with your Joule Thief samples.
Can you come up with an action plan to convince us that your Joule Thief is an over unity device? My suggestion is that you and/or Aaron Quant make measurements on a device and post photographs, scope shots, and submit all of the relevant data to back up your claim, and that even includes what the power source is. This would have to include data that shows the alleged "pseudo resonance" mode because you are clearly indicating that you only get over unity when the Joule Thief is operating in "pseudo resonance" mode. If you want to make your case even more convincing then you should also document the ordinary operating mode for the Joule Thief with full data. You would also have to clearly and unambiguously state how you got the device to switch from normal operating mode to "pseudo resonance" mode. Since I assume that we are discussing a fully assembled unit, perhaps it will only operate in "pseudo resonance" mode. You would have to clearly and definitively document all of these things so that there are no uncertainties and ambiguities before you send a new device to Poynt.
If all of this makes sense to you then you can send the new device to Poynt and he can double-check your measurements. The burden of proof rests on your shoulders Lawrence, and only on your shoulders, to send a device to Poynt that he can simply attach a power source to and attach his scope probes and then make measurements that will either confirm or refute your claim of over unity.
This is just my personal suggestion for an action plan. I submit it for your and Poynt's consideration.
MileHigh
There is nothing more I can do.
I tested a device after repairing a loose solder joint, and the device came up COP<<1. Lawrence now claims that my solder repair spoiled the tuning of the unit which is why it no longer operates in pseudo-resonance mode, and therefore not COP>>1.
I would like to send the unit back to Lawrence in hopes that he will use what he has learned here about testing, and re-test this unit after he re-tunes it for optimum performance.
Like I said, I've done what I can. Now it's up to Lawrence to prove to himself that this unit is COP>>1. I expect he will post the results when they become available.
In the mean time, yes; Lawrence, if you have another unit you feel is worthy of testing, by all means I am open to testing it. Just be sure that the unit is properly soldered so that it can not change in any way in transport. The testing I've done on that unit has turned out to be totally in vain, and I would hate for anyone to waste their time on testing when a simple solder joint repair invalidates any tests performed on it.
.99
Quote from: poynt99 on 2011.02.15, 01:49:44
There is nothing more I can do.
I tested a device after repairing a loose solder joint, and the device came up COP<<1. Lawrence now claims that my solder
repair spoiled the tuning of the unit which is why it no longer operates in pseudo-resonance mode
This is possible.
Don't forget that the resonant frequency of a LR, CR circuit depends on R, and if you replaced a badly
soldered joint with significant resistance with one with less, then it would throw resonance out.
Yeah,
And that applies quadruply to pseudo-resonance! That cold solder joint is probably the key to the OU puzzle and, if all solder joints could be made that way, most common circuits and electrical appliances would most likely begin producing copious free energy.
Damn soldering textbooks, anyhow! C.C
Humby
I checked with a colleague at the University from which I retired. I told him I'm looking for a DSO that has:
1. Math multiply function, to permit display and calculations on the P = V*I waveform as we have discussed
2. Built-in calculation of the MEAN (or average), that is, a number for Paverage, as we have also discussed.
Here are the oscilloscopes available for me to borrow and bring to my home lab at this time, and I ask if anyone can help determine whether the two functions above are available on any of these:
Tektronix 2445, 2215 and 2213
Hitachi V-222
LeCroy WaveAce 101.
Thanks.
PhysicsProf:
Your best course of action would be to use Google to find the user manuals, download them, and look for the information yourself.
MileHigh
@ Paul-R
That's a rather stupid statement. Your blinders are affecting what is reality! There is NO excess energy in this Lawrence set-up - bad solder joint or not. Just lack of electronics understanding and not knowing how to measure power. Seriously people ....
cheers
chrisC
Quote from: PhysicsProf on 2011.02.15, 20:05:51
I checked with a colleague at the University from which I retired. I told him I'm looking for a DSO that has:
1. Math multiply function, to permit display and calculations on the P = V*I waveform as we have discussed
2. Built-in calculation of the MEAN (or average), that is, a number for Paverage, as we have also discussed.
Here are the oscilloscopes available for me to borrow and bring to my home lab at this time, and I ask if anyone can help determine whether the two functions above are available on any of these:
Tektronix 2445, 2215 and 2213
Hitachi V-222
LeCroy WaveAce 101.
Thanks.
Professor,
I don't think any of those are purely digital, nor have the required functions. They are all quite old and mostly analog.
Did you recently buy a scope? I suppose it does not have what we need?
The ideal would be to borrow the relatively new DPO Tek scope you were using there....but I suppose they don't want to lend that one off campus....can't blame them.
.99