For various reasons, asking that all posts in this thread be Terse and Technical. Q's here OK if terse, or by PM. Eschew dialog. No term "C OP" or "O U" pls. Pls replicate.
Vin 2.5 V AA's
Rb 51K
MPS2222
C-B 151 pF
D red LED
L-B, L-O bifilar 9turns, ferrite toroid 1"OD, ~90uH each
Ro 0
Rr 0
Cout 10,000uF @16V, disconnected normally
I(t) by V over 1ohm CSR's
Then
DVM 5-7: ~zero (DSO shows Vpp 12.7V, f ~ 1.4MHz)
DVM 5(red)-6(black) + 5.1V (DSO: Vpp ~5.1V), Co charges to ~4.8V in ~5 secs when added
DVM 6(black)-7(red) +0.6V (DSO: Vpp 3.4V) Dim red.
Reps should show ~same as first step.
To
Ro 9.8K
Rr 3.1ohm
Rb 2K
No Co
I(t) measured over 1ohm CSR's
Then see Tek3032 attached, Pin left red waveform, Pout right.
For Pin, DPO gnds @2 (see schematic); Photo
Probes always x10
Ferrite toroid 1"OD, 0.5"ID, 7/16" high, electronic goldmine G6683 or whatever you've got - try it.
Insulated Cu, 22 gauge (20-23). Now fun part: Bifilar winding CW, see attachment, 9 (+/-) turns.
Combined pair to point 7, then one wire to C-B, other to point 4 (schm). O0
Contrast circuit sj1 with "standard" Joule Thief circuit, attached, JT has L's BEFORE xistor, and has much higher P consumption (Pin).
sj1 has LB-CB tank circuit, opens&closes xistor gate rapidly, sets freq~. Also RB and Rr limit current flow, thus limit Pin in sj1 circuit.
Goal is small Pin (and large Pout in closed "OUT" portion, with Pout dumped on [R-D or C-O] =load). OUT circuit loop allows larger current flow there. Diode (or LED) direction in post 1 is correct.
Attached scope shots show Pin (green P waveform oscillates around 0, mean~small) whereas Pout integral clearly non-zero -- that's the pattern seen with sj1.
Two examples of Pin for sj1 type circuits attached; one shows Pout waveform comparison (Vin @ 2.92V); other shows mean Pin values around 1.3mW (Vin@2.06V).
OTOH, JT circuit typically shows saw-tooth waveform -- much different.
Tek3032 displays for sj1 type circuit @2.06Vin. Cf Pin and Pout waveforms (red) and values. Different wound-toroid and circuit board from reply#1 above; replication effciency results tend to corroborate.
I welcome technical comments. O0
Schematic repost, easier to see JPEG I think. Can someone re-do this schematic with SPICE or equiv? thx, much obliged.
Editing: argh, on re-checking everything, I found one probe on 1X rather than 10X -- moved inadvertently, somehow.
So -- here are corrected results = with both probes on 10X
DVM shows potential low is at pt. 6 in circuit. So I attached grnds of ATTEN DSO x10 probes to pt 6.
Vin 2.5V, Rr=0, no Cap Co. (Different build than in post #1)
Channel 1 probe to pt 7; across LED, yellow waveform.
Channel 2 probe to pt 5; across Ro = 9.7K, blue waveform.
Results interesting, displayed on attached (sorry a bit fuzzy); note that ch1 crosses zero (I added black dashed line at V=0) goes flat, and note large voltages:
Ch1 (LED) rms 2.56 V, pp 6.16, mean 1.84
Ch2 (Ro) Rms 3.00 V, mean 1.60
Professor
I have been following this thread with interest. You stated your goal in one of the posts.
QuoteGoal is small Pin (and large Pout in closed "OUT" portion, with Pout dumped on [R-D or C-O] =load).
With this in mind, as an old hand at switchmode design, can I ask a question and make a few recommendations ?
1) In light of your goal, what function does Rr perform? You state that it is used as a current limit. Considering this, it will certainly waste power.
2) Consider operating at higher Vin, where semiconductor Vdrop will be less significant. You will need to readjust bias.
3) Use a good quality Schottky diode rather than LED. Determine the sweet spot value for the output load resistor. The idea is to keep current through the diode low so as not to waste power there, unless you want to calculate that power and use it as part of the output consideration, even though it is unusable except as a heater.
4) With the output capacitor connected, there will be high peak currents in the diode. Best to not use it unless you need a filtered output. A small choke will limit peak current if it Cout must be used. Consider synchronous rectification for lowest loss in the rectifier device.
5) Readjust turns ratio so that power is not wasted in unnecessary base drive. There is nothing magic about a 1:1 ratio. Personally I like to use the heaviest gauge wire I can for the main winding, while leaving enough room on the core for just enough turns of a thinner gauge to sustain oscillation. Since any decent transistor will have a current gain of at least 100, keep the ohmic losses in the main winding as low as possible.
6) If you can operate above 5 volts, consider using a FET instead of BJT, as the drive requirements will be lower.
7) If size is not a problem, use a larger core with same number of turns. The operating frequency will be lower, but this will mean less switching loss overall. Generally if you can keep the operating frequency low, core loss and transition loss will also be reduced. Judging from the waveforms in "try3" scope shot, I would hazard a guess that you have lots of switching loss from the rise and fall slopes.
Finally, I see nothing magic about the common collector configuration vs. the common emitter configuration of a standard blocking oscillator. The only slight advantage may be that base drive also appears in the inductor ramp up current, but this is two orders of magnitude less important than other lossy circuit problems.
Those are my first cut suggestions. While some of these suggestions are more applicable to power supply design in the 10's of Watts range, if you are really trying to squeeze the last bit of efficiency, they can be good "rules of thumb".
You may wish to restate your goal if the above has violated any constraints.
Where there are constraints, engineering involves compromise.
If the only goal is maximum efficiency, no holds barred, we have a lot of room to redesign.
If you must stay with that particular core, operating voltage, switching frequency and circuit configuration, then we can only make minor gains in efficiency.
thats a very old circuit there..
last year i thought i made a new circuit setup only to find that it was a hartley..
i've have built this circuit more than a hundered times.. and you dont need all the resistors in there..
i liked this video cause i half light the neon with only 120 micro amps and 1.1 volts..
i have ran this circuit upto 34 volts.. and down to 0.1 volts..
this is a very efficient circuit far better than any collector type circuit..
watch my hartley.. videos and you will see it run alot of leds in parallel
http://www.youtube.com/user/koolerization#p/u/2/wav6mvzUdZs (http://www.youtube.com/user/koolerization#p/u/2/wav6mvzUdZs)
Quote from: ION on 2011.05.21, 01:06:03
Professor
I have been following this thread with interest. You stated your goal in one of the posts.
With this in mind, as an old hand at switchmode design, can I ask a question and make a few recommendations ?
1) In light of your goal, what function does Rr perform? You state that it is used as a current limit. Considering this, it will certainly waste power.
2) Consider operating at higher Vin, where semiconductor Vdrop will be less significant. You will need to readjust bias.
3) Use a good quality Schottky diode rather than LED. Determine the sweet spot value for the output load resistor. The idea is to keep current through the diode low so as not to waste power there, unless you want to calculate that power and use it as part of the output consideration, even though it is unusable except as a heater.
4) With the output capacitor connected, there will be high peak currents in the diode. Best to not use it unless you need a filtered output. A small choke will limit peak current if it Cout must be used. Consider synchronous rectification for lowest loss in the rectifier device.
5) Readjust turns ratio so that power is not wasted in unnecessary base drive. There is nothing magic about a 1:1 ratio. Personally I like to use the heaviest gauge wire I can for the main winding, while leaving enough room on the core for just enough turns of a thinner gauge to sustain oscillation. Since any decent transistor will have a current gain of at least 100, keep the ohmic losses in the main winding as low as possible.
6) If you can operate above 5 volts, consider using a FET instead of BJT, as the drive requirements will be lower.
7) If size is not a problem, use a larger core with same number of turns. The operating frequency will be lower, but this will mean less switching loss overall. Generally if you can keep the operating frequency low, core loss and transition loss will also be reduced. Judging from the waveforms in "try3" scope shot, I would hazard a guess that you have lots of switching loss from the rise and fall slopes.
Finally, I see nothing magic about the common collector configuration vs. the common emitter configuration of a standard blocking oscillator. The only slight advantage may be that base drive also appears in the inductor ramp up current, but this is two orders of magnitude less important than other lossy circuit problems.
Those are my first cut suggestions. While some of these suggestions are more applicable to power supply design in the 10's of Watts range, if you are really trying to squeeze the last bit of efficiency, they can be good "rules of thumb".
You may wish to restate your goal if the above has violated any constraints.
Where there are constraints, engineering involves compromise.
If the only goal is maximum efficiency, no holds barred, we have a lot of room to redesign.
If you must stay with that particular core, operating voltage, switching frequency and circuit configuration, then we can only make minor gains in efficiency.
"1) In light of your goal, what function does Rr perform? You state that it is used as a current limit. Considering this, it will certainly waste power." The idea I had is to limit Pin, to increase the Pout/Pin ratio. I have some further results that indicate this may be the case (later).
Yes, my goal is maximum efficiency, and I very much appreciate your comments, Ion.
Quote from: kooler on 2011.05.21, 03:14:12
thats a very old circuit there..
last year i thought i made a new circuit setup only to find that it was a hartley..
i've have built this circuit more than a hundered times.. and you dont need all the resistors in there..
i liked this video cause i half light the neon with only 120 micro amps and 1.1 volts..
i have ran this circuit upto 34 volts.. and down to 0.1 volts..
this is a very efficient circuit far better than any collector type circuit..
watch my hartley.. videos and you will see it run alot of leds in parallel
http://www.youtube.com/user/koolerization#p/u/2/wav6mvzUdZs (http://www.youtube.com/user/koolerization#p/u/2/wav6mvzUdZs)
Clearly, the general blocking oscillator (also Joule Thief) circuit is quite old, no question.
I enjoyed your video, Kooler, but did not see a schematic... Could you then provide the schematic, showing especially the LEDiode direction, and the placement of the emitter in the circuit? Also the neon. Would be great -- a sketch will do and will allow comparisons.
here are some of my older schematics..
http://www.overunity.com/index.php?topic=9733.0 (http://www.overunity.com/index.php?topic=9733.0)
and i appreciate your interest in this type of osc.. so i will make you a diagram of the flyback on this circuit and share more circuits of this type..
but it may be in 10 or so hours before i get around to making any..
but i will give you alot of schematics.. even my one that uses the collector and emitter to pulse the ferrite or air core..
the only problem is that is in the 106 khz range and higher..
but one circuit is 657v for 1.5 mins on a 3300uf cap.. very low current..
robbie
Hey, Robbie -- I look forward to you continued input on this one! I'm truly learning a lot, having fun for a retired guy. (Retired at age 57 few years ago)
OK -- I tried something I was sure would work. I put an LED in the OPPOSITE direction across the resistor Ro at 9.7Kohms. See schematic -- this added LED would be in place of cap Co, LED "pointing" down. (Also, I replaced the original red LED with a yellow one to allow seeing them both, no mistaking reflections one to the other.) So now we have TWO LED's in opposite directions -- and guess what, they both light up! Photos show the data; sorry one is particularly fuzzy w/my little HP camera.
Data: Ch 1 (yellow LED in "D" location on schematic): Vpp 7.04 V, Vmean 720 mV.
Ch 2 (red LED across Ro, parallel to Ro): Vpp 7.20 V, Vmean 640 mV, so voltages close to = over each LED, with current flowing (one could say) in both directions out of point 6.
Given that Ro is at 9.7Kohms, I'd guess not much current flows through it.... very roughly, (few V/9.7Kohms) -- so of order, perhaps a few hundred microamps through resistor.
So where do the LED-currents come from, sloshing through the opposing LED's, both joined to point 6 as current-source point?
Quite a puzzle, to me...
Quote from: PhysicsProf on 2011.05.21, 05:35:36
...
So now we have TWO LED's in opposite directions -- and guess what, they both light up!
...
When reverse polarized, a LED presents a capacitance (from pF to tens pF) that allows high frequency currents to pass.
So I presume that in turn at each alternation, a LED light up with current in normal direction while the other acts as a capacitor in series.
You may be right, exnihilo, about the LED's acting as caps. Of course, the connected and opposed LED's both lighting up is a curiosity compared to the Pout/Pin ("efficiency") that I'm getting with this little circuit already using the best measurement methods I know of at this time. If you look at post #1, you see 79mW/10mW, which is interesting. Then post #6 shows roughly the same ratio; as I wrote:
Quote" Different wound-toroid and circuit board from reply#1 above; replication effciency results tend to corroborate."
So yes, I think this circuit is worth pursuing a bit further, and checking out. I may be wrong, but I think it is more than a curiosity. Thanks, Kooler, for the thread you pointed to -- lots of good material there.
And as ION noted,
QuoteIf the only goal is maximum efficiency, no holds barred, we have a lot of room to redesign.
That really is the goal, max Pout/Pin, with the
best possible and redundant measurements to accompany the "redesign" development. I thank you for joining me on this quest for understanding. OK, I'm going to call it a quest for
"energy truth"... What is going on here? is what I'd like to know. Data-based, though; I'm not so interested in fanciful theories at this stage.
(BTW -- Saturday, will be gone most of the day. Look forward to comments.)
..
Kooler,
The two circuits on the right of your diagram
seem to have connection errors in that the
power source is directly attached to the transistor
collector and emitter.
Will you have a look at them again?
Quote from: Dumped on 2011.05.21, 20:35:31
Kooler,
The two circuits on the right of your diagram
seem to have connection errors in that the
power source is directly attached to the transistor
collector and emitter.
Will you have a look at them again?
thanks for looking over it.. cause i am known to make mistakes..
but yes the circuits are correct..
looks weird doesn't it..
Quote from: kooler on 2011.05.21, 05:14:53
here are some of my older schematics..
http://www.overunity.com/index.php?topic=9733.0 (http://www.overunity.com/index.php?topic=9733.0)
and i appreciate your interest in this type of osc.. so i will make you a diagram of the flyback on this circuit and share more circuits of this type..
but it may be in 10 or so hours before i get around to making any..
but i will give you alot of schematics.. even my one that uses the collector and emitter to pulse the ferrite or air core..
the only problem is that is in the 106 khz range and higher..
but one circuit is 657v for 1.5 mins on a 3300uf cap.. very low current..
robbie
Intriguing circuits, Robbie. Can you point to schematic, this one: "one circuit is 657v for 1.5 mins on a 3300uf cap"? Amazing. Have you ever measured Pout/Pin on this?
I'd be glad to do this measurement if you will permit a replication...
PS -- I hope some will replicate the little sj1 circuit from post #1.... and help draw a proper schematic of it.
Pps -- LCR Meter for $17. delivered: http://cgi.ebay.com/LCR-RCL-INDUCTANCE-CAPACITANCE-RESISTANCE-METER-W-Leads-/260749773429?pt=LH_DefaultDomain_0&hash=item3cb5e5ca75 I ordered one so I can measure L especially, on my hand-wound toroids etc.
Quote from: PhysicsProf on 2011.05.20, 20:54:25
...
Can someone re-do this schematic with SPICE or equiv? thx, much obliged.
...
Here is a rough LTspice simulation in the attached txt file. It must be renamed in ".asc" in order to be open in LTspice (the forum doesn't accept files with asc extension although LTspice asc are text files. Could it be corrected?).
It is just a base to go further.
K1 is a directive for establishing the mutual coupling coefficient of L0/Lb.
On the jpg, the green trace is the emitter voltage. The blue trace is the CSRin current.
We see that we are far from Mhz frequencies. Either the real inductances have lower values or the toroid core is saturated, phenomenon that LTspice doesn't deal with, except if we give it the model for a real toroid transformer including the saturation curve (rather tedious to make).
We have also to replace ideal NPN and diode with real ltspice models (not yet checked, but I think there is one for 2N2222 and for 1N4148) and then to see what is the input/output energy. It is unlikely that LTspice shows overunity because it is based on the Kirchoff's laws of currents. Only real experiments could show OU.
Guys,
"asc" files have now been enabled for attachments.
Thanks,
.99
Thanks for the asc, poynt!
Thanks for this, exnihiloest...
In my head, I get about 100 KHz for your simulation results... which indeed is far from the 1.4 MHz observed.
Its late, and I need to retire, but I've checked the frequency for the Lb-Cb tank circuit... which controls the transistor gating, clearly. I get 1.37 MHz (see attached), which agrees with observed.
I'm surprised the sim is so far off from the observed freq.
Thanks again!
Thanks for the change,
.99. Any chance you'll do a replication of this circuit, and tests including Pout / Pin? In my 2nd post above, I noted good results with Rr = 3.1 ohms.
QuoteRo 9.8K
Rr 3.1ohm
Rb 2K
No Co
I(t) measured over 1ohm CSR's
There are reasons, mainly good results, that
I would like to show the Rr ("return resistor") in the circuit, XNest, as I had in my schematic drawing. Could you add that to your nicely drawn schematic kindly? It will not affect the sim much, adding 3.1 ohm, but I sometimes increase that resistance and take Pout/Pin measurements, so I would like to SHOW the presence of this Rr in the diagram, if you would pls.
One of my main concerns is that the Pout calculation in the Tek 3032 may have a DC offset in the voltage measured across the CSRout, which would result in an incorrect quantitative measurement of Pout, although qualitatively it does appear positive on both the Tek and on the ATTEN scopes, and significantly larger than MEAN Pin. MEAN Pin is close to zero as the input current oscillates around zero. But being QUANTITATIVE accurately is not so easy. It is an intriguing little circuit.
I get the two opposing LED's in series to light up brightly now, with Rout removed entirely from the circuit.
Quote from: PhysicsProf on 2011.05.23, 13:33:36
Thanks for the change, .99. Any chance you'll do a replication of this circuit, and tests including Pout / Pin?
Professor,
I'm currently on vacation. When I return, I will likely be building the oscillator Rose is currently working on. Glen has sent me some additional MOSFET's, so I can build the full version. I'll be doing some testing on that circuit.
Hopefully the sim for your circuit can be developed to produce results closer to what you are seeing. That way you'll be able to probe anywhere in the circuit and perhaps determine where the differences are between the sim and bench unit.
Regards,
.99
PhysicsProf,
Ex has pointed out K1 (coupling). This, and the circuit values of the secondary components, will have an effect upon the result frequency compared to the results of the web calculator you used.
An increased secondary load will increase the frequency since this decreases the primary inductive impedance reactance (more appropriate).
I would expect the measured frequency to be a bit higher than the web calculator says.
When all of these factors are entered into the sim the simulator should reflect what you are seeing, as far as frequency goes. I'll not speculate on the rest.
Quote from: poynt99 on 2011.05.23, 17:41:34
Professor,
I'm currently on vacation. When I return, I will likely be building the oscillator Rose is currently working on. Glen has sent me some additional MOSFET's, so I can build the full version. I'll be doing some testing on that circuit.
Hopefully the sim for your circuit can be developed to produce results closer to what you are seeing. That way you'll be able to probe anywhere in the circuit and perhaps determine where the differences are between the sim and bench unit.
Regards,
.99
Good luck on that Rose oscillator testing. And have a good vacation ;)
I'm more interested in real replications of the sj1 circuit than sim runs, fun as they are, and appreciate exnihiloest for the schematic-drawing.
I'm consistently double-checking things. The Pout/Pin results seem quite solid, results given above.
About the "opposing LED's" however, I
double checked the yellow LED which I had not used before -- and bottom line is that I got the direction wrong for it.
So I had two LED's in series in the SAME direction... nothing surprising there that they both lit up. Again, that was not the basic circuit in any case (was just for fun and learning) and in no way affects the measurements done on Pin and Pout given.
Quote from: PhysicsProf on 2011.05.23, 13:33:36
Thanks for the change, .99. Any chance you'll do a replication of this circuit, and tests including Pout / Pin?
In my 2nd post above, I noted good results with Rr = 3.1 ohms.
...
I have added Rr and replaced the standard diode by a low power LED from the LTspice library. It doesn't change much. We see that the frequency is less than 100 Khz, which is compatible with the time constant 1/Rb*Cb.
See first trace in: FEosc-traces.png
Then I have considered the capacitance between Lb and L0 due to their proximity. In real life, there are always parasitic capacitors :(. So I added Cp=20pF connecting the two coils.
The simulation seems much more realistic, see the second trace in: FEosc-traces.png.
The time constant 1/Rb*Cb still plays its role, but modulates a new signal due to Lb/Lo/Cp ringing (around 2 Mhz).
Then I replaced the ideal NPN by the LTspice model of 2N2222.
Now the time constant 1/Rb*Cb does no more appear except for the first burst.
See the third trace in: FEosc-traces.png
Finally I visualized the power and the energy (COP<1, no surprise). The output power is calculated from the product of the voltage available at the emitter by the current passing through the LED/R0/CSRout.
See the final circuit and the traces of the power in FEosc-Power-Energy-schematics.png
For those not familiar with LTspice, one can get the energy or rms values by a ctrl+click on the title of the trace in the simulation window.
LTpsice model is in FEosc-2N2222.asc.
May be I will duplicate the circuit if time permits.
Thanks again, XN.
So the SIM can't get the Pout/Pin that we see on the Tek 3032 -- as you say, not surprising that the SIM can't get it.
I wonder if it can do another task. With these conditions:
Rb = 2K
Rr = 0
D = 1n4148 diode
Vin = 2.6 V
10,000 uF (16V) cap IN PLACE OF Rout, oriented as shown in schematic below.
Then, I measure with a stop-watch that the Co cap charges from 0 to 7 volts in 2.5 seconds, as soon as I turn the circuit on. Can the SIM get any thing like that?
I thought this was a significant result. Eout can be calculated from E = 1/2 CV**2.
I'm open for suggestions on determining Ein with this Cap-charging test. Remember that if I connect a charged cap to an equal C uncharged cap, I lose half the energy in the process, so I' rather not use a charged-cap as the input Ein.
I also bought a few high-C caps, hoping to use these in lieu of a battery for Vin. BUT -- I found that they "leak" somehow, rather steadily. The 10,000uF caps hold the charge better, but I run into the problem alluded to above ( if I connect a charged cap to an equal C uncharged cap, I lose half the energy in the process...).
I put the conditions:
Rb = 2K
Rr = 0
D = 1n4148 diode
Vin = 2.6 V
Cout = 10,000 uF (16V)
I have had to reduce the calculation step otherwise the simulation fails. With short steps, the sim is very slow (each HF period must be sampled in many steps and there are more a million periods/s). I must end the sim after around 10ms.
Nevertheless the sim shows that the capacitor charges very linearly at 6.5 mV/ms and the power provided by the battery is a stable 272mW.
After 1s, we can estimate that the sim would give a wasted energy of 272mJ and the capacitor would be charged at 6.5v, storing an energy of 0.5*C*U2 = 0.5*10-2*6.52 = 211mJ.
A sim will never show OU (if there is no bug :-) ).
I suggest that you build a second strictly identical circuit but no battery, and you power it with the capacitor of the first one. A 2.6v zener in parallel with the capacitor will limit the voltage at the same value as the battery.
If the second circuit can also charge its own capacitor to 2.6v, there is suspicion of OU. Then to be sure, measure the currents in the Csrin of the 2 circuits. If they are equal, or if the second one is more than the first one, there is clearly OU, and to loop the two circuits without the battery should be conceivable in a next step.
I'll try your suggestion, XN, or equivalent... Again, just discharging a cap into a 2nd equal cap will necessarily involve a loss of half the initial stored energy. (Doing the math, E = 1/2CV**2) But I'll give your idea a go to see what happens, experimentally. I have several errands this morning first...
Meanwhile, could you kindly post your circuit diagram for the above from SPICE? thanks for doing all that work, BTW -- very helpful.
- deleted due to calculation errors
(deleted per request of person to whom I was replying)
Thanks for doing the SIM, though -- these can be a lot of work, and its a learning opportunity for us all.
- deleted due to calculation errors
This is interesting... Glad we agree on the use of the MEAN rather than RMS.
I like your positioning of Rr in this instance also.
- deleted due to calculation errors
(deleted per request of poster to whom I was replying)
- deleted due to calculation errors
- deleted due to calculation errors
Quote from: PhysicsProf on 2011.05.25, 14:25:46
I'll try your suggestion, XN, or equivalent... Again, just discharging a cap into a 2nd equal cap will necessarily involve a loss of half the initial stored energy.
The loss of half the energy is true only when the capacitor is charged from a voltage source, not true when it is charged from a current source. When a magnetic field through a coil collapses, the coil acts as a current generator, charging a capacitor without losses (allowing for example LC oscillators weakly damped in spite that the capacitor is periodically charged and discharged from the coil). We see that it is the case here, because the capacitor is negatively charged: only the coil can provide such a current. So this point is not worrying.
There is an equivalent way to charge a capacitor theoretically without losses and from a voltage source, which is named "adiabatic charge". It had been studied by IBM to reduce the consumption of VLSI circuits due to the gate capacitance of MOSFETs: the voltage source must be controlled. It is increased slowly and maintained very near and just above the capacitor voltage.
Quote
Meanwhile, could you kindly post your circuit diagram for the above from SPICE? thanks for doing all that work, BTW -- very helpful.
See the attachment. The Trtol parameter of the simulation (in control panel, "SPICE" tab) is put to 0.01 instead of 1, its default value.
Another point: when the capacitor is disconnected, the sim shows that the emitter of the 2N2222 is periodically at a positive voltage up to 10v (even more at the begining). We had seen elsewhere that this transistor presents a negative resistance when reverse polarized around 6v. I had checked that LTspice fails in simulating this negative resistance. In the circuits with negative resistance, the transistor base is disconnected but I had also tested the negative resistance with the base connected in some ways. This point could be of interest in the functioning of the real circuit.
Quote from: duff on 2011.05.26, 07:04:33
In an effort to optimize Rb again (now looking at 2ms of data), I step through standard values of resistance from 51K to 20Meg.
The highest COP resulted with Rb = 3.3Meg which gave a COP = 4.1.
Frequency of 1.3MHz.
t2: v(vout)=2.0v AT 1.49641e-006
t1: v(vout)=2.0v AT 7.29945e-007
freq1: 1/(t2-t1)=1.30469e+006
ir1: AVG(i(csrin))=-1.20317e-005 FROM 0 TO 0.00209
ir2: AVG(i(csrout))=-1.01898e-005 FROM 0 TO 0.00209
ir3: AVG(i(rr))=2.22216e-005 FROM 0 TO 0.00209
vout: AVG(v(vout))=0.00748531 FROM 0 TO 0.00209
vr1: AVG(v(vcsrin))=-1.20317e-005 FROM 0 TO 0.00209
vr2: AVG(v(vcsrout))=-1.01898e-005 FROM 0 TO 0.00209
vr3: AVG(v(vrr))=2.22216e-005 FROM 0 TO 0.00209
cop: ( (vr3*ir3) + (vr2*ir2) ) / (vr1*ir1)=4.12836
AVG(v)*AVG(i) is not equal to AVG(v*i).
You must use instantaneous values of u*i and integrate the product over time.
(LTspice obeys Kirchoff's law: sum of currents in a node must be zero. So it can't show COP>1).
Quote from: exnihiloest on 2011.05.26, 07:33:00
AVG(v)*AVG(i) is not equal to AVG(v*i).
You must use instantaneous values of u*i and integrate the product over time.
(LTspice obeys Kirchoff's law: sum of currents in a node must be zero. So it can't show COP>1).
Thanks
I would like to thank Duff for his input. And if there was a mistake, so what? we are here to learn. And to probe the frontiers of energy research. Not likely we'll do this without a small mistake, or two.
We don't expect LTSpice to tell us when a device is "overunity". IMO, only experiments can do that -- experiments including careful measurements on Pinput and Poutput. I agree with exnihiloest that a self-running device would be a compelling demonstration, if it could be replicated and done repeatedly. Repeatability is the heart of good science and progress. Such a device should also allow us to determine the source of the "anomalous" incoming energy.
Sterling Allan came by yesterday and videotaped some of the work I'm doing... So you'll be able to see more of the device and results.
Traveling today and then a long trip next week -- cheers, then.
Quote from: duff on 2011.05.26, 11:49:10
Thanks
Think nothing of it. I saw that you used "AVG", a key word that I didn't know. I just found it now in LtSpice help. This opens me interesting possibilities. I thank you also.
I apologize if these questions are stupid...
Regarding your CB - LB tank circuit.
90uH and 151pF = 1.37MHz (using my web LCR calculator)
151pF seems like a odd value so I was wondering what are your constant items and what are variable?
90uH is most likely constant (could be varied with +/- windings)
Was 151pF constant or a variable capacitor?
Is 1.37MHz a constant value you were seeking or just the end result of using those 2 components?
Thank you for your time,
DonL
Quote from: dllabarre on 2011.05.27, 19:11:46
I apologize if these questions are stupid...
Regarding your CB - LB tank circuit.
90uH and 151pF = 1.37MHz (using my web LCR calculator)
151pF seems like a odd value so I was wondering what are your constant items and what are variable?
90uH is most likely constant (could be varied with +/- windings)
Was 151pF constant or a variable capacitor?
Is 1.37MHz a constant value you were seeking or just the end result of using those 2 components?
Thank you for your time,
DonL
Good questions, DonL. 151 pF is what I had several of, readily available -- a constant cap.
Yes, once the toroid was wound, it was about 90uH and could be varied with +/- windings, but left as a constant.
The combination in a tank circuit gives 1.37 MHz, as you say, the end result of using those 2 components.
I should add that sometimes the scope has difficulty picking out the frequency... until I properly adjust the trigger (sometimes, I don't bother with that).
@PhysicsProf,
Congratulation in making my circuit, the one I made you aware of at 2011-03-11, 23:10:17, go over unity.
( http://www.overunityresearch.com/index.php?topic=764.0 ) I'm really pleased to see that it was possible
to tune my circuit design in the way you did. Keep up the good work. :-)
Groundloop.
Quote from: PhysicsProf on 2011.05.20, 02:28:21
To
Ro 9.8K
Rr 3.1ohm
Rb 2K
No Co
I(t) measured over 1ohm CSR's
Then see Tek3032 attached, Pin left red waveform, Pout right.
@PhysicsProf,
First of all, congrats on your progress! But I think we should also be very careful with the results. I noticed that the frequency shown in your picture are dramatically different (3.926MHz on the Pin measurement, 37.10MHz on the Pout measurement). Therefore I am wondering if it is because the circuit is so sensitive that by putting the probes at different places, the whole thing changed. If that's the case, your measurement might make little sense...
I suspect that the frequency should be the same on both the input and output side....but I could be wrong. Therefore I think it would be great to measure the frequency on both input and output sides at the same time. Here is a possible way of measurement: put the ground at point 4. Then one probe at Point 2, another probe at Point 5. This way we can measure if the frequencies on both sides are the same or not.
lanenal
Quote from: PhysicsProf on 2011.05.23, 08:05:28
Thanks for this, exnihiloest...
In my head, I get about 100 KHz for your simulation results... which indeed is far from the 1.4 MHz observed.
Its late, and I need to retire, but I've checked the frequency for the Lb-Cb tank circuit... which controls the transistor gating, clearly. I get 1.37 MHz (see attached), which agrees with observed.
I'm surprised the sim is so far off from the observed freq.
Thanks again!
I noticed exnihiloest used Rb=51K in his schematic, and got roughly 100KHz. But in your second post, your value was Rb=2K. When I changed to Rb=2K in exnihiloest's schematic, I got 11/7 = 1.57 MHz, which is also close to your 1.4MHz observation...
lanenal
P.S.: when I changed the "K1 Lb L0 0.9" to "K1 Lb L0 0.5", Kb=2K, I got 1.38MHz with simulation.
I built the real circuit yesterday. I have some problem for it to work. I was unable to make oscillate a 2N2222A.
With a transistor better designed for HF frequencies, such 2N2369 or 2N2218, the circuit oscillates but at frequencies around 80 Mhz! Obviously this frequency is not related to L and C values. I tried with 2 different toroids and coils, and I got similar results. Moreover one of the two coils can be reversed and the oscillation remains the same.
The only result according to the simulation is the "modulation" of the HF signal by a signal of much lower frequency, linked to the time constant RbCb, only when Cb is exceeding a certain threshold value (that I don't know, I used a variable capacitor).
For me it is a variant of a Colpitt's oscillator, so I don't see what is wrong. :(
Quote from: PhysicsProf on 2011.05.20, 04:12:47
Ferrite toroid 1"OD, 0.5"ID, 7/16" high, electronic goldmine G6683 or whatever you've got - try it.
Insulated Cu, 22 gauge (20-23). Now fun part: Bifilar winding CW, see attachment, 9 (+/-) turns.
Combined pair to point 7, then one wire to C-B, other to point 4 (schm). O0
This might be of interest for some who would like to replicate:
[quote link=http://forum.allaboutcircuits.com/archive/index.php/t-39439.html]
I just ran some tests on that 5-pack of toroids I received from Electronic Goldmine: G6683
The Al value is around 10,000, which is quite high.
10 turns of magnet wire gave me 936uH; 20 turns were 4mH.
[/quote]
Quote from: exnihiloest on 2011.05.28, 11:03:46
I built the real circuit yesterday. I have some problem for it to work. I was unable to make oscillate a 2N2222A.
With a transistor better designed for HF frequencies, such 2N2369 or 2N2218, the circuit oscillates but at frequencies around 80 Mhz! Obviously this frequency is not related to L and C values. I tried with 2 different toroids and coils, and I got similar results. Moreover one of the two coils can be reversed and the oscillation remains the same.
The only result according to the simulation is the "modulation" of the HF signal by a signal of much lower frequency, linked to the time constant RbCb, only when Cb is exceeding a certain threshold value (that I don't know, I used a variable capacitor).
For me it is a variant of a Colpitt's oscillator, so I don't see what is wrong. :(
It seems to me, the frequency is related to Rb, C, L, as well as the factor K between Lb and Lo.
The 80MHz of your build is sooo high...could that be the reason that your 2N2222 does not work?
Could you simulate the values of your Rb, C, L in your LTSpice?
Quote from: PhysicsProf on 2011.05.28, 06:23:13
Good questions, DonL. 151 pF is what I had several of, readily available -- a constant cap.
Yes, once the toroid was wound, it was about 90uH and could be varied with +/- windings, but left as a constant.
The combination in a tank circuit gives 1.37 MHz, as you say, the end result of using those 2 components.
I should add that sometimes the scope has difficulty picking out the frequency... until I properly adjust the trigger (sometimes, I don't bother with that).
I've already noted, laneal, that " sometimes the scope has difficulty picking out the frequency... until I properly adjust the trigger (sometimes, I don't bother with that)." Haven't you experienced that? The trigger can pick out two prominent peaks, for example, and end up with twice the frequency shown on the scope, than when just one cycle is picked out by the trigger. But note that the f determination does not affect the measurement of the MEAN power, which is taken over Numerous cycles.
It is interesting to evaluate the resonant frequency of the as-built device, and that freq may well depend on various factors, besides Lb and Cb, I agree. It will be fun to map out how the resonant f changes with variations.
When I take the time to adjust the trigger, then the freq measured on the input side, and on the output side, is the same, @ 1.37 MHz -- with small variations which I ascribe to temperature changes and the like.
@Groundloop -- I give you full credit for pointing me in this direction! Before your posts, I had not tried the "reverse JT circuit". Very helpful, and that's why I say "we are in this together" in this open-source development. As I recall, you had the LED oriented in the "wrong direction" -- corrected by EMDevices. I also added a resistor in series with the LED (later replaced with a diode for some tests in my lab), and a resistor Ro -- both variable resistors which allowed for tuning. Thus, we work together in developing a functioning, tunable device.
I hope I emphasized sufficiently on the vid by Sterling that this device needs to be replicated and thoroughly checked out by others before one is sure of overunity! XN --
very glad that you are building the circuit! Glad that you got it to ring with a different transistor. I used an MPS2222 transistor, as I said in an early post.
Les Kraut already did a replication, and it did ring on the first try actually. He used 2Kohms on the Rb, and about 9 turns on the bifilar-wound toroid.
Quote from: PhysicsProf on 2011.05.28, 12:26:05
I've already noted, laneal, that " sometimes the scope has difficulty picking out the frequency... until I properly adjust the trigger (sometimes, I don't bother with that)." Haven't you experienced that? The trigger can pick out two prominent peaks, for example, and end up with twice the frequency shown on the scope, than when just one cycle is picked out by the trigger. But note that the f determination does not affect the measurement of the MEAN power, which is taken over Numerous cycles.
It is interesting to evaluate the resonant frequency of the as-built device, and that freq may well depend on various factors, besides Lb and Cb, I agree. It will be fun to map out how the resonant f changes with variations.
When I take the time to adjust the trigger, then the freq measured on the input side, and on the output side, is the same, @ 1.37 MHz -- with small variations which I ascribe to temperature changes and the like.
Thanks for the explanation. Unfortunately, I don't have such a nice scope. I only worked with the LTSpice simulator, which has a psudo scope.
I don't quite understand the scope shot you have got there in the 2nd post. On the left, the math duration was 2.0us, on the right, the math duration was 200ns -- could you also explain that difference of time duration? Another question: if the frequency was 1.37MHz, then why your scope shots looks so irregular on the right with only 200ns duration? I suggest that for better comparison, both shots should have the same time duration on the scope (since you also seem to agree that both sides have the same frequency).
lanenal
Quote from: PhysicsProf on 2011.05.28, 12:26:05
I've already noted, laneal, that " sometimes the scope has difficulty picking out the frequency... until I properly adjust the trigger (sometimes, I don't bother with that)." Haven't you experienced that? The trigger can pick out two prominent peaks, for example, and end up with twice the frequency shown on the scope, than when just one cycle is picked out by the trigger. But note that the f determination does not affect the measurement of the MEAN power, which is taken over Numerous cycles.
It is interesting to evaluate the resonant frequency of the as-built device, and that freq may well depend on various factors, besides Lb and Cb, I agree. It will be fun to map out how the resonant f changes with variations.
When I take the time to adjust the trigger, then the freq measured on the input side, and on the output side, is the same, @ 1.37 MHz -- with small variations which I ascribe to temperature changes and the like.
@Groundloop -- I give you full credit for pointing me in this direction! Before your posts, I had not tried the "reverse JT circuit". Very helpful, and that's why I say "we are in this together" in this open-source development. As I recall, you had the LED oriented in the "wrong direction" -- corrected by EMDevices. I also added a resistor in series with the LED (later replaced with a diode for some tests in my lab), and a resistor Ro -- both variable resistors which allowed for tuning. Thus, we work together in developing a functioning, tunable device.
I hope I emphasized sufficiently on the vid by Sterling that this device needs to be replicated and thoroughly checked out by others before one is sure of overunity!
XN -- very glad that you are building the circuit! Glad that you got it to ring with a different transistor. I used an MPS2222 transistor, as I said in an early post.
Les Kraut already did a replication, and it did ring on the first try actually. He used 2Kohms on the Rb, and about 9 turns on the bifilar-wound toroid.
@PhysicsProf,
Yes, open source is the way to go. I did not use a LED in my circuit. I did use a diode to indicate a power output.
I'm now trying out the same emitter following method in my figure8 circuit (posted in my bench) to see if I
get better results. The figure8 circuit is a nonlinear circuit and very difficult to get running and tuned. I did
blow my old BUX80 transistors and did replace them with BUX98A transistors. Now the circuit does not
perform as it was new. I can hardly get any output at all. So back to the lab an find out whats wrong. LOL
The figure8 circuit is my first attempt to scale it up to some useful power levels.
Groundloop.
Quote from: lanenal on 2011.05.28, 12:18:05
It seems to me, the frequency is related to Rb, C, L, as well as the factor K between Lb and Lo.
The 80MHz of your build is sooo high...could that be the reason that your 2N2222 does not work?
Hi lanenal,
Problem solved. You are right. My toroidal coil had not enough inductance and probably the ferrite was not suited for the involved frequencies. I replaced it by a big toroid of quite high mu ferrite, and 2 x 16 turns coils. Now all transistors work at frequencies compatible with the simulation, included the 2N2222. Further tests needed.
Quote
Could you simulate the values of your Rb, C, L in your LTSpice?
In a future step. My priority is now the measurements of the real circuit because LTspice is of poor help outside conventional electronics.
Quote from: exnihiloest on 2011.05.28, 13:35:46
Hi lanenal,
Problem solved. You are right. My toroidal coil had not enough inductance and probably the ferrite was not suited for the involved frequencies. I replaced it by a big toroid of quite high mu ferrite, and 2 x 16 turns coils. Now all transistors work at frequencies compatible with the simulation, included the 2N2222. Further tests needed.
In a future step. My priority is now the measurements of the real circuit because LTspice is of poor help outside conventional electronics.
Right! thanks for your replication XN, hope to hear regarding your results here.
@laneal --
Quote"
I suggest that for better comparison, both shots should have the same time duration on the scope (since you also seem to agree that both sides have the same frequency)."
Please see my post #5, which data now I wish had been in post #1 -- since most seem to look at those data rather than the data in post #5, which has both shots of the same time duration. I agree this is helpful -- and I repost the data here for all.
Note that the mean output Power -- taken over many cycles -- is approx 10 mW, whereas the mean input Power is less, around 1.3 mW. Data acquired using a Tek 3032. This was actually one of my first tests of this sj1-circuit, taken April 1, 2011. Not an April fool's joke!
I have re-checked the results and I don't think I made a mistake in measurement. Also, the replication by Les Kraut showed more Pout than Pin. And so I'm asking others to replicate and check this circuit.
Quote from: PhysicsProf on 2011.05.28, 13:48:35
Right! thanks for your replication XN, hope to hear regarding your results here.
Good luck, XN!
Quote
@laneal -- Please see my post #5, which data now I wish had been in post #1 -- since most seem to look at those data rather than the data in post #5, which has both shots of the same time duration. I agree this is helpful -- and I repost the data here for all.
Note that the mean output Power -- taken over many cycles -- is approx 10 mW, whereas the mean input Power is less, around 1.3 mW. Data acquired using a Tek 3032. This was actually one of my first tests of this sj1-circuit, taken April 1, 2011. Not an April fool's joke!
I have re-checked the results and I don't think I made a mistake in measurement. Also, the replication by Les Kraut showed more Pout than Pin. And so I'm asking others to replicate and check this circuit.
This give me much more confidence in your measurements, thanks for pointing this out to me!
Have you thought of using a dc-dc converter to loop the circuit? See
http://en.wikipedia.org/wiki/DC-to-DC_converter
Note that "Most DC to DC converters are designed to move power in only one direction, from the input to the output."
So it would be ideal to use Cout and connect to the input of a dc-dc converter, which will supply the power for the JT. Of course, will need the battery in the beginning to get it started.
cheers,
lanenal
First measurements. See the asc for the circuit.
Rout is the charge. It's value is 1K which fits well the output impedance (when connected, the output voltage from the diode is approximately divided by 2).
I monitor:
- the emitter voltage to check the oscillation (Oscillo Tektronix 2232, probe 1)
- the voltage Vout on Rout (probe 2)
- the collector voltage Vc (HP3468A multimeter)
- the current I from the power voltage, which can be adjusted (analog multimeter 20.000ohm/volt, in series with the power supply and previously checked with the high precision HP3468A).
Input power = Vc*I.
Output power = Vout2/Rout. As Rout=1K, Vout2 gives directly the output power in mW.
All values are cc
I have tested different conditions :
Diode as on the schematics or reversed. Best efficiency is with the diode as on the schematics.
Rb tested from 82K down to 2.2K. The higher the value, the better the efficiency.
Cb from hundreds of pF down to 15pF. 15pF is the best, with a frequency around 1 Mhz.
I took several measures while changing step by step the voltage from the power supply.
The efficiency varies between 25% and 60%, depending on all these parameters. OU is not attained with this method.
Second measurement method, see the asc.
I measure the input power Vc*I as in the first method. I=2.7mA, Vc=10,18v => P=27.5mW=27.5 mJ/s
After 1mn C1 (2200µF) is charged at 20.2v. The C1 voltage is permanently monitored with the HP3468A multimeter. Note that its connection adds parasitic capacitances between the output of each diode and the ground: it is a well known phenomenon of this double diode configuration, that improves the charge of C1 (at the expense of the current that has to be provided).
The energy in the capacitor is: E=1/2*C*U²=0.5*2.2*10-3*20.22= 449mJ
The input power was: 27.5*60=1650mJ. Efficiency: 449/1650=27%. No ou.
I noted that the capacitor voltage increases at a constant rate, for example 0.3v/s. The energy being proportional to the voltage squared, in the same unit of time more energy is provided when the capacitor is already charged than when it is empty.
So I led another measurement with C1 charging from U1=10v up to U2=18.6v. It needs 30s. Energy was 1/2*C*U22-1/2*C*U12=270mJ.
During this charge, the input was I=1.5mA and Vc=10v, P=15mW. Input energy=15*30=450mJ.
Efficiency 270/450=60%. Still no overunity but we see that when the charge is less because the capacitor is already partly charged and don't "shortcuts" the diodes, the efficiency is much better and nevertheless more energy is provided by each unit of time.
Hi everyone,
I couldn't resist testing this circuit a la gotoluc way ;)
I winded on a Metglas 1" o.d. with 1/2" i.d. 7 turns to get close to 90uH
The only resistor I used in the circuit (other than the load) is a 1 meg pot adjusted to 266K to the 2N2222 base. I connected a diode in series to a 30,000uf cap where the output led is in the circuit and used a 4700 Ohm 2% resistor (close to a led load) as load on to cap. A DVM measures the DC voltage across the caps resistor to find true Pout.
I used an old 3 vdc battery as input power to the circuit connected through my special dual 3,900uf capacitors with dual volt meters to get source voltage and voltage across a 10 Ohm 1% resistor to obtain an accurate current reading. This measurement meter was a suggestion by .99 and I have been using it for the past 3 years on pulse circuits to obtaining real time current draw. It has proven to be a very accurate measurement method.
After my first tests I decided to re-wind the toroid to a higher inductance (now about 3.5mH) which reduced the input current by 20% while maintaining the same voltage on the output load.
The shots below are my best results to date but no OU
Luc
Quote from: gotoluc on 2011.05.29, 04:26:29
Hi everyone,
I couldn't resist testing this circuit a la gotoluc way ;)
I winded on a Metglas 1" o.d. with 1/2" i.d. 7 turns to get close to 90uH
The only resistor I used in the circuit (other than the load) is a 1 meg pot adjusted to 690K to the 2N2222 base. I connected a diode in series to a 30,000uf cap where the output led is in the circuit and used a 4700 Ohm resistor (close to a led load) as load on to cap. A DVM measures the DC voltage across the caps resistor to find true Pout.
I used an old 3 vdc battery as input power to the circuit connected through my special dual 4,700uf capacitor with dual volt meters to get source voltage and voltage across a 10 Ohm 1% resistor to obtain an accurate current reading. This measurement meter was a suggestion by .99 and I have been using it for the past 3 years on pulse circuits to obtaining real time current draw. It has proven to be a very accurate measurement method.
After my first tests I decided re-wounded the toroid to a higher inductance (now about 3.5mH) which reduced the input current by 20% while maintaining the same voltage on the output load.
The shots below are the best results but no OU
Luc
Luc,
Please check your uploaded data again. They are identical?
Edit: Thanks for fixing the upload. :-)
GL.
Quote from: Groundloop on 2011.05.29, 04:30:48
Luc,
Please check your uploaded data again. They are identical?
GL.
Yes I know! I'm trying to fix it but it keeps coming back that way for some reason ???
Luc
EDIT Okay I fixed it! the problem was me :P
I did more changes I took the reading on the pot on the wrong side. It's adjusted to 266K Ohm to get the input to 1 ma. I also had a small change on the Pout
Quote from: gotoluc on 2011.05.29, 04:26:29
Hi everyone,
I couldn't resist testing this circuit a la gotoluc way ;)
I winded on a Metglas 1" o.d. with 1/2" i.d. 7 turns to get close to 90uH
The only resistor I used in the circuit (other than the load) is a 1 meg pot adjusted to 266K to the 2N2222 base. I connected a diode in series to a 30,000uf cap where the output led is in the circuit and used a 4700 Ohm 2% resistor (close to a led load) as load on to cap. A DVM measures the DC voltage across the caps resistor to find true Pout.
I used an old 3 vdc battery as input power to the circuit connected through my special dual 3,900uf capacitors with dual volt meters to get source voltage and voltage across a 10 Ohm 1% resistor to obtain an accurate current reading. This measurement meter was a suggestion by .99 and I have been using it for the past 3 years on pulse circuits to obtaining real time current draw. It has proven to be a very accurate measurement method.
After my first tests I decided re-wind the toroid to a higher inductance (now about 3.5mH) which reduced the input current by 20% while maintaining the same voltage on the output load.
The shots below are my best results to date but no OU
Luc
I appreciate your doing a replication, Luc, but it certainly is not clear HOW YOU PERFORMED THE POUT MEASUREMENTS. Pls explain, including where you attached your probes.
I have attached your measurements, juxtaposed for convenience. V= 1.067 Volts across the 4700ohm Rout resistor -- is that correct? RMS, Mean, what? Also, how do you get the current in the output leg, and what is it?
I'm responding also to queries over at OU -- linking here if you wish to read:
http://www.overunity.com/index.php?topic=10773.15
I'm pursuing alternative/additional methods for checking results as explained on the vid and elsewhere; in brief:
I realize there are limits to the oscilloscope method outlined and used, and that is why I am trying other methods as well.
Another method I've started is to compare the temperature rise in the two matched (1-ohm) CSR's. If indeed there is more current circulating in the output leg of the circuit, that resistor CSR(out) should show a greater temp rise than CSR(in). That is the simplest non-oscilloscope test I have thought of so far...[/b
What do you think -- is this a good method, or not (and why)? :
Quotecompare the temperature rise in the two matched (1-ohm) CSR's. If indeed there is more current circulating in the output leg of the circuit, that resistor CSR(out) should show a greater temp rise than CSR(in).
The theory is sound Professor, but there is the issue of the relatively low powers you are dealing with, i.e. 10's of mW, and it is likely that the CSR's will not heat up much above ambient. This could pose a problem especially if the true Pin and Pout are close in value.
You're also facing the requirement for a temperature meter with 10ths of degrees of resolution, or better.
.99
Quote from: poynt99 on 2011.05.29, 05:31:29
The theory is sound Professor, but there is the issue of the relatively low powers you are dealing with, i.e. 10's of mW, and it is likely that the CSR's will not heat up much above ambient. This could pose a problem especially if the true Pin and Pout are close in value.
You're also facing the requirement for a temperature meter with 10ths of degrees of resolution, or better.
.99
Thank you for confirming the approach, .99 -- I like the simplicity of it. Now, as to how much the CSR's heat up:
1. depends on how they are insulated
2. depends on how much we can scale up the input and output power -- especially the output power!
If the output CSR heats up while the input CSR heating is below what we can measure, that result is still significant (further comments welcomed).
Quote from: PhysicsProf on 2011.05.29, 05:17:08
I appreciate your doing a replication, Luc, but it certainly is not clear HOW YOU PERFORMED THE POUT MEASUREMENTS. Pls explain, including where you attached your probes.
I have attached your measurements, juxtaposed for convenience. V= 1.067 Volts across the 4700ohm Rout resistor -- is that correct? RMS, Mean, what? Also, how do you get the current in the output leg, and what is it?
Hello PhysicsProf,
all my measurements are done in DC, no scope used.
Pre the circuit a DVM in DC measures the voltage (milli volts) across a 10 Ohm resistor which is connected between two 3,900uf capacitors to smooth out pulses. So Pin is measured in DC Amps by V/R=I
Yes, Pout is V= 1.067 Volts across the 4700ohm Rout resistor... however this is also now DC (30,000uf cap)
I use this online calculator to get the shots I posted above: http://www.sengpielaudio.com/calculator-ohm.htm (http://www.sengpielaudio.com/calculator-ohm.htm)
I hope my explanation helped. If not, I'm sorry but I can't do any better as I have no schooling. All I know and do has been self taught.
Many here know my research style and they can help if you need more information on this.
Luc
Luc:
QuoteYes, Pout is V= 1.067 Volts across the 4700ohm Rout resistor... however this is also now DC (30,000uf cap)
I don't understand -- where is the 30K uF cap placed? just use words if you can't do a schematic drawing.
1.067 V across the 4700 Ohm Rout is significantly low, compared to my circuit and the replication by Les K.
This is a start, but would you also kindly address how you measured the
output current as I asked you?
So far, your Pout measurement makes little sense, sorry. Pls add a little more explanation.
PS -- @ .99 :
I will also move CSRin to the other side of Vin for this test, to further isolate CSRin from the matched CSRout. Both greater than 1ohm, certainly. Rb would remain, but other resistors would be eliminated. Replace the battery with a big Cap, to eliminate internal resistance of Vinput.
The goal is to compare Pin vs Pout using CSRin and CSRout only, comparing heat rise in these... Not sure I can see how to make this work, especially for Pin via a sensing resistor. Pout via a dump-resistor is relatively easy. BUT any comments on how to make this heat-rise-in-resistors method work would be appreciated.
Thanks, Steve.
Hi all,
i made an attempt to replicate this circuit, but as we already found out before, the problem is the Pout.
How to accurately measure that.
Anyway, i made a coil, both windings are 12 turns, and measure 0.93mH (930uH)!
Rb is a 2.2K resistor,
Cb is 140pF,
Rr is a 3.3 Ohm resistor
Do is a white led
Ro is a combo, 5.6K resistor and 5K potentiometer
csrin and out are both 1 Ohm resistors
Using a 2N2222A transistor
Frequency turns out to be around 525Khz.
Using 1 rechargeable AA battery with about 1.195V
Using the dual DMM method (low pass filter) for Pin, i found it to be around 40mW
But now how to measure the Pout correctly, that's the question.
I got some suggestions to dump the output (over the Ro) into a cap, but still then, how to measure the Power?
The csrout is floating, so no way to measure there with my scope.
Anyway, i am open for suggestions and meanwhile will try to find some way.
(i am thinking to separate the output pulses directly taken from the collector via a 1:1 transformer/toroid,
use a FWBR and a cap to create DC, then put a load on it (bulp), and measure there the voltage/current).
I video of this attempt is to be found here: http://www.youtube.com/watch?v=vV4xMeZ_41Y
Regards Itsu
Quote from: Itsu on 2011.05.29, 11:35:56
Hi all,
i made an attempt to replicate this circuit, but as we already found out before, the problem is the Pout.
How to accurately measure that.
Anyway, i made a coil, both windings are 12 turns, and measure 0.93mH (930uH)!
Rb is a 2.2K resistor,
Cb is 140pF,
Rr is a 3.3 Ohm resistor
Do is a white led
Ro is a combo, 5.6K resistor and 5K potentiometer
csrin and out are both 1 Ohm resistors
Using a 2N2222A transistor
Frequency turns out to be around 525Khz.
Using 1 rechargeable AA battery with about 1.195V
Using the dual DMM method (low pass filter) for Pin, i found it to be around 40mW
But now how to measure the Pout correctly, that's the question.
I got some suggestions to dump the output (over the Ro) into a cap, but still then, how to measure the Power?
The csrout is floating, so no way to measure there with my scope.
Anyway, i am open for suggestions and meanwhile will try to find some way.
(i am thinking to separate the output pulses directly taken from the collector via a 1:1 transformer/toroid,
use a FWBR and a cap to create DC, then put a load on it (bulp), and measure there the voltage/current).
I video of this attempt is to be found here: http://www.youtube.com/watch?v=vV4xMeZ_41Y
Regards Itsu
Your replication is much appreciated, Itsu. I understand the difficulty in measuring Pin and Pout...
I see the .99 approach to measure the input power Pin by measuring the current across CSRin and multiplying by the battery voltage. However, when I look at the INSTANTANEOUS Pin waveform on the Tek 3032, I see that Pin fluctuates around zero, and the MEAN (not RMS) value of the Pin is close to zero. (Same result using my ATTEN scope and looking at the power waveform, integrating by hand over one cycle.) I believe this result, from the Power waveform on the oscilloscope, disagrees with the .99 method.
I would like to see a direct comparison of the two methods for Pin.
Further, when I ran my sj1 circuit using a single AA rechargeable battery overnight, the battery voltage had not dropped measurably the next morning, over nine hours running. So I do not think that the circuit was drawing 40 mW.
So I'm questioning even the measurement method for Pin here.
I would like to find a method using heat-rise in resistors and comparing Pin and Pout more directly, but not sure how to do this either. I'm thinking now of a separate circuit for Pin -- in which the voltage from a battery is MATCHED, divided equally between the sj1 circuit and a simple power-sensing resistor (PSRin). Then I think the power dissipated in this PSRin will match the power dissipated in the test circuit, since the voltages from the battery are EQUAL.
Finally, the output power is determined by looking at the Power dissipated in the output resistor Ro.
Again, thanks for this effort to find understanding via careful measurements.
Here's another method, from a discussion (my portion just now) at OU:
Quote from: Omega_0 on 2011.05.29, 07:38:02
Prof Jones,
You are indeed doing a work of great value and I respect your open-mindedness and understanding. I hope something interesting will come out of this circuit.
If you seriously consider heat measurements then you will need a high end calorimeter. It needs to be scaled up into watts range to be above error margins. The heat is not much in this version but the good thing with calorimetry is that you can leave it running for hours and have a cumulative effect. It is the final measurement for any OU setup.
...
QuoteMy reply: Hmmm... I may have access to a high-end calorimeter. But I'm trying to figure out just how one would use it. Perhaps put the entire device in the calorimeter -- except for the output leg of the circuit (Diode + resistor Ro). The CSR resistors are superfluous in this measurement method and are removed. Measure the heat-rise for this "input" portion of the DUT as total Pin, using the calorimeter. Then place the isolated output leg of the device in the calorimeter and measure the heat-rise separately, as total Pout.
Does this make sense?
This method would require that I obtain use of a high-end calorimeter, of course... And I'm wondering where to put the toroid itself, in the input or output leg? Perhaps that won't make much difference...
So I'm wondering about this approach mentioned earlier as an alternative:
QuoteI'm thinking now of a separate circuit for Pin -- in which the voltage from a battery is MATCHED, divided equally between the sj1 circuit and a simple power-sensing resistor (PSRin). Then I think the power dissipated in this PSRin will match the power dissipated in the test circuit, since the voltages from the battery are EQUAL.
Finally, the output power is determined by looking at the Power dissipated in the output resistor Ro.
Comments welcomed. Note this on a "thermal converter" -- including a Fluke meter using this method: http://en.wikipedia.org/wiki/True_RMS_converter
(Thanks to Omega at OU for calling this to my attention.)
OK -- more pondering. The simplest experimental test I can think of using a calorimeter -- place the ENTIRE circuit in a calorimeter with the only source of energy being a capacitor (say 10F) in place of the battery. The available energy Ein is known from Ein = 1/2 CV**2. Then turn the device on (inside the calorimeter) and let it run. Calculate the total energy OUTPUT Eout using the calorimeter.
n = Eout / Ein.
Professor
Perhaps reconsider this method that I have used successfully:
http://www.overunityresearch.com/index.php?topic=702.0
Quote from: PhysicsProf on 2011.05.29, 06:13:34
Luc:
I don't understand -- where is the 30K uF cap placed? just use words if you can't do a schematic drawing.
1.067 V across the 4700 Ohm Rout is significantly low, compared to my circuit and the replication by Les K.
Hi PhysicsProf
I used your schematic and edited it to my test.
Co is where is the 30K uF cap placed and Ro is the 4700 Ohm load on the cap. I used a 1N4148 Diode at Do as it worked best from all diodes I tested.
Components:
Transistor is 2N2222
Vin is 2.86vdc battery
Rb is 1 Meg pot adjusted to 266K Ohm
Cb is 125pf cap
Lb and Lo are 3.5mH each on Metglas 1" o.d. x 1/2" i.d. toroid core
Do is 1N4148 diode
Ro is 4.7K Ohm 2% 1/2 watt resistor
Co is 30K uf 30vdc Mallory cap
Quote from: PhysicsProf on 2011.05.29, 06:13:34
Luc:
1.067 V across the 4700 Ohm Rout is significantly low, compared to my circuit and the replication by Les K.
@PhysicsProf,
have you or Les K. measured Pout the way I did? If so please point me to where I can see this data.
Relying on scopes only can be misleading!... I don't think anything beats a real load test.
If I'm wrong, then I have had OU two or more years back and could not prove it using this method.
Thanks for sharing
Luc
Thanks for posting your schematic, Luc.
You still have not said how you measure the output current -- there is no CSRout in your circuit, for instance.
With the Co in parallel with Ro as you have it, the Co will charge rapidly at first, then the current will be shared between the Ro and Co, with more and more "output" current going through Ro, probably until Co-voltage reaches some maximum value... a while back I had this arrangement, but don't see how it helps you measure the output current.
So please tell us how you measure Pout -- especially how you measure Iout -- with this variation.
@Ion -- thanks for the URL; will study.
--Steve
Quote from: PhysicsProf on 2011.05.29, 22:28:41
@Ion -- thanks for the URL; will study.
Seems again the problem is measuring...that's always the way to declare that it is not there. It was the same to declare that cold fusion is not there.
I believe here is a simple way to show it is there beyond any doubt:
Forget about scopes.
Use two big caps (electrolytic is OK -- bigs ones are usually of that kind), both can be precharged, but put one at the input, one at the output. No more measuring resistors to reduce energy waste, no more resistors at output.
Then, after a few seconds, remeasure the voltages of both capacitors. Calculate the stored energy by 0.5*C*V^2. Find the change in energy, and compute the n by: (Input Energy)/(Output Energy).
This is simple, this is direct, this should resolve the measuring problem.
Quote from: lanenal on 2011.05.30, 03:14:06
...Use two big caps (electrolytic is OK -- bigs ones are usually of that kind), both can be precharged, but put one at the input, one at the output...
It might be good to precharge the cap at the output -- to make sure in each cycle, the gating diode shuts off.
Quote from: PhysicsProf on 2011.05.29, 22:28:41
Thanks for posting your schematic, Luc.
You still have not said how you measure the output current -- there is no CSRout in your circuit, for instance.
With the Co in parallel with Ro as you have it, the Co will charge rapidly at first, then the current will be shared between the Ro and Co, with more and more "output" current going through Ro, probably until Co-voltage reaches some maximum value... a while back I had this arrangement, but don't see how it helps you measure the output current.
So please tell us how you measure Pout -- especially how you measure Iout -- with this variation.
--Steve
Humm ???
I did say I did not use a scope so why would I have a CSRout in the circuit.
I used an online calculator http://www.sengpielaudio.com/calculator-ohm.htm (http://www.sengpielaudio.com/calculator-ohm.htm) to obtain Pin and Pout. I also posted the shots of the calculators!... so I thought it was clear enough!
Anyways, here's how I did it. I click on the link above C.C and entered the voltage measured across the load resistor and enter the value of the resistor and click calculate. This gives Pout. You can also do V x V / R to get P
For Pin I used the same site but used the lower calculator. Enter input current and voltage and click calculate to get Pin.
Been using this method for a while. Works great ;)
Luc
Hi XN:
I notice your replication is different from Steve's build in a number of ways, such as transistor, toroid core, etc. could those play a role in barring you from observing the same effect as that by him?
lanenal
If there was a gain of 8 times the energy, the conversion to DC would show large overunity in despite of losses.
With random shape signals, probe capacitances or transient signals are enough to falsify the measurements.
After many measurements in DC where errors are unlikely, my conclusion is that there is no overunity. The best efficiency is around 60%.
We face a new "rosemary's circuit": a conventional circuit behaving in accord with all laws of electronics, only the blind faith from observers being able to make it miraculous. It is a pity to see the incongruous announce in pesn.com, contributing to the urban legend of the 'Joule Thief' circuit.
Overunity doesn't need a battery. Nevertheless if an energy source is used, at least the device must be running during a much longer time than the source allows. Overunity claims imply a selfsustaining and duplicable device. If these points are not met, we have the right to shut up and to continue to work.
The lack of deontology of most actors in free energy is very disappointing.
Quote from: lanenal on 2011.05.30, 05:32:11
Hi XN:
I notice your replication is different from Steve's build in a number of ways, such as transistor, toroid core, etc. could those play a role in barring you from observing the same effect as that by him?
lanenal
Excellent point, and I await REPLICATIONS, actual replications. You may be right, XN -- further replications and tests will tell. I have not given up, certainly, and I'm encouraged by the apparent good will by most experimenters in this field.
Your point regarding caps is a good one also, laneal. This weekend I have placed four 10,000 uF caps in parallel, and have managed to get the system to feed back into these caps. There is NO battery in the system, only caps. The voltage across the caps is nearly constant now, dropping very slowly with LED lit and no CSR's --
my problem is that the caps detached from the circuit drop in voltage at a measurably significant rate. This particular system does not appear to have demonstrable OU, but again the leaky caps are a problem.
AS noted earlier (perhaps at OU), I am trying to find caps that do not leak so fast. Any ideas on this would be helpful.
A brief comment on
a possible source of anomalous energy that we know very little about (except for its existence):
"What Is Dark Energy?
More is unknown than is known. We know how much dark energy there is because we know how it affects the Universe's expansion. Other than that, it is a complete mystery. But it is an important mystery. It turns out that roughly 70% of the Universe is dark energy. Dark matter makes up about 25%. The rest - everything on Earth, everything ever observed with all of our instruments, all normal matter - adds up to less than 5% of the Universe. Come to think of it, maybe it shouldn't be called "normal" matter at all, since it is such a small fraction of the Universe. "...
"Another explanation for dark energy is that it is a new kind of dynamical energy fluid or field, something that fills all of space but something whose effect on the expansion of the Universe is the opposite of that of matter and normal energy. Some theorists have named this "quintessence," after the fifth element of the Greek philosophers. But, if quintessence is the answer, we still don't know what it is like, what it interacts with, or why it exists. So the mystery continues. "
Read more: http://science.nasa.gov/astrophysics/focus-areas/what-is-dark-energy/
PS -- I will be gone most of the day today (Memorial day).
Quote from: PhysicsProf on 2011.05.30, 15:19:07
This weekend I have placed four 10,000 uF caps in parallel, and have managed to get the system to feed back into these caps. There is NO battery in the system, only caps. The voltage across the caps is nearly constant now, dropping very slowly with LED lit and no CSR's -- my problem is that the caps detached from the circuit drop in voltage at a measurably significant rate. This particular system does not appear to have demonstrable OU, but again the leaky caps are a problem.
I think we can still make a quicky estimate of our system even with the Caps dynamics. Let's say:
-D - I + O = netD= capacitor leak power, negative denotes losses
I = Input power, negative denotes losses
O = Output power
net = Input power with leaks
(this is conservative estimate because we left out circuit efficient)
If we have ideal Caps and 100% efficient circuit, then -I + O = 0 , then we can say -I + O = net + D for non ideal circuit. If net + D >0 , we have overunity. In another word, we compares the rate of Cap discharge under looped operation to the rate of Caps leak when detached. We can conclude the circuit is OU or not. Of course this is valid on the assumption that Caps leak rate is independent of its commission.
Quote from: GibbsHelmholtz on 2011.05.30, 18:18:32
I think we can still make a quicky estimate of our system even with the Caps dynamics. (snip)
If we have ideal Caps and 100% efficient circuit, then -I + O = 0 , then we can say -I + O = net + D for non ideal circuit. If net + D >0 , we have overunity. In another word, we compares the rate of Cap discharge under looped operation to the rate of Caps leak when detached. We can conclude the circuit is OU or not. Of course this is valid on the assumption that Caps leak rate is independent of its commission.
Right. Limited time today (holiday w/ family) -- but I did a quick test. Four 10K uF caps, to run the sj1 circuit.
By measuring the volts before and after 30 seconds on the caps, I can calculate input power!delta-E = 1/2 C(Vi**2 - Vf**2) , Pin = deltaE / delta-T , 30 seconds. C = 40mF.
Start, Vinitial = 1.385V , Vfinal = 1.255V
So delta-E = 6.8 mJoules.
and Pin = 6.8mj/30sec= 0.23 mW, pls check my math.
EDITED to replace minutes with seconds above.
in reasonable agreement with the Tek-scope measurement under similar conditions, perhaps... (see reply #1 for the data, Pin on the left).
Again, the Tek3032 is distant from here, so I can't do the two measurements within minutes, but I think this tends to verify the scope method.
I would ask Itsu to do the same thing on his sj1 circuit, and compare with the dual-DMM method. I like to check things out.
Quote from: PhysicsProf on 2011.05.30, 18:51:24
Right. Limited time today (holiday w/ family) -- but I did a quick test. Four 10K uF caps, to run the sj1 circuit. By measuring the volts before and after 30 seconds on the caps, I can calculate input power!
delta-E = 1/2 C(Vi**2 - Vf**2) , Pin = deltaE / delta-T , 30 seconds. C = 40mF.
Start, Vinitial = 1.385V , Vfinal = 1.255V
So delta-E = 6.8 mJoules.
and Pin = 6.8/o.5min = 13.6 mW, pls check my math.
in reasonable agreement with the Tek-scope measurement under similar conditions, 10 mW (see reply #1 for the data, Pin on the left).
Again, the Tek3032 is distant from here, so I can't do the two measurements within minutes, but I think this tends to verify the scope method.
I would ask Itsu to do the same thing on his sj1 circuit, and compare with the dual-DMM method. I like to check things out.
This is exciting, professor! We should definitely do it this way -- just two caps and two DMMs! This would lower the bar so that the majority of us can replicate what you have done. Please give exact specifications of your components and circuit without any suggestions for substitutions/alternatives in a single post once this simple and self-evident method of measuring corroborates your results.
lanenal
Thanks, laneal... working on it. However -- I need to note that I'm going on travel Thurs am, so I won't have much lab time for about ten days... sigh...
Over at OU, Nul-pts and Xee2 caught an error I made in the math above -- I asked that someone check my math as I was rushing out the door, and they did.
W= J/s of course, not J/min.
And this means that the power input is really close to zero, at
Pin = 6.87J/30s = 0.23mW[/quote]
The mean power is close to zero as seen on both the Tek 3230 and my little ATTEN, as I've been saying.... often bouncing around zero on this circuit.
Now -- the LED on the output leg still lights up, though dimly... at 0.23mW input power... hmmm...
I've repeated the measurement now several times at voltages between approx. 1.2 and 2 Vin from the cap...
Always the input power is in this small range, around 0.2 - .3 mW input power...
There are three of us now in my small town working on this little circuit! Bob is planning to increase Cb, slow the thing down, and see what happens. All three of us have our own scopes, so this is getting fun...
BUT... it is still just "evidence for" OU at this stage, not a proof yet. I hope that is clear.
(PS -- I helped my expecting daughter most of the day, that's why I was slow in responding. She's preparing for the baby coming soon! proud grand-pa here...)
Quote from: PhysicsProf on 2011.05.31, 04:20:17
Thanks, laneal... working on it. However -- I need to note that I'm going on travel Thurs am, so I won't have much lab time for about ten days... sigh...
Over at OU, Nul-pts and Xee2 caught an error I made in the math above -- I asked that someone check my math as I was rushing out the door, and they did.
W= J/s of course, not J/min.
And this means that the power input is really close to zero, at
Pin = 6.87J/30s = 0.23mW
The mean power is close to zero as seen on both the Tek 3230 and my little ATTEN, as I've been saying.
Now -- the LED on the output leg still lights up, though dimly... at 0.23mW input power... hmmm...
I've repeated the measurement now several times at voltages between approx. 1.2 and 2 Vin from the cap...
Always the input power is in this small range, around 0.2 - .3 mW input power...
There are three of us now in my small town working on this little circuit! Bob is planning to increase Cb, slow the thing down, and see what happens. All three of us have our own scopes, so this is getting fun...
BUT... it is still just "evidence for" OU at this stage, not a proof yet. I hope that is clear.
(PS -- I helped my expecting daughter most of the day, that's why I was slow in responding. She's preparing for the baby coming soon! proud grand-pa here...)
Oh no, professor, 10 days! ... I can't just wait so I will probably have my own build...but have a nice trip! With so much excitement going on, life is good! I'm glad for you, professor.
lanenal
Hey, that's great laneal -- so glad to hear you're going to do a replication.
So I think we have a straightforward way to measure the input power Pin without an oscilloscope, using a cap and a stopwatch.
Measuring Pout will be more difficult.
On the output leg of the circuit, the voltage shows large swings, typically 12 V or so Vpp. One could put a rectifier FWBR in this output leg, then charge a cap... As long as that did not adversely affect the circuit performance.
I would replace the LED with a diode in the same direction, so that less power is dumped on the diode-LED, and more on the output Capacitor... Wish I had more time for this, but I've a long-planned road trip coming up Thursday, so away from my home lab for about ten days at that point... sigh... missing out on some of the fun here...
Quote from: lanenal on 2011.05.30, 05:32:11
Hi XN:
I notice your replication is different from Steve's build in a number of ways, such as transistor, toroid core, etc. could those play a role in barring you from observing the same effect as that by him?
lanenal
You are right, lanenal. I confirm that if you have not the exact ferrite toroid of Physicsprof, the same x ohms resistances at less than 0.00001% margin of error compared to his, the same oscilloscope with probes of same capacitance, the same battery of x V within 0.00001% error, the same transistor of the same lot, the same capacitors of x pF, you can't achieve overunity. It is a general problem in free energy and a very big difference with conventional science: without departing from the spirit of the invention, nobody having exactly the same components as the inventors, nobody is able to duplicate the miraculous machines. Too bad! ;)
Dear Professor
Measuring power out (into a load resistor or LED or combination thereof) vs. power in misses a part of the picture. How do you account for energy that may arise or be absorbed in the form of heat in the ferrite core or other device that is part of the system. You will miss it.
You can carefully measure the temperature rise of each component and try to sum these heat gains or losses.
I chose the dual chamber thermal balance bridge method in order to account for all gains and losses using a null balance technique against a known power input and heat source in a second test chamber.
This allows real time steady state monitoring and logging of real power in and out of the DUT.
As we have said before, other types of energy loss may be difficult to account for, such as radiated RF, but the test setup can be modified with absorbers to turn the radiated RF into heat energy, such that it can be measured as part of the black box I/O.
Quote from: ION on 2011.05.31, 12:26:16
Dear Professor
Measuring power out (into a load resistor or LED or combination thereof) vs. power in misses a part of the picture. How do you account for energy that may arise or be absorbed in the form of heat in the ferrite core or other device that is part of the system. You will miss it.
You can carefully measure the temperature rise of each component and try to sum these heat gains or losses.
I chose the dual chamber thermal balance bridge method in order to account for all gains and losses using a null balance technique against a known power input and heat source in a second test chamber.
This allows real time steady state monitoring and logging of real power in and out of the DUT.
As we have said before, other types of energy loss may be difficult to account for, such as radiated RF, but the test setup can be modified with absorbers to turn the radiated RF into heat energy, such that it can be measured as part of the black box I/O.
ION -- I would like to understand your method better, but found your other thread (which was closed IIRC) to be hard to understand. Since you are here, could you explain the method more thoroughly? Is it basically a calorimetric method as it appears?
I have suggested above in this thread a straightforward calorimetric test along these lines:
1. The initial energy is provided by a capacitor charged to a measured voltage, Vcap-start.
Ecap-start = 1/2 *C* Vin**2
2. The ENTIRE device is placed inside a precision calorimeter, with just one or two thin wires to permit "turning on" the device.
I have not worked out how this is done in detail; it could simply be that a switch is turned to on with a single thin wire.
3. From runs down previously, it is know how long it takes for the cap to go from Vin to a lower value (above the cut-off point for the DUT). For example, one could go from Vin = 2.5 V to Vstop = 2 V in 10 seconds, turning off the switch when the device is inside the calorimeter at 10 seconds. (The ending voltage Vstop is measured soon after the run.)
4. Ein is simply Ecap-start minus Ecap-stop.
5. Eout from the entire system is determined by the calorimeter.
6. The efficiency is simply n = Eout / Ein.
There is a concern that the operation of the device may be affected by the shielding of the calorimeter, but I think this approach is worth a try.
Do you see any problems with this method? And -- is it similar to your method?
@exnihiloest -- I somehow missed the detail on how you measured Eout or Pout. Pin is relatively straightforward -- but could you delineate just how you determine Pout or Eout? Thanks in advance.
Hi PhysicsProf,
i saw your latest input info with the 30 sec. cap test.
You say 0.23mW as in 230uW? That is extremely low in my opinion, especially when you say that the output p2p is around 11V (which is the same as with me), and with the led dimly on.
My latest attempt using the dual DDM shows 13,7mA at 1.37 V which means 18.7mW.
When trying to use 4 x 6.3V/10KuF caps (40mF) as the source, it drained them within 10 seconds.
I will try with 10 x 10KuF (0.1F) later tonight and make a video of it, but don't think it will last for 30 seconds.
Using your cap math on my values (startV 1,54, endingV 0,768V, time 5 seconds, C=40mF), the Pin comes at 7.127mW, but can hardly be considered trustworthy this way.
Anyway, have a nice time traveling,
regards Itsu
Quote from: Itsu on 2011.05.31, 18:54:04
Hi PhysicsProf,
i saw your latest input info with the 30 sec. cap test.
You say 0.23mW as in 230uW? That is extremely low in my opinion, especially when you say that the output p2p is around 11V (which is the same as with me), and with the led dimly on.
My latest attempt using the dual DDM shows 13,7mA at 1.37 V which means 18.7mW.
When trying to use 4 x 6.3V/10KuF caps (40mF) as the source, it drained them within 10 seconds.
I will try with 10 x 10KuF (0.1F) later tonight and make a video of it, but don't think it will last for 30 seconds.
Using your cap math on my values (startV 1,54, endingV 0,768V, time 5 seconds, C=40mF), the Pin comes at 7.127mW, but can hardly be considered trustworthy this way.
Anyway, have a nice time traveling,
regards Itsu
Yes, as in less than a milliWatt. Something is different, obviously, in our two systems. Do you have variable resistors Ro and Rr, that you vary to minimize Pin (using the scope)? I look at the power waveform, which oscillates around zero, and try to get the above-zero spikes to equal (in area) the below-zero spikes. Of course, for this one needs the MATH function, V(t)*I(t).
What is the resistance in your toroidal windings? Do you use a red LED? IIRC, you used white -- that may make a difference. The red is dim under these conditions, but visible.
Also, would you pls check using the cap/stopwatch method the Pin -- and compare with the dual-DMM method? that's an important test you are in a position to do.
Itsu -- I just watched your latest vid: http://www.youtube.com/watch?v=P7x2Y0gdXWQ&feature=uploademail
Quite well done. Thank you for this research effort.
You are using about 2Kohms for Rb, whereas I'm using 51Kohms for Rb, in the test I did yesterday (and see my post #1). Pls try with about 51kOhms, and a red LED, would you? could bring us more into line.
QuoteRb 51K
MPS2222
C-B 151 pF
D red LED
L-B, L-O bifilar 9turns, ferrite toroid 1"OD, ~90uH each
-- PhysicsProf, post#1
Notes:
1. Yesterday, the dual - DMM method of .99 gave 34mW, today 18.7mW -- and you noted you had NOT changed the circuit.
Strange... Does the DMM method give variable results? or is it the circuit which changes?
2. You did the Cap/stopwatch method and have some results -- interesting. I attach a screen-shot of your results.
You say that the highest value is most in line with the DMM method, 12.2 mW. Yes, but this still does not appear to agree with the dual-MM method, 18.7mW or 34mW.
Conclusion: dual DMM-method vis-a-vis the cap/time method needs further checking.
Thanks again, Itsu.
Great, you found the video allready.
Yes, i agree, we have some differences between the 2 circuits, and indeed the dual DMM method creates different results all the time.
I understood you used with your latest circuit the 2K Ohm Rb resistor, but probably i misunderstood.
I will try to match the components as close as possible, starting with a new coil, as mine now has an inductance of 930mH each (12 turns) .
But i think this Hartley oscillator is doing what it suppose to do; oscillate....broadband
Regards Itsu.
Quote from: Itsu on 2011.05.31, 23:32:00
Great, you found the video allready.
Yes, i agree, we have some differences between the 2 circuits, and indeed the dual DMM method creates different results all the time.
I understood you used with your latest circuit the 2K Ohm Rb resistor, but probably i misunderstood.
I will try to match the components as close as possible, starting with a new coil, as mine now has an inductance of 930mH each (12 turns) .
But i think this Hartley oscillator is doing what it suppose to do; oscillate....broadband
Regards Itsu.
" indeed the dual DMM method creates different results all the time." This is significant! please provide a few more examples of how this varies, would you? And then i hope for some comment from .99 ;)
Thanks so much, Itsu. I have used Rb of 51Kohms and sometimes 2kohms, and I try to indicate when I use a particular value; apologize for the confusion.
Quote
inductance of 930mH each (12 turns) .
Pls check that -- seems very high. Mine are about 90microH each.
Itsu -- note that Hoppy at OU has also gone to a higher Rb and gets 4 mW:
QuoteAnyway, with 56K I get a sinusoidal waveform scoped emitter to ground and the in / out power levels measured across 1R shunt resistors (without the additional 3R) appear to be fairly closely matched on my scope at around 4mW... Hoppy
So pls let us know what you get. I've asked him just how he measures 4mW...
PhysicsProf,
OK, i will keep on experimenting, but i am on my work right now, so that will be later today.
Concerning the: " indeed the dual DMM method creates different results all the time." , what i meant is that it probably is not the dual DDM method itself (i think its accurate), but rather
the nature of this circuit (oscillator) which keeps on influencing the components (capacitive feedback) and creating different oscillations all the time (and current pull).
Concerning the inductance of the coils, i think yours are very low (90microH) :)
I wound another one (smaller od, 8 windings each, so more in line with yours), and measured it to be 330mH (1 third of my former one).
Anyway, i am following the severall forums, so i will keep up with the results others are presenting.
Regards Itsu
Professor
First a little background on the single chamber test method.
The idea behind the single chamber method is that the DUT is placed into a thermal chamber with some fixed losses to ambient. A heat rise will be generated within that chamber that will eventually stabilize at some temperature above ambient.
Power input in the steady state is noted, and the test is repeated without the DUT in the chamber, except in this case we try to obtain the same rise over ambient using a resistive heater and manual adjustment of the power supply, again noting the final power input to obtain the identical heat rise over ambient.
If there is no anomalous heat gain in the DUT, similar results should be obtained, and power input should be the same. This method requires two steps and has been so named the two step method.
The dual chamber thermal method is a refinement and automation of the single chamber test method shown below. In this case the entire process is automated using a PID control algorithm to "servo" the control chamber to the same temperature rise as the DUT.
As long as the test chambers are nearly identical, with roughly the same fixed losses to ambient, we do not need to care what those fixed losses are, as we are attempting to obtain a "null" reading. We can indeed "characterize" the test chambers using fixed heaters in each to determine the limits of accuracy of the test.
Any device that we are testing will have many heat sources to consider. The thermal chamber method sums up all the heat losses, which should be equal to power input. If the final heat rise over ambient seems to be greater than power input in the control chamber, we can begin to look for a source of heat gain.
In the dual chamber method shown below, the DUT and load resistor should be placed in the chamber. Thermocouples should not be taped to the resistors, but placed at the top of the chambers. This diagram was made for a different purpose, where open air testing of load resistors was being compared, but you should get the idea.
I'm sorry if this is not clear. I have many years of experience using these methods so I tend leave a lot of holes in my explanations. Think of it as a balance scale in that that we are attempting to find the unknown quantity by comparing it to a known quantity.
PhysicsProf (happy traveling),
concerning the the 30 second measurement i did, here are again the figures:
Vstart V5sec V15sec V30sec
1.537 1.067 0.636 0.522 (volts)
Pstart P5sec P15sec P30sec
???? 12.24 6.5 3.48 (mW)
When i make a graph of these, is it not fair to extrapolate the unknown Pstart the way i did below, meaning that the purple part points to a Pstart of 18mW, which
is in line with the dual DMM measurement i did on this same circuit (18.7mW)?
If you agree, than this again means to me that the dual DMM method is accurate for DC, even with high AC/RF pollution .
Regards itsu
Itsu, actually I find that the higher Vin, the faster the cap drains, and the LED is brighter also.
So you need a capacitor that will drain slowly, but not leak significantly when disconnected -- measuring over a smaller V drop on the cap. And compare THAT Power result with the DMM measurement!
Today, I tried my own replication of the DUT and used a 10F cap to do the measurement:
Over 6.0 minutes, 2.39 to 2.34 volts => 3.2 mW +/- due to the small V change.
Over 40 minutes, 1.661 V to 1.490 V => 1.1 mW.
(someone check my math?)
I just report the results as I see 'em. Rb again @ 51Kohms -- are you doing this, Itsu??
I note also that when the cap is first charged, one has to wait for it to "settle down" -- I did this.
This is a simple test IMO, to see if YOUR replication is in the same ballpark as mine.
Note: I have no opportunity to test this replication with the Tek 3032, until I return from my trip.
On Terminology.
"Overunity" carries baggage, connoting energy out of nowhere, implying a violation of the laws of Physics.
Don't believe in violating laws of physics at all, so I try to avoid the term OU -- and suggest instead "super-efficiency", meaning
n = (electrical power out) / (electrical power in) > 1.
n > 1, super-efficiency -- allowing for anomalous energy input to the device. (And I favor "dark energy" personally -- 70% of all the mass-energy in the universe is this little-understood stuff, as I posted earlier.)
I avoid "COP" in favor of "n" -- same reason, baggage carried with the term COP >1 going with violating laws of physics.
And finally, I avoid "free energy" because it also connotes now energy out of nowhere (google it).
IMHO, something like Novel Electrodynamic Energy ( or simply, novel energy) would be a much better term.
All, IMHO.
Of course, we have to prove a device exists with n>1 before this is really useful! ;)
TERMinology in a nascent field of science is critically important.
I went shopping today for a 3 Ohm Potentiometer, but only got lucky with a 25 Ohm Rheostat (took it anyway). Did some quick calculation and realized how important/crucial it is to find correct POT for precision tuning. Let's say you need to tune your POT to +/- .1 Ohm to get what you want. Let's also say that an average person can touch a pot and tuned it 1/50 of the whole turn.
Having a 3 Ohm pot will give 3/50 = .06 or +/- .03 Ohm
.1/.03 = 3.3 or 330% accuracy.
However, having a 25 Ohm pot will give 330/8.3 = 39% accuracy! (since 25Ohm / 3Ohm = 8.3). Before I used a 5K pot to tune it, so the precision is 330/1666 = .19% accuracy! That means if I touch it 500 times, I should get it right. lol
I realized this is crucial because during my calculation I came up with a tuning characterized by L/i where L is the coil inductance and i is the input current. If the coil is 250 microH and I have a 25 Ohm POT with 1.5V input. The current I can play with is 1.5/25 = .06 amps . .06/ 250 micro = 240 . It means for each unit of current I adjust, the sensitivity is 240 times. By having a 3 Ohm POT, the sensitivity can be reduced to 30. Sorry if you understand this already, just my share for those who in the same boat.
My 1/50 dollar.
Hi Professor,
I'm just curious: have you ever successfully replicated it yourself? XN says that OU is simply not duplicable because it was an extreme coincidence. Therefore, I wonder if you could replicate it by yourself, having two independent sets of components assembled that show similar results?
lanenal
GibbsH
You could put a resistor in parallel with the rheostat to bring it's resistive range down
Quote from: Peterae on 2011.06.02, 10:55:08
GibbsH
You could put a resistor in parallel with the rheostat to bring it's resistive range down
Oh nice. Thank you for your expertise Peterae.
Rtotal = RtuneRparallel / (Rtune + R parallel) . Good tune!
GH
Thanks @all for comments.
Quote from: lanenal on 2011.06.02, 10:02:12
Hi Professor,
I'm just curious: have you ever successfully replicated it yourself? XN says that OU is simply not duplicable because it was an extreme coincidence. Therefore, I wonder if you could replicate it by yourself, having two independent sets of components assembled that show similar results?
lanenal
Yes, I have built my own replication, and as noted above:
QuoteToday, I tried my own replication of the DUT and used a 10F cap to do the measurement:
...
Over 40 minutes, 1.661 V to 1.490 V => 1.1 mW.
...
This is a simple test IMO, to see if YOUR replication is in the same ballpark as mine.
Note: I have no opportunity to test this replication with the Tek 3032, until I return from my trip.
XN may be correct, don't know yet.
My first test showed Pin of only 0.23 mW as I noted earlier, using the cap and stopwatch method.
I think this rather simple test of Pin allows a first test, to check out other builds, to see whether they are in the same ballpark.
laneal -- was it you who had the information/specs on the 1"OD, 0.5"ID, 7/16" high, electronic goldmine G6683 toroid that I used?
Can you give what information you have on this toroid?
Thanks -- soon to hit the road again! sorry no computer contact for a while.
Quote from: GibbsHelmholtz on 2011.06.02, 12:10:07
Oh nice. Thank you for your expertise Peterae.
Rtotal = RtuneRparallel / (Rtune + R parallel) . Good tune!
GH
Please be aware that connecting an external resistor to the outside terminals of a potentiometer or rheostat does not do exactly what folks may think. The variable resistance will not change as expected.
On larger variable resistance values this can be a real problem.
PhysicsProf,
About concerns of not seeing EM energy using calorimetry.... You can enclose the DUT within a small, double layer Faraday cage (with no electrical connection to the outside or between layers). The purpose is to provide an opportunity for EM energy to be converted to heat for measurement. Double-layer because the second layer minimizes capacitive coupling to the measurement container.
Hi all,
i replicated this circuit as close to the original as possible now, and made again some measurements.
The diagram is as can be seen below, and the video is to be seen here: http://www.youtube.com/watch?v=rJttV_kzTFY
My input current measured with the dual DDM method is:
with Rb=50K is 2.2mA, and with a Vbat=1.31V this gives a Pin of 2.52mW
with Rb=100K is 1mA, and with a Vbat=1.31V this gives a Pin of 1.31mW
Using the Cap bank (40mF) as source, the 30sec calculations gives Pin of 0.66mW, but when calculating after 20sec, this gives a Pin of 0.79mW.
I guess this should give both the same Pin, right? Or does the depletion of the voltage across the caps influences the current being drawn?
If so, then this cap math method is only accurate when very little current is drawn, and the voltage keeps kind of stable over the 30s period.
My feeling says, and experiments show, that the dual DDM method is more reliable in this circuit.
I will now concentrate on the output power, but still have not found a reliable method for measuring that.
User TinselKoala proposed to calibrate the used led with a phototransistor or so, and a known current or voltage source, and then measure the led output again with this phototx in this circuit.
But as can be seen, this (red) led hardly is lit, so i don't think this will work, but i will see what i can do.
Regards Itsu
Quote from: Itsu on 2011.06.02, 23:27:48
Hi all,
i replicated this circuit as close to the original as possible now, and made again some measurements.
The diagram is as can be seen below, and the video is to be seen here: http://www.youtube.com/watch?v=rJttV_kzTFY
My input current measured with the dual DDM method is:
with Rb=50K is 2.2mA, and with a Vbat=1.31V this gives a Pin of 2.52mW
with Rb=100K is 1mA, and with a Vbat=1.31V this gives a Pin of 1.31mW
Using the Cap bank (40mF) as source, the 30sec calculations gives Pin of 0.66mW, but when calculating after 20sec, this gives a Pin of 0.79mW.
I guess this should give both the same Pin, right? Or does the depletion of the voltage across the caps influences the current being drawn?
If so, then this cap math method is only accurate when very little current is drawn, and the voltage keeps kind of stable over the 30s period.
My feeling says, and experiments show, that the dual DDM method is more reliable in this circuit.
I will now concentrate on the output power, but still have not found a reliable method for measuring that.
User TinselKoala proposed to calibrate the used led with a phototransistor or so, and a known current or voltage source, and then measure the led output again with this phototx in this circuit.
But as can be seen, this (red) led hardly is lit, so i don't think this will work, but i will see what i can do.
Regards Itsu
Hi Itsu,
Giving the about the same current draws between the Caps and Battery, the result should be similar between the 2 methods. I know the voltage drop a bit in the cap but that should not change the result by a factor of 2-3 in my opinion.
I think the observation of more power in higher voltage Cap is this:
q=CV , dq = CdV . I think the circuit draw about the same current at high and low voltage, that means dq = dV for both cases. But this, say at 10-9V , the energy is 100-81=19, at say 5-4V, the energy is 25-16 = 9 .
WaveWatcher,
Thanks
Quote from: PhysicsProf on 2011.06.02, 13:55:31
laneal -- was it you who had the information/specs on the 1"OD, 0.5"ID, 7/16" high, electronic goldmine G6683 toroid that I used?
Can you give what information you have on this toroid?
Thanks -- soon to hit the road again! sorry no computer contact for a while.
Here is what I have got from their catalog
(http://www.goldmine-elec.com/pdf/243_pgs30-37.pdf):
These are prime ferrite cores made by Nippon Ferrite LTD. Size is 1"
outer diameter x 5/16" thick with an inside diameter of 1/2"
G6683 5 / $ 1 . 0 0
In the other post I also have its Al Value, which is 10,000 (nH/t^2).
That means, if you had T turns on the toroid, you are supposed to
get about 10,000 X T^2 nH of inductance.
For example, when you got 9 turns on it, you are supposed to get
10,000 X 81 nH = 810uH of inductance. But what you get is not
that high (around 90uH as you said), that's probably because you
are using wires with thick encapsulation, causing good amount of
leaking. If you use magnetic wire, the inductance will be much closer.
lanenal
Quote from: PhysicsProf on 2011.06.02, 13:35:46
Thanks @all for comments.
Yes, I have built my own replication, and as noted above:
Note: I have no opportunity to test this replication with the Tek 3032, until I return from my trip.
Thanks for the confirmation. Hopefully I can manage to make some progress by then.
lanenal
Ok Gibbs, thanks for your insights.
So the very second the cap voltage is decreasing, the energy in the system is decreased, and therefor the current drawn (and input power) is decreased.
So for this Cap math method to be accurate, one should take regular voltage measurements say like every 2 seconds, put that in a graph, and then extrapolate the unknown starting power (Pin) like i did in my post #101
Is that fair to say?
Regards Itsu
Quote from: Itsu on 2011.06.03, 10:06:10
Ok Gibbs, thanks for your insights.
So the very second the cap voltage is decreasing, the energy in the system is decreased, and therefor the current drawn (and input power) is decreased.
So for this Cap math method to be accurate, one should take regular voltage measurements say like every 2 seconds, put that in a graph, and then extrapolate the unknown starting power (Pin) like i did in my post #101
Is that fair to say?
Regards Itsu
Well, the first test you did with the graph showing exponential decay in voltage suggested that the voltage and current decreased over time. I think if you tuned your circuit good and have a large cap like the second test, voltage should drop linear (implies same current draw). During your test, did you observed current drop? Like 2 mA to 1ma ? What I was saying is the cap voltage decrease, the energy decrease, the current is the same.
I also think that you did accurately measured input for both methods, but with a factor of 2-3 off like that. One method is wrong.
PS: are you the one that paint a dog biting a man's hand on youtube?
Quote
Well, the first test you did with the graph showing exponential decay in voltage suggested that the voltage and current decreased over time. I think if you tuned your circuit good and have a large cap like the second test, voltage should drop linear (implies same current draw). During your test, did you observed current drop? Like 2 mA to 1ma ? What I was saying is the cap voltage decrease, the energy decrease, the current is the same.
I also think that you did accurately measured input for both methods, but with a factor of 2-3 off like that. One method is wrong.
PS: are you the one that paint a dog biting a man's hand on youtube?
Ok, Gibbs,
i do see also the current drop in a similar way as the voltage, see this video at the 11:08 mark:
http://www.youtube.com/watch?v=P7x2Y0gdXWQ
where i just hooked up the circuit to the Cap bank
11:08 V = 1.314V and decreasing fast
i = 11.3mA and decreasing fast.
So my money still is on the dual DDM method for most accurate input method.
And no, i am not the one that painted the dog biting a man's hand.
Regards Itsu
Quote from: Itsu on 2011.06.03, 14:31:04
Ok, Gibbs,
i do see also the current drop in a similar way as the voltage, see this video at the 11:08 mark:
http://www.youtube.com/watch?v=P7x2Y0gdXWQ
where i just hooked up the circuit to the Cap bank
11:08 V = 1.314V and decreasing fast
i = 11.3mA and decreasing fast.
So my money still is on the dual DDM method for most accurate input method.
And no, i am not the one that painted the dog biting a man's hand.
Regards Itsu
I see that the circuit draw down the Cap bank quickly. I also see that for the first 5 seconds, the power is the same as the dual DMM method. So at least there is an agreement on both methods. But your tuned circuit does not agree. I see if you use 5 seconds on your tuned circuit, the power is still low (if 30 seconds gives .66mW, 20 sec gives .79mW, then 5 second would be around 1mW). 1mW is still lower than dual meter method, not sure what Rb is when you do this. So we should investigate why both methods agree on low efficient circuit and disagree on high efficient circuit.
Quote from: GibbsHelmholtz on 2011.06.03, 20:14:21
I see that the circuit draw down the Cap bank quickly. I also see that for the first 5 seconds, the power is the same as the dual DMM method. So at least there is an agreement on both methods. But your tuned circuit does not agree. I see if you use 5 seconds on your tuned circuit, the power is still low (if 30 seconds gives .66mW, 20 sec gives .79mW, then 5 second would be around 1mW). 1mW is still lower than dual meter method, not sure what Rb is when you do this. So we should investigate why both methods agree on low efficient circuit and disagree on high efficient circuit.
You make an important point, Gibbs: "
So we should investigate why both methods agree on low efficient circuit and disagree on high efficient circuit."
Still traveling, home late tonight if all goes well. Then back to research on this circuit tomorrow.
I'm back from the trip to California where I spoke to colleagues. A brief summary of where we stand on the blocking oscillator OU? device seems in order.
1. After months of development and learning, the initial device was tested by me using a Tektronix 3032 scope to measure I(t) as voltage drop across a 1-ohm resistor and V(t) and the instantaneous Power was displayed as the product: P(t) = V(t) * I(t). Then the Tek3032 was used to calculate the MEAN power from this waveform, over numerous cycles. Finally, I calculated n = Poutput / Pinput and found n ~ 8 for this circuit, by this method.
My early measurements involved hand-integration of the energy of the power waveforms for Pin and Pout, P(t) = V(t) * I(t), for one cycle, and these integrations yielding Ein and Eout also showed n>1 (back in the Feb-March 2011 time frame).
2. The same procedure using the Tek3032 was followed for an "exact-as-possible" replication of the circuit by Les Kraut, which showed again n ~ 8. At this point, I noted that we had "evidence for" (NOT "proof of") OU and shared the straightforward circuit design publicly, inviting any who wanted to test/develop the circuit to do so. It was the success of the replication and pushing by Sterling Allan that induced me to release the development publicly -- to those willing to build and test the device. I am a strong supporter of open source development of alternate-energy devices.
3. In both cases, mine and Les', the input power was tuned (using especially the variable resistors in the circuit) to be close to zero.
4. The low value of the input power was checked using the input-capacitor + stop-watch method, Ein = 1/2 CV**2 and Pin = Ein /time, and the values came out: 0.23 mW for my initial circuit and 1.1 mW for Les Kraut's replication, with the output LED dimly but visibly lit in both cases.
5. I urged replicators to assure that the Pinput was in this low range with the output LED lit, as a first test of whether the replication was in the same ballpark as our DUT's.
6. Chris built a replication then several versions, measuring n>1 but by a different method which was challenged... Chris found that sometimes the circuit would stop producing n>1 (by his measurement), and he worked on the stability of the circuit. He is attempting to build a self-running version as am I.
7. A few others built or are building replications, but again the power-measurement is a difficult issue, especially for the output power.
8. We discussed various power or energy measurement methods to check/complement the math-mean method using an advanced scope to evaluate power, including use of capacitors and use of a calorimeter. But the self-running system would be the most compelling (in my opinion and that of others).
9. It was noted that Russian work shows an apparently similar long-running blocking-oscillator circuit, again all solid state, but the Russian is hard to read and how long it runs is not yet understood.
Quote from: gyulasun on 2011.06.08, 17:44:50
Thanks for this link too and if you happen to figure out info on the runtime, please mention it here.
Gyula
Agreed -- again the link to the Russian blocking oscillator is here: http://freeenergylt.narod2.ru/vladimir_pantiuhov/ I have been to Russia for conferences on fusion energy a couple of times, but I do not read enough Russian to be of help in understanding Pantiuhov's work. If someone could post the "best" schematic from this Russian research, it would be appreciated.
10. As for myself, I'm planning to work on the circuit using first the method of capacitor-in and output-capacitors -- in order to better understand energy-flow in this circuit. Then I will proceed with an effort to build a self-runner. This may take several days or even weeks -- patience requested.
(Posted also at OU.com, where Chris has posted his development of this circuit, along with others.)
Thanks for the summary Professor.
:)
.99
Professor,
Glad you are back to work. This is a good summary!
I'd like to propose a self-runner schematic, which is quite simple and direct I think. Instead of a bifiler toroid, here we will need a tri-filer toroid. A big Cap Cs is used to collect the output and power the circuit (a Zener's Diode is used to keep it from running away -- this is possible when n is big enough). The battery should be switched off when the Cap is fully charged. The breakdown voltage of diode Z should be a little higher than the battery voltage.
lanenal
I have been testing these types of circuit for a couple of years now.
The closest I ever did come to COP=1 (circuit was COP<1) is the
attached mosfet version. However, I did find that the circuit was an
excellent light dimmer for 12 Volt light bulbs. The circuit itself do
not heat up and all the input energy goes to the light bulb.
GL.
Here is a simulation of the self-run circuit. Of course, it did not turn off the battery. Just a proof of concept.
Here I replaced the Schottkey diode with a scilicon diode.
...
Quote from: lanenal on 2011.06.10, 04:32:04
Here is a simulation of the self-run circuit. Of course, it did not turn off the battery. Just a proof of concept.
Good job Lane. Looks like you have the frequency about right.
Have you considered trying the sim with a non-linear core rather than the simple linear coupling? How close are those waveforms to what the Professor is seeing?
.99
Quote from: poynt99 on 2011.06.10, 13:34:50
Good job Lane. Looks like you have the frequency about right.
Have you considered trying the sim with a non-linear core rather than the simple linear coupling? How close are those waveforms to what the Professor is seeing?
.99
Thanks .99. This is not exactly like his model, but close, and the frequency in the simulation is of the same magnitude. I'm not sure how to work with a non-linear core yet, but that's a good idea! I'll see if there is anything available. I'll also try to replicate his model too. Just need a 151p cap and some battery...
Interesting idea, laneal. Evidently you are adding the third winding to use that to extract "output" energy from the system. Not a bad idea, but note that this will certainly change the dynamics of the circuit. Certainly worth a try.
Let me emphasize a curiosity in the original schematic (attached again for convenience) -- which is that point 6 is the "low point" in the potential, as measured by a DMM and diode Do in the direction shown. Right now, I'm running with a capacitor in place of Ro, with the NEGATIVE terminal of the cap at point 6, as shown in the diagram. This charges the cap quite rapidly under good conditions to about 6 volts, with 2.5 volts input.
Quote from: PhysicsProf on 2011.06.10, 18:23:44
Interesting idea, laneal. Evidently you are adding the third winding to use that to extract "output" energy from the system. Not a bad idea, but note that this will certainly change the dynamics of the circuit. Certainly worth a try.
Let me emphasize a curiosity in the original schematic (attached again for convenience) -- which is that point 6 is the "low point" in the potential, as measured by a DMM and diode Do in the direction shown. Right now, I'm running with a capacitor in place of Ro, with the NEGATIVE terminal of the cap at point 6, as shown in the diagram. This charges the cap quite rapidly under good conditions to about 6 volts, with 2.5 volts input.
Thanks professor. Could you maybe give an exact specification for replication when you are more or less certain of the best values you have got in your experiments? There are many changes here and there now and I am not sure which is the best way to follow, especially when you might have got better configurations.
I agree that that's the right way to put a capacitor. But to loop the device, it is much harder. Using a third coil is one of the simplest way I can think of. You are right that will change the dynamics a bit, but won't be too much -- because the L0 and L1 will be equivalent when it comes to pick up the collapsing magnetic field. With L1, it then becomes easy to loop the device.
OK, I have my replication running...
I found that when R0=10k, the LED is really dimly lit. So I reduced it to R0=1k, with much brighter LED, and the input is almost the same. My input is about 3mA averaged over time, peak is probably 40mA, duty cycle about 10%. The frequency is measured at 1.37MHz. I have no scope, all measurements done with my DMM. My input V = 5Volts. This seems to be a promising replication, next I think I should try to do a dual DMM with caps test to measure input/output energy ratio.
I will keep it running for a day or so, with a DC converted from the grid.
What do you have in mind for a Pout measurement method Lane?
.99
I'd try to charge a capacitor. I've got some 10000uF caps, hopefully will do. Any suggestions?
Quote from: lanenal on 2011.06.11, 10:49:50
OK, I have my replication running...
I found that when R0=10k, the LED is really dimly lit. So I reduced it to R0=1k, with much brighter LED, and the input is almost the same. My input is about 3mA averaged over time, peak is probably 40mA, duty cycle about 10%. The frequency is measured at 1.37MHz. I have no scope, all measurements done with my DMM. My input V = 5Volts. This seems to be a promising replication, next I think I should try to do a dual DMM with caps test to measure input/output energy ratio.
Glad to hear you have a DUT, lanenal. You asked about what conditions I would recommend; to summarize,
Vin 2.5 V
Rb 2Kohms
MPS2222
C-B 151 pF
D red LED
L-B, L-O bifilar 9turns, ferrite toroid 1"OD, ~120uH each (recently remeasured the toroid giving best results)
Ro 9.7 Kohms
Rr 4.1 ohms
Testing the input Power using a Cap and a stopwatch, as described earlier in this thread, and independently using the TEK 3230 method, I find Pin ~ 10-17 mW with this DUT If your Pin is much larger than this, with the conditions given above, then the replication needs to be re-checked for consistency. With Rb ~ 51Kohms, the LED is seen to glow but very dimly, and Pinput drops (as described above).
Chris S. on the OU.com forum has made several tests and mods and his version 7 is shown below; his "best", IIRC. I hope Chris does not mind my citing his work; posted here: http://www.overunity.com/index.php?topic=10773.210
Chris has been very diligent in testing the circuit and making small mods.
Thanks professor, will try your configuration. My configuration is far from yours. I can hardly charge my 1000uF capacitor in place of R0 to above 2V, after 2 minutes of charging with my input at 5V (after a while the red LED won't lit anymore, and I noticed the circuit stops working after a while, because the frequency drops to zero), this shows that there must be something special in your configuration. Have you tried to replace one of your crucial components at a time (Diode, Transistor, Toroid core) and observe significant changes in your measurements?
Quote from: lanenal on 2011.06.12, 14:23:09
Thanks professor, will try your configuration. My configuration is far from yours. I can hardly charge my 1000uF capacitor in place of R0 to above 2V, after 2 minutes of charging with my input at 5V (after a while the red LED won't lit anymore, and I noticed the circuit stops working after a while, because the frequency drops to zero), this shows that there must be something special in your configuration. Have you tried to replace one of your crucial components at a time (Diode, Transistor, Toroid core) and observe significant changes in your measurements?
Clearly, our DUT's are different. I don't know why, but I have some further results to share. I'm working with two capacitors now, 10,000 uF = 10mF each, so a total of 20 mF. Charged to 2.58V using 2 rechargeable AA's. I use the caps to run the device until the voltage drops to 1.5V on the caps and note the time t required using a stop-watch. Thus, I have a reliable measure of Ein and Pin:
E = 1/2 C V**2, P = E/t,
So Pin = 1/2 20mF (2.58V**2 - 1.5V**2) /t = 44.06mJ/t
So 44.06 mJ are put into the system, and dividing this Ein by the time, I derive Pin. The measurements have proven repeatable -- and interesting.
Here are some results this morning:
Basic conditions:
Rb 52Kohms
MPS2222
C-B 151 pF
D red LED
L-B, L-O bifilar 9turns, ferrite toroid 1"OD, ~120uH each
Ro = 1Kohm
Rr= 0, CSRout = 0 (removed)
Oscilloscope ATTEN used to look at the Pin waveform most of the time in my little lab at home.
1. With the above conditions, the caps discharge from 2.58V to 1.5V in t = 37.8 seconds, so Pin = 44.06mJ/37.8s = 1.17 mW.
LED dim but clearly visible throughout the run.
2. Next, I removed the Ro/LED from the system, so current flows back to ground through L1 loop only, t = 37.0 s, Pin = 1.19mW.
A bit of a surprise, repeatedly, with the Ro/LED out of the circuit, without that load, the input power drain INCREASES.
A bit hard to explain without OU perhaps, but not sure. In any case, the power drain through the LED and 1Kohm resistor is small (if not negative ;) )...
3. So I take the output of L1 and connect it to point 6 (instead of to point 4), so that the return path is through the 1Kohm Rout. Rout/LED back in the circuit.
Now the LED is extremely dim, but visible in a darkened room. t= 52s, Pin = 0.85mW. Pin went down, as might be expected since the current through L1 is impeded by the 1Kohm Ro.
4. Next, change Ro from 1Kohm to 220ohms. Same test as in 3, now t = 62.2 sec, Pin = 0.71mW. Now this is surprising to see Pin go down with Ro reduced, since the current through L1 is impeded LESS by 220ohms than by 1Kohm Ro -- we are approaching the situation in 1 where the L1 output goes directly to ground.
5. Back to condition as in 1, but with Ro=220ohms, t= 38.2 s, Pin = 1.15mW (about the same as with Ro=1Kohm).
6. Next, I removed the Ro/LED from the system, so current flows back to ground through L1 loop only, t = 37.1 s, Pin = 1.19mW.
A bit of a surprise, repeatedly, with the Ro/LED out of the circuit, without that load, the input power drain INCREASES. As before.
Next I went to my own "replication", with the conditions as in 4 above, and found t=54 seconds (0.82mW), rather than 62.2 s with the 1st DUT. So small variations make a difference (not too surprising) -- in particular, the toroid/windings differ in my own replication. But the power draw (Pin) is much less than your Pin, Laneal...
Now, this gives us an idea of how the circuit behaves, and a measure of the low power consumption.
Now with my replication, more results:
1. I take the output of L1 and connect it to point 6 (instead of to point 4), so that the return path is through the 220ohm Rout. Rout/LED back in the circuit.
Now the LED is dim, but clearly visible. t= 54s, Pin = 0.82mW. (repeat from above)
2. Connect L1 direct back to point 4. t= 50.8s, Pin = 0.87mW. I observed that with this condition, the LED is growing dimmer as the voltage from the caps drops -- until about 1.2Vin, then the LED suddenly gets brighter -- unexpected. I repeated this experiment and observation.
3. Place a 220-ohm R in series with L1, then to point 4. t= 59.8s, Pin = 0.74mW.
I can play with the circuit until Pin is quite small, and the LED lights up for longer and longer on the two caps...but I need to get Pout measured in the best way possible, without an oscilloscope preferably. I want to check the scope results and derive a method I can do in my home lab. The goal is maximum Pout/Pin. Working on that.
Meanwhile, this weekend I have been puzzling over a relativity puzzle. Also a lot of fun, and a mental challenge. At a conference a week ago, a fellow posed a question and I've modified it so that now it looks like a real puzzle, in that momentum conservation appears to be violated... which "cannot happen".
OK, so show me what's wrong. I like to pose puzzles like this to other physicists, and if any of you have answers, pls let me know. I admit haven't found an answer yet... and somehow, it MIGHT relate to what we're doing here. Of course, at present its a thought-experiment, not done physically yet...
Consider two loops of wire facing each other, A and B, 3 cm apart as shown in the attached. We're going to use the fact that magnetic fields propagate at the speed of light so that it takes time for the field generated in A to reach B, and vice versa,
t = 3cm/3X10**10cm/s = 0.1 nanosecond (ns) for a field generated at A by a current pulse to reach B.
Sure, edge effects, etc. -- I'm not worrying about those, yet.
We send a pulsed current through A and (with time delay as shown) through B. A "positive current pulse" is such that a North pole points to the right, a negative current pulse generates a North pole pointing left. When the first + pulse goes through A, B is off. But as the field from A arrives, B switches on with a negative pulse and thus is REPELLED, pushed to the right. We can end the thought experiment there, with A turned off now and so feels no effects, while loop B (free to move) travels to the right. There is motion to the right only, which does not conserve momentum... Can you find an error?
But we can go on and get loop A to move also. The field from B travels back towards A and when it arrives 0.1 ns later, A has a negative pulse and so is pulled to the right and begins to move right. THAT field from A propagates at the speed of light towards B... which is pushed to the right when it arrives, and so on.
Thus, the whole "system" moves to the right (one can connect the loops on a platform, or a space-ship ;) , with an apparent violation of conservation of momentum...
I haven't found the error if there is one... let me know why this won't work. Thanks. :)
Quote"We send a pulsed current through A and (with time delay as shown) through B. A "positive current pulse" is such that a North pole points to the right, a negative current pulse generates a North pole pointing left. When the first + pulse goes through A, B is off. But as the field from A arrives, B switches on with a negative pulse and thus is REPELLED, pushed to the right. We can end the thought experiment there, with A turned off now and so feels no effects, while loop B (free to move) travels to the right. There is motion to the right only, which does not conserve momentum... Can you find an error?
But we can go on and get loop A to move also. The field from B travels back towards A and when it arrives 0.1 ns later, A has a negative pulse and so is pulled to the right and begins to move right. THAT field from A propagates at the speed of light towards B... which is pushed to the right when it arrives, and so on.
Thus, the whole "system" moves to the right (one can connect the loops on a platform, or a space-ship ;) , with an apparent violation of conservation of momentum...
I haven't found the error if there is one... let me know why this won't work. Thanks. :)
I can't believe this would work ...
But if it did... following some experiments certainly...
Then, I would arrange the loops (small, short-wire coils probably) into a wheel, and let the push-pull described above generate circular motion -- There's your motor, to drive a generator or a car... you see where this could lead.. ^-^
Quote from: PhysicsProf on 2011.06.13, 20:12:17
I can't believe this would work ...
But if it did... following some experiments certainly...
Then, I would arrange the loops (small, short-wire coils probably) into a wheel, and let the push-pull described above generate circular motion -- There's your motor, to drive a generator or a car... you see where this could lead.. ^-^
Very thought provoking...it seems that Newton's third law does not hold here (that's is, force and its anti-force are not symmetric).
I can certainly lower my input to your level (my LED is brighter than yours and my input power is about 15mW). Would be very interesting to see your alternative measurements of output. Charging a cap via a diode is probably a good idea. The power wasted on the diode can also be easily calculated: because there is a nearly constant voltage drop over the diode when there is current through it (thus the total energy spent on the diode is: C * deltaV * Vdiode, where C is the capacitance, deltaV is the voltage change over the capacitor, and Vdiode is the constant forward voltage drop on the diode).
lanenal
Thanks, laneal -- yes, violating Newton's third law is essentially equivalent to non-conservation of momentum.
More studies with the circuit, trying to achieve low Pin (< 1mW), then look at Pout --
Conditions:
Ein by 10,000uF cap = 10mF
Rb 52Kohms
MPS2222
C-B 372 pF (note increase)
D red LED
L-B, L-O bifilar 9turns, ferrite toroid 1"OD, ~120uH each
L1 connected to point 6 (between LED and Ro)
Ro = 220 ohms
Rr= 0, CSRout = 0 (removed)
Then, using Cap + stopwatch method, Pin is 0.41mW with LED glowing dim but easily seen in lit room
Next, to look at Pout -- crudely. I simply replaced Ro with a 10mF capacitor
Ro = > replaced by 10,000 uF cap, same as the input-power cap (yes, I need to measure the actual capacitances -- using my colleague's meter).
Using cap and stop-watch method for BOTH caps, get following results.
Ein = 1/2 C (Vstart**2 - Vfinal**2) -- discharging to provide Pin
Eout = 1/2 C (Vofinal**2 - Vostart**2) -- charging to estimate Pout. Note that the LED is glowing, and that energy is NOT captured, (various losses also not counted) so this provides a conservative estimate of Pout.
n = Eout/Ein -- since Cin ~ Cout,
n ~ (Vofinal**2 - Vostart**2) / (Vstart**2 - Vfinal**2) -- conservative estimate, made using matched capacitors for input and output energy.
Typical result:
n ~ (1.58**2 - 1.30**2) / (2.54**2 - 2.25**2) = 0.58 = 58% (conservative)
Note, Pin = Ein/time ~ 0.25 mW (Pin is less with this system, with the cap replacing the 220 ohm Ro)
Best result tonight:
n ~ (1.366**2 - 1.258**2) / (2.0**2 - 1.89**2) = 0.7 = 70% (conservative)
Now, this is with matched caps for input energy and output energy, with the energy in the LED "thrown away", as it was lit visibly. Hope you're following what I'm doing here -- achieve a low Pin while LED is still lit visibly (quite bright), then get a FIRST CONSERVATIVE ESTIMATE for Pout (or Eout in this case).
Next -- a simple test, 10mF cap at 2.53V, so Ein = 32 mJ, connect matching cap in parallel --> both caps now at 1.22 volts (volts drops by half as the charge is shared, conservation of charge). But
E = 1/2 10mF V**2
So Ein = 32 mJ, Eout = 15 mJ (measured using voltages before and after joining caps in parallel) -- so HALF THE ENERGY IS LOST JUST CHARGING ONE CAP WITH A CHARGED-CAP (= capacitances). n = 50% is the best one can do under these circumstances. Not certain if that applies in this circuit... a sim might tell that...
Here, I get n ~ 58% typically and n~ 70% best run.
Conclusion: interesting, not definitively OU in this configuration -- still would like a better way to measure Eout or Poutput routinely so I can tune caps and resistors -- and the wound-toroid -- to maximize Pout/Pin = Eout/Ein.
Quote from: PhysicsProf on 2011.06.14, 04:07:37
Thanks, laneal -- yes, violating Newton's third law is essentially equivalent to non-conservation of momentum.
More studies with the circuit, trying to achieve low Pin (< 1mW), then look at Pout --
Conditions:
Ein by 10,000uF cap = 10mF
Rb 52Kohms
MPS2222
C-B 372 pF (note increase)
D red LED
L-B, L-O bifilar 9turns, ferrite toroid 1"OD, ~120uH each
L1 connected to point 6 (between LED and Ro)
Ro = 220 ohms
Rr= 0, CSRout = 0 (removed)
Then, using Cap + stopwatch method, Pin is 0.41mW with LED glowing dim but easily seen in lit room
Next, to look at Pout -- crudely. I simply replaced Ro with a 10mF capacitor
Ro = > replaced by 10,000 uF cap, same as the input-power cap (yes, I need to measure the actual capacitances -- using my colleague's meter).
Using cap and stop-watch method for BOTH caps, get following results.
Ein = 1/2 C (Vstart**2 - Vfinal**2) -- discharging to provide Pin
Eout = 1/2 C (Vofinal**2 - Vostart**2) -- charging to estimate Pout. Note that the LED is glowing, and that energy is NOT captured, (various losses also not counted) so this provides a conservative estimate of Pout.
n = Eout/Ein -- since Cin ~ Cout,
n ~ (Vofinal**2 - Vostart**2) / (Vstart**2 - Vfinal**2) -- conservative estimate, made using matched capacitors for input and output energy.
Typical result:
n ~ (1.58**2 - 1.30**2) / (2.54**2 - 2.25**2) = 0.58 = 58% (conservative)
Note, Pin = Ein/time ~ 0.25 mW (Pin is less with this system, with the cap replacing the 220 ohm Ro)
Best result tonight:
n ~ (1.366**2 - 1.258**2) / (2.0**2 - 1.89**2) = 0.7 = 70% (conservative)
Now, this is with matched caps for input energy and output energy, with the energy in the LED "thrown away", as it was lit visibly. Hope you're following what I'm doing here -- achieve a low Pin while LED is still lit visibly (quite bright), then get a FIRST CONSERVATIVE ESTIMATE for Pout (or Eout in this case).
Next -- a simple test, 10mF cap at 2.53V, so Ein = 32 mJ, connect matching cap in parallel --> both caps now at 1.22 volts (volts drops by half as the charge is shared, conservation of charge). But
E = 1/2 10mF V**2
So Ein = 32 mJ, Eout = 15 mJ (measured using voltages before and after joining caps in parallel) -- so HALF THE ENERGY IS LOST JUST CHARGING ONE CAP WITH A CHARGED-CAP (= capacitances). n = 50% is the best one can do under these circumstances. Not certain if that applies in this circuit... a sim might tell that...
Here, I get n ~ 58% typically and n~ 70% best run.
Conclusion: interesting, not definitively OU in this configuration -- still would like a better way to measure Eout or Poutput routinely so I can tune caps and resistors -- and the wound-toroid -- to maximize Pout/Pin = Eout/Ein.
Thanks professor for sharing those measurements.
So with a 10mF capacitor in place of R0, the voltage rises to 1.58V from 1.30V.
For a red LED, the forward voltage drop is about 1.67V (I measured mine with a DMM, please replace it with your own measurement).
Therefore, the power spent on the LED is: C * deltaV * Vdiode = 10mF * (1.58 - 1.30) * 1.67 = 4.676mJ.
The total energy stored in the cap is (1.58**2 - 1.30**2) * 10mF /2 = 4.03
So, total output is: 4.676+4.03 = 8.706mJ.
Total input: (2.54**2 - 2.25**2) * 10mF/2 = 6.9455
Therefore n = 8.706 / 6.9455 = 1.253
Hey, that's more like an OU :). But clearly it needs an accurate measurement of your Vdiode.
P.S.: my calculation shows that as long as the Vdiode > 1.04125V, we will have n>1.
Quote from: PhysicsProf on 2011.06.14, 04:07:37
Best result tonight:
n ~ (1.366**2 - 1.258**2) / (2.0**2 - 1.89**2) = 0.7 = 70% (conservative)
For this best case, if Vdiode > (1.366V+1.258V)/2 = 1.312V, then n > 140%.
Vdiode is the voltage drop over the diode at the moment when the capacitor is being charged.
For this best case, if Vdiode > 0.667, n>1.
Of course, this is still a conservative computation of n, as we have not calculated power wastes in the transistor and toroid.
I wrote this Python function to compute n and vpar. see below:
def computen(vi0, vi1, vo0, vo1, vd, ci, co=None):
'''compute the n = Eout/Ein, where
vi0 = initial voltage of input capacitor,
vi1 = final voltage of input capacitor,
vo0 = initial voltage of output capacitor,
vo1 = final voltage of output capacitor,
vd = forward voltage drop over the diode.
ci = capacitance of input capacitor,
co = capacitance of output capacitor,
it defaults to None, meaning co = ci.
Ein = (vi0**2 - vi1**2)*ci/2,
Eout= (vo1**2 - vo0**2)*co/2 + vd*(vo1-vo0)*co
This function also computes the critical vpar
such that when vd = vpar, n = 1.
'''
if co is None: co = ci #default
Ein = (vi0**2 - vi1**2)*ci/2.
Eoc = (vo1**2 - vo0**2)*co/2.
Eout = Eoc + vd*(vo1-vo0)*co
vpar = (Ein - Eoc)/(vo1-vo0)/co
return Eout/Ein, vpar
I further modified my simulation, now the voltage power turns off (drops to zero volts) after 2 ms. See below the simulation of the first 10 ms. It clearly runs on the charged capacitor for the last 8 ms. V(vtop) - V(vbot) is the voltage difference over the voltage power source. I(L1) and I(C1) are the currents in L1 and C1.
Good points, laneal. Here I attach a photo of the set-up, using one cap for Ein and the second cap to capture some of the Eout. Two DMM's read the voltage, one on each cap.
Also a screen shot of the voltage across the LED, while powered by the Ein cap alone, with the second cap charging (from approx 0 volts, starting voltage).
The measurements read:
Vmean 40 mV, which is approx what I read with the DMM across the LED... but --
Vpp 4.84V
Vrms 680mV
Vtop 3.12V -- this is the voltage in the forward direction, the direction of the current flow allowed by the LED
So you tell me, you asked for the voltage across the LED -- but which voltage does one use??
Quote from: PhysicsProf on 2011.06.14, 13:55:52
Good points, laneal. Here I attach a photo of the set-up, using one cap for Ein and the second cap to capture some of the Eout. Two DMM's read the voltage, one on each cap.
Also a screen shot of the voltage across the LED, while powered by the Ein cap alone, with the second cap charging (from approx 0 volts, starting voltage).
The measurements read:
Vmean 40 mV, which is approx what I read with the DMM across the LED... but --
Vpp 4.84V
Vrms 680mV
Vtop 3.12V -- this is the voltage in the forward direction, the direction of the current flow allowed by the LED
So you tell me, you asked for the voltage across the LED -- but which voltage does one use??
Professor, thanks for the diagrams. Conceptually it should be V4 - V3 as labeled in your schematic 2Cap13Jun2011.jpg. However, there is a simple way to get a good estimation of it, see below (quoted from http://www.repairfaq.org/sam/semitest.htm#sttdj):
On a (digital) DMM, there will usually be a diode test mode. Using this, a silicon diode should read between .5 to .8 V in the forward direction and open in reverse. For a germanium diode, it will be lower, perhaps .2 to .4 V or so in the forward direction. Using the normal resistance ranges - any of them - will usually show open for any semiconductor junction since the meter does not apply enough voltage to reach the value of the forward drop. Note, however, that a defective diode may indeed indicate a resistance lower than infinity especially on the highest ohms range. So, any reading of this sort would be an indication of a bad device but the opposite is not guaranteed.
Therefore, you may also replace the LED with a germanium diode and probably get better n (without calculating the energy wasted on the LED).
Another way to get better n is this: let the output capacitor having a higher initial voltage relative to the diode's forward voltage.
lanenal
Quote from: lanenal on 2011.06.11, 15:00:36
I'd try to charge a capacitor. I've got some 10000uF caps, hopefully will do. Any suggestions?
For a possible Pout measurement, we can try using a single diode or FWBR output to charge a capacitor. Then connect a potentiometer across this output cap and monitor the DC output with a DMM. Slowly decrease the resistance of the potentiometer until the DC output voltage begins to drop steadily. Back it off until the output voltage holds steadily. Now use the output voltage and resistance reading on the pot to compute power. I would suggest somewhere between a 10k to 50k pot.
.99
Professor, Lane,
I have tried my own sim based on the values shown in schema01.png. I used a modeled non-linear core of 1" diameter and 9 turns each coil.
Fo is approximately 1.68MHz. Correction, the Fo is about 174kHz. It was about 1.68MHz before I added the LED and RO to the output.
Is the wave form shown, taken off the emitter, similar to what you are seeing?
I've not done any power measurements yet, as I want to know if I'm close or not first.
Regards,
.99
Edit: Corrected Fo frequency.
I like this idea, .99:
Quote"For a possible Pout measurement, we can try using a single diode or FWBR output to charge a capacitor. Then connect a potentiometer across this output cap and monitor the DC output with a DMM. Slowly decrease the resistance of the potentiometer until the DC output voltage begins to drop steadily. Back it off until the output voltage holds steadily. Now use the output voltage and resistance reading on the pot to compute power. I would suggest somewhere between a 10k to 50k pot.
We can know the Pinput from the cap/stop-watch tests and can then run off a battery at a steady Vin, and with your cap/poten. measurement for Poutput, we can keep conditions steady. It concerned me that with variable voltages on Vin and on the output cap, one could not really "tune" the circuit. This way is much better. Thank you!
Quote from: poynt99 on 2011.06.15, 14:26:31
Professor, Lane,
I have tried my own sim based on the values shown in schema01.png. I used a modeled non-linear core of 1" diameter and 9 turns each coil.
Fo is approximately 1.68MHz. Correction, the Fo is about 174kHz. It was about 1.68MHz before I added the LED and RO to the output.
Is the wave form shown, taken off the emitter, similar to what you are seeing?
I've not done any power measurements yet, as I want to know if I'm close or not first.
Regards,
.99
Edit: Corrected Fo frequency.
Yes, the output waveform is similar to what I'm seeing, so pls carry on with power measurements.
I would like especially to see:
1. Is the average input power small, as I'm observing with the actual DUT -- that is around 1 mW?
2. What is the shape of the input power waveform?
These will provide (IMO) a good test of the SIM.
Professor, as the circuit stands, the various average power measurements are as follows:
PLED=0.31mW
PRO= 0.76mW
PBAT=-1.82mW
The Battery power wave form is attached.
.99
Quote from: poynt99 on 2011.06.15, 14:26:31
Professor, Lane,
I have tried my own sim based on the values shown in schema01.png. I used a modeled non-linear core of 1" diameter and 9 turns each coil.
...
.99, great that you have been able to do the non-linearity sim, it would definitely contribute our understanding. thank you.
professor, I wonder if you could also try (with a battery or a cap at the input -- or both, doesn't matter) to measure Pout with that nice scope when a diode and a cap is joined at the output. There are two ways to estimate Pout:
1. Only measure the output energy dumped into the cap
( use math: PoutCap = MEAN[V4*Iout] ).
2. Measure the output energy dumped into the cap and that wasted on the LED/diode
(use math: PoutBoth = MEAN[V3*Iout] ).
Here Iout is the output current through the diode/cap, which can be measured by a CSR as you have done many times before.
In the above measurements, the ground is at the joint of the cap and L1.
The purpose of this experiment is to verify:
PoutCap/PoutLED = (Vc1 + Vc0)/(2*Vdiode)
where
PoutLED = PoutBoth - PoutCap
Vc1: final voltage over the output cap.
Vc0: initial voltage over the output cap.
Vdiode: forward voltage drop over the diode.
If that can be verified, we can greatly simplify our calculations without using a scope later on.
lanenal
Professor,
Have you tried changing the polarity of the LED? What differences do you see?
When I reverse the polarity in the sim, the frequency goes back up to 1.68MHz or so, and there is more dissipation and light power in the LED.
.99
Quote from: poynt99 on 2011.06.16, 19:00:18
Professor,
Have you tried changing the polarity of the LED? What differences do you see?
When I reverse the polarity in the sim, the frequency goes back up to 1.68MHz or so, and there is more dissipation and light power in the LED.
.99
Yes, reversing the LED results in larger Pinput, that is, more dissipation as you say.
Here are the conditions and results for the latest low-power-consumption run:
I used the Cap/stopwatch method.
P = Ein/time, measuring Ein using a cap, from 2.55 V to 1.5 V, so around 2Volts in.
Best result (to date): 12.7 seconds using a 1000 uF cap for Piin, so
P = 1/2 10mF (2.55**2 - 1.5**2)/12.7s
= 0.17 mW = 170 uW.
That is, "170 microwatts".
Frequency -- power to output leg in spikes, f = 1.44Khz (not MHz) at 2.5 V, and at 1.5V, f = 1.31KHz.
Conditions: Rb = 47 Kohms
Cb = 223 pF (ceramic cap)
MPS 2222 transistor
Ro = 220 ohms
Lo ~ Lb ~ 130 uH
No CSRout or Co or Rr or CSRin.
Red LED lit visibly in lighted room.
With LED reversed, time for 1mF cap to drop from 2.55V to 1.5V is 11.9s, so 180 uW.
My JT could put out 4uW if the math is right.
7/1000 volts every 10 seconds if I remember correctly.
4700uF 1.3 volts start
Frequency function broke
I also shunt battery through several MOhm resistors in series with LEDs to study input. Seems like they are close in current.
Quote from: GibbsHelmholtz on 2011.06.17, 17:33:13
My JT could put out 4uW if the math is right.
7/1000 volts every 10 seconds if I remember correctly.
4700uF 1.3 volts start
Frequency function broke
I also shunt battery through several MOhm resistors in series with LEDs to study input. Seems like they are close in current.
I'm not sure I understand you here, Gibbs. Are you saying you have a JT that will light up an LED visibly with only 4uW input power?
(Please explain)
Professor,
Is there a specific requirement to have a diode or LED in the output?
Could a resistor only be there for determining Pout?
.99
Quote from: PhysicsProf on 2011.06.17, 19:04:18
I'm not sure I understand you here, Gibbs. Are you saying you have a JT that will light up an LED visibly with only 4uW input power?
(Please explain)
hehehe I'm sorry Professor,
I meant to say the input but I think it could be mistaken with "put out".
I have to go now but I'll provide the schematic when get back. But like you said, the LED is tuned to just visible condition to achieve this status. I'm working on increasing lighting condition with little input as possible.
Alright, back
A few comments about my JT experiment
This design is basically derived from the Professor's JT. I used a PNP. It was 3906. The assorted bag came in 3 types I think, but I never tried the other two types. I know the NPN types make a different. The resistor was 1MOhm pot tuned to max. I tried to add a couple more MOhms but it makes little difference. The LED is barely visible. I can see the red LED but I like pink LED better. There are also 2 positions one can place the LED. It makes little difference in power consumption if detectable. I like the emitter-collector position better because it gets a voltage boost from the source. The Cap and the bifilar can have a good range under similar operation. I used a .01uF metalized film and forgot what coil I used since I had so many.
WaRnINg,
I've seen the capacitor/resistor combination gives out a pulse once a full moon so I can only give warranty that it does not operates at 1-5Hz . No refund if you find it operates at 10 or 60Hz. lol
Quote from: poynt99 on 2011.06.17, 21:22:31
Professor,
Is there a specific requirement to have a diode or LED in the output?
Could a resistor only be there for determining Pout?
.99
How do you determine Pout for an "erratic" AC output with just a resistor ? Unless you are using a Thermal Wattmeter or equivalent -- I'm still trying to hunt down one of those. In any case, no I would not make a "specific requirement to have a diode or LED in the output" if you're going to measure Pout. The LED in the output was for the cap/stopwatch test for Pinput, with the LED glowing... something any one can do without a scope.
CORRECTION: to measure Pout properly, without drastically altering the DUT, the diode in the output leg of the circuit must be present.
Quote from: PhysicsProf on 2011.06.18, 01:38:55
How do you determine Pout for an "erratic" AC output with just a resistor ?
Quite simply! (I'll get to that shortly)
How does adding a diode make it any more possible?
.99
Quote from: GibbsHelmholtz on 2011.06.18, 00:17:12
Alright, back
A few comments about my JT experiment
This design is basically derived from the Professor's JT. I used a PNP. It was 3906. The assorted bag came in 3 types I think, but I never tried the other two types. I know the NPN types make a different. The resistor was 1MOhm pot tuned to max. I tried to add a couple more MOhms but it makes little difference. The LED is barely visible. I can see the red LED but I like pink LED better. There are also 2 positions one can place the LED. It makes little difference in power consumption if detectable. I like the emitter-collector position better because it gets a voltage boost from the source. The Cap and the bifilar can have a good range under similar operation. I used a .01uF metalized film and forgot what coil I used since I had so many.
WaRnINg,
I've seen the capacitor/resistor combination gives out a pulse once a full moon so I can only give warranty that it does not operates at 1-5Hz . No refund if you find it operates at 10 or 60Hz. lol
Right, 60Hz operation would imply "poaching energy" from the grid, not what we're looking for ... lol.
Can you put that circuit or one like it back together, and:
1. Measure Pinput by the Cap/Time method described above?
Here are the straightforward equations:
Ecap = 1/2 C V**2
so
Pinput = 1/2 C (Vstart**2 - Vstop**2)/time
2. Measure the frequency?
Yeah, we better check that it's not 60 Hz.
Also, somehow we need to quantify a MINIMUM brightness for the LED to make useful comparisons between circuits...
I have another idea -- experimenter can replace his LED with a known diode, 1N4148 which is common, and then use the stop-watch/Cap method.
@.99 -- Look, I'm trying to get at ways to compare INPUT power for various circuits -- and I'm recommending a STANDARD using the cap/stop-watch method.
For Pout, adding a diode is not the answer... a Thermal Wattmeter may be a very good answer, as I've noted before.
I use the LED to detect if my circuit is on. I also use it as a rule of thumb for energy output since the amount of light is proportional to energy (just a rule of thump for nonscopers). I also use it to charge up a capacitor to quantify somewhat of the output.
Professor,
I think my last results holds. .007V/10 seconds. I also see a person name xee on OU that shares the same circuit in NPN configuration. His stated power was 10uW so I think we do have something in common. My frequency function broke.
I tried to charge up an identical 4700uF cap and the input:output is 3:1 ratio (every 3 unit volt input I recover 1 unit volt ouput). I believe the rate is constant though. Also the LED stop when I charge the recover cap to about 2.4V from outside source. This indicate my peak voltage is below 4-5V. The diode also leaks.
I have put all this behind and moved on. Now I have roughly the same input with 2 LEDs lit. LOL that means my peak voltage is double? LOL how could this be? Of course there is a trick to this. hehehe
Quote from: PhysicsProf on 2011.06.18, 01:50:40
Also, somehow we need to quantify a MINIMUM brightness for the LED to make useful comparisons between circuits...
Yes, that's why I made that post at OU.
QuoteI have another idea -- experimenter can replace his LED with a known diode, 1N4148 which is common, and then use the stop-watch/Cap method.
This does not solve the problem of having a baseline for comparison. What criteria would there now be for the circuit operating?
Quote
@.99 -- Look, I'm trying to get at ways to compare INPUT power for various circuits -- and I'm recommending a STANDARD using the cap/stop-watch method.
I'm trying to help you in both regards; Pin and Pout. No need to get aggravated.
Quote
For Pout, adding a diode is not the answer...
You misread me. I was not trying to suggest that. I was trying to suggest the contrary actually.
.99
Quote from: GibbsHelmholtz on 2011.06.18, 02:08:03
I use the LED to detect if my circuit is on. I also use it as a rule of thumb for energy output since the amount of light is proportional to energy (just a rule of thump for nonscopers). I also use it to charge up a capacitor to quantify somewhat of the output.
Professor,
I think my last results holds. .007V/10 seconds. I also see a person name xee on OU that shares the same circuit in NPN configuration. His stated power was 10uW so I think we do have something in common. My frequency function broke.
I tried to charge up an identical 4700uF cap and the input:output is 3:1 ratio (every 3 unit volt input I recover 1 unit volt ouput). I believe the rate is constant though. Also the LED stop when I charge the recover cap to about 2.4V from outside source. This indicate my peak voltage is below 4-5V. The diode also leaks.
I have put all this behind and moved on. Now I have roughly the same input with 2 LEDs lit. LOL that means my peak voltage is double? LOL how could this be? Of course there is a trick to this. hehehe
Hmm... I'm asking both you and Xee2 to use this STANDARD cap/time method if you would.
1. Measure Pinput by the Cap/Time method described above?
Here are the straightforward equations:
Ecap = 1/2 C V**2
so
Pinput = 1/2 C (Vstart**2 - Vstop**2)/timeBut I can see that a minimum frequency might also be needed for the test... see below.
Some of you may not realize this discussion is going on in parallel over at OU.com -- and to understand the dialog, you may want to read there also:
http://www.overunity.com/index.php?topic=10773.330
My last reply:
QuoteQuote from: poynt99 on 2011.06.18, 02:28:34
This does not solve the problem of having a baseline for comparison. What criteria would there now be for the circuit operating?
With your current criteria, I am certain I can beat everyone. I can build a circuit that will drop from 2.5V to 2.4V in 10 seconds.....do I win?
.99
Required is a JT-type circuit, with at least one transistor and one bifilar-wound toroid and it must light an LED to observable brightness in a lighted room (all as I said before IIRC) before the final test with a 1N4148 replacing the LED.
And the winner will check the FREQUENCY also...
I'm going to add a provisional stipulation that the frequency of input-power pulses be above 200 Hz, to prevent pulsing "once in a blue moon" and also to avoid 60-Hz pick-up. If you can persuade me this is not a "fair" stipulation, I'm listening.
The winning device cannot be poaching from the grid.
How's that? and I do appreciate your checking the criteria, .99. Rigor without rancor.
@.99 -- do you have something new for measuring Pout? I'm very interested if you've come with something better than the Thermal Wattmeter...
I don't realized that there is a contest going on here. I'm not entering the contest but my interest is to compares input and how much light one can get.
1.3V - 1.293V on 4700uF in 10 seconds
Pin = 1/2 * .0047 *(1.3*1.3 - 1.293*1.293)/ 10sec
Pin = .000004267 W
Pin = 4.26 uW
I'll check on the frequency later.
There are also an old video I see that related.
http://www.youtube.com/user/jonnydavro#p/u/14/tQecrP5ZPYQ
@2:15 the voltage reading is 14.03
@2:20 the voltage reading is 13.87
2200uF
joule ringer
frequency-unknown
Pin=1/2 * .0022(14.03*14.03 - 13.87-13.87)/5 sec
Pin = .000982 W
Pin = 982 uW
@4:20 6.56V
@4:32 6.27V
Pin = 1/2 *.0022(6.56*6.56 - 6.27*6.27)/12 sec
Pin = .000341 W
Pin = 341uW
Quote from: GibbsHelmholtz on 2011.06.18, 04:24:21
... my interest is to compares input and how much light one can get.
1.3V - 1.293V on 4700uF in 10 seconds
Pin = 1/2 * .0047 *(1.3*1.3 - 1.293*1.293)/ 10sec
Pin = .000004267 W
Pin = 4.26 uW
I'll check on the frequency later.
Remarkable, Gibbs! is there a video of this? I agree with your goal of comparing Pin and light output.
I don't suppose you could dig up (or rebuild) the device and get similar results with a 1N4148 diode replacing the LED, and taking the measurement at 2 volts? so we can have a standard for comparisons with other circuits ...
Quote from: PhysicsProf on 2011.06.18, 05:17:04
Remarkable, Gibbs! is there a video of this? I agree with your goal of comparing Pin and light output.
I don't suppose you could dig up (or rebuild) the device and get similar results with a 1N4148 diode replacing the LED, and taking the measurement at 2 volts? so we can have a standard for comparisons with other circuits ...
No video but sure I'll take a note of the 1N4148 and 2 volts.
There is something unusual about the video I posted. The meter next to the voltage read .42 mA @ 13.87V and .57mA @ 6.27V
13.87 * .42 = .0058 = 5.8mW Cap method is = .982mW
6.27 * .57 = .00357 = 3.57mW Cap method is = .341mW
Ah... I have to think about this.
Quote from: GibbsHelmholtz on 2011.06.18, 05:46:15
No video but sure I'll take a note of the 1N4148 and 2 volts.
Wow, did you just say 1N4148? That was exactly the diode I used in my simulation. See reply #125 on page 6 of this thread.
Thanks, Gibbs -- pls get back to us on this.
Quote from: lanenal on 2011.06.18, 09:46:14
Wow, did you just say 1N4148? That was exactly the diode I used in my simulation. See reply #125 on page 6 of this thread.
Yes, I see that. I take it your frequency there was approx 1/1.8us ~ 550KHz, is that right? Do you have a measure of the Pinput from this simulation?
The discussion on the contest is continuing with some enthusiasm at OU.com, so I will focus on that, at this time.
e.g., from nul-pts: http://www.overunity.com/index.php?topic=10773.msg291497#msg291497
Quote
i've been running my SJ1 variant tests from a single (depleted) AAA NiMH, and the voltage has been around the 1.12-1.15V range
just for fun i constructed another unit to add to the comparative results you're getting back from folks
the circuit is a minor mod to one of my earlier tests with a variant of your SJ1 circuit (see schematic below for this test circuit)
C1 0.022uF
D1 1N5187 (schottky)
Q1 2N3906
T1 approx 50:50:100; 0.45mm wire; tri-toroid (ferrite)
C2 1000uF (nominal)
L1 approx 0.5mH
LED1 6mm(?) HiBrite (visible, but not bright)
i'm using a tertiary winding to decouple the AC o/p from the DC operating conditions of the oscillator
this circuit takes 433 seconds (7min 13sec) to discharge a nominal 1000uF cap from 2.24V to 1.5V
C2 1000uF (nominal)
2.24V => 2.509mJ
1.50V => 1.125mJ
-------
Ein: 1.384mJ
Pav: 1.384/433 = 3.2uW
this is all just ballbark at the moment, obviously - to be more accurate, the 1000uF cap would need measuring... " etc.
This is close to the number you mentioned, Gibbs... y'all may wish to follow the discussion there at OU.
Quote from: PhysicsProf on 2011.06.18, 13:11:19
Yes, I see that. I take it your frequency there was approx 1/1.8us ~ 550KHz, is that right? Do you have a measure of the Pinput from this simulation?
I looked into the way of doing it in LTSpice, and it can do RMS and MEAN calculations over many cycles. I found that input is 955uW and output is 213uW. Not surprising here. These numbers shows us that a lot of power is spent at the transistor/coil/core/diode. Therefore when we get a n=50% when only considering the power stored in a cap at the output, that is a strong indicator of ou. As I have shown in the other posts, including the power dissipation on the diode might be enough to show ou.
Quote from: PhysicsProf on 2011.06.18, 14:30:11
The discussion on the contest is continuing with some enthusiasm at OU.com, so I will focus on that, at this time.
e.g., from nul-pts: http://www.overunity.com/index.php?topic=10773.msg291497#msg291497
This is close to the number you mentioned, Gibbs... y'all may wish to follow the discussion there at OU.
Yes, I just read that. I am very impressed. He beat me but not just by a little bit, a whole lot! I have to put my hands around my LED to see it and the worst case, turned off the light lol. I'm also glad we coming to a standard agreement. Just shake it around until everyone feel generous to give in.
Quote from: lanenal on 2011.06.18, 09:46:14
Wow, did you just say 1N4148? That was exactly the diode I used in my simulation. See reply #125 on page 6 of this thread.
Lane,
I was very impressed with your looping technique. I had a lot of fun trying to looped the JT past weeks with your technique. This is why I know COP>1 is real, but it won't be in the form of charge. One can see if the circuit reached COP>1 easily with
both cap/stopwatch and dual DMM method! The cap/stopwatch method is always be less than or equal to Dual DMM method. The energy consuming from the capacitor is a real component, there is no if, and, or but about it. The dual DMM method is not totally invalid. We can use it to determine the phase angle and calculate the ratio of heat&radiation: energy dissipate from cap. It is now important to get this thing in a chamber where not only heat, but radiation can be efficiently captured.
I think we should reconsider the condition for this contest. Here is my argument.
What if a guy with a circuit COP 5 that has an input of 5 watts vs a circuit with COP .5 with 1 watt input? The only way we can tell is to compare input to output and it's not an easy task up to now. I think the quality of a circuit should depends on its COP and not input or output. But how?
I propose that we should want the quality of the circuit in relation to how much it is out of phase. As I mention earlier about the cap/watch method and the dual DMM method. The cap/watch method is a sure way to know how much we put in. But what does the dual DMM method means? Here is what I see:
Consider the attached image below. With the dual DMM method when voltage and current are in phase, it should multiply the Average of the voltage and the average of the current. This is the power consumption and should agree with the cap/watch method. Now how about an out of phase scenario? The average of the voltage and current are exactly the same as the in phase condition and thus the dual DMM method gives the same measurement. But we know this is incorrect if there is an out of phase condition, thus by comparing the real input from the cap to the dual DMM method, we have a mean to know how much the phase angle is off. The quality of the circuit can be express as the ratio of the two.
Gibbs,
If your input source is DC, the phase angle between the input voltage and current is always 0º.
Aside from the very low net input currents involved, and the potential difficulty in getting a nice clean average reading of it, the INPUT power measurement with a non-inductive CSR and single (or dual) DMM is a snap.
.99
Quote from: poynt99 on 2011.06.19, 03:21:08
Gibbs,
If your input source is DC, the phase angle between the input voltage and current is always 0º.
Aside from the very low net input currents involved, and the potential difficulty in getting a nice clean average reading of it, the INPUT power measurement with a non-inductive CSR and single (or dual) DMM is a snap.
.99
Poynt,
Why do you require a non-inductive CSR?
Quote from: lanenal on 2011.06.18, 16:01:27
I looked into the way of doing it in LTSpice, and it can do RMS and MEAN calculations over many cycles. I found that input is 955uW and output is 213uW. Not surprising here. These numbers shows us that a lot of power is spent at the transistor/coil/core/diode. Therefore when we get a n=50% when only considering the power stored in a cap at the output, that is a strong indicator of ou. As I have shown in the other posts, including the power dissipation on the diode might be enough to show ou.
Good points, Lane... and also Gibbs.
The discussion at OU -- note that I found this morning on my variant of Xee2's variant, that the LED dims quickly down (Pinput ~ 700mW), then the LED turns back ON (clearly) while the Pinput has dropped dramatically.
/2 C V**2
So Pinput = 1/2 C (Vstart**2 - Vend**2)/time
= 1/2 10mF (2.8**2 - 1.7**2) / 3.4 seconds = 7.3 mW =
7300 uW (above Vcritical)
Not too exciting so far, but I noticed that the LED came BACK ON at approx 1.63 Volts! Not bright, but clearly glowing. Note that Xee2 on his schematic specifies 1.366V as Vinput, so this would be in the range of much lower power consumption, Pinput.
Note that the power consumption is hundreds of times LESS:
Pinput = 1/2 C (Vstart**2 - Vend**2)/time
= 1/2 10mF (1.431**2 - 1.243**2) / 200 seconds = 0.0126 mW = 12.6 uW !! (below Vcritical)
The input power has dropped by a factor of over 500.
Next I put a 1ohm resistor in series with the LED, and the Pinput dropped to about
9.3 uW. As in "microwatts"!!
BTW, I noticed the same pattern for the sj1 circuit a while back, dimming, then the LED comes back on and glows for a lot longer thereafter. I thought it was just a curiosity when I discussed this with smartscarecrow last week, but now I see this as an important effect -- and very dramatic with Xee2's circuit, a huge reduction in Pinput.
With a red LED instead of Green, the Pinput is roughly the same (11 uW from 1.291V to 1.117V in 200s). The red LED goes OFF at about 1.44 Vin (from the cap) and back on at lower Pinput, at about 1.406 Vin.
Now this is exciting -- but why the huge drop in Pinput at a critical voltage?? I really don't know, but would like to understand. Will some of you jump in and let's see if we can figure this out? My GUESS is that the lower Pinput range is the most interesting in terms of seeking OU.
I just did some testing on this effect.
I use my PNP configuration because it seems to give a clearer window than NPN.
The pot is 50K and the capacitor is 1uF.
V critical is different depends on LED. The white and pink LED have higher V critical than the red.
Actually, I observed 3 stages.
1st stage: highest voltage, LED blinking fast, meter read 77-170 Hz
2nd stage: frequency start to increase to 200-400 Hz and suddenly jumped to 1.2-1.5Mhz for 1-1.5 seconds. This sucks alot of energy. LED becomes bright.
3rd stage: LED dimmed out, then back to blinking mode after that 1-1.5 seconds with reading 30-40 kHz and dropping
I think the stage where you stated 7300uW is the 2nd stage. The low power is the 3rd stage.
Quote from: GibbsHelmholtz on 2011.06.19, 04:28:25
Poynt,
Why do you require a non-inductive CSR?
With inductance comes inductive reactance (http://www.ndt-ed.org/EducationResources/CommunityCollege/EddyCurrents/Physics/selfinductance.htm) to high frequencies and transients. This skews the current measurement (makes it look like less current, hence less power), and that is not what we want.
Even with a non-inductive CSR, keep the probe leads as close as possible to the CSR body.
.99
Thanks for doing that testing, Gibbs, and for presenting your results clearly.
Note that over at OU, the circuit with the LOWEST Pinput is now that of Nul-pts:
QuoteSteven
i made an interesting discovery - the tertiary winding for the o/p in my previous circuit is redundant! (see below for updated schematic)
the latest circuit, using a 1000uF 35V cap as C2 (all other other parts as posted above) takes 687 seconds (11min 27sec) to discharge a nominal 1000uF cap from 2.55V to 1.5V
C2 1000uF (nominal)
2.55V => 3.251mJ
1.50V => 1.125mJ
-------
Ein: 2.126mJ
Pav: 2.126/687 = 3.1uW (including cap leakage)
thanks
np
Xee2 has a circuit with Pinput of 3.9uW, also measured using the cap/time method for Pinput. (But still working towards consistency in measurements.)
To me, these results are remarkable. I've achieved about 7-10uW with my "best" circuit. Interestingly, I have put a 21-ohm R in series with the LED and can still get Pinput ~ 10 uW.
@Gibbs -- can you measure the Pinput for your circuit? As I also asked of the guys there:
Quote1. Can you test from 1.37 V down to 1.10 V, using the cap/time method? That should give plenty of time to get an accurate reading. You may use different caps as you wish, but please give the correction for cap leakage (done by seeing how the cap voltage drops over the same time as the run -- but without cap connected to the DUT. Subtract this "effective cap-power loss" from the measured Pinput. This leakage correction will be smaller for shorter time in the run).
2. AND, if you have a 1N4148 diode available, pls repeat the test with this "standard diode" in place of the LED.
Quote from: poynt99 on 2011.06.21, 03:45:31
With inductance comes inductive reactance (http://www.ndt-ed.org/EducationResources/CommunityCollege/EddyCurrents/Physics/selfinductance.htm) to high frequencies and transients. This skews the current measurement (makes it look like less current, hence less power), and that is not what we want.
Even with a non-inductive CSR, keep the probe leads as close as possible to the CSR body.
.99
From the link you provided:
"...When there is inductive reactance present in the circuit, the phase of the current will be shifted so that its peaks and valleys do not occur at the same time as those of the voltage...."
Can this happens to a DC source with phase angle is always 0 degree? Or are you referring to a special case where DC is pulsing?
Professor,
I will do it tomorrow.
Quote from: GibbsHelmholtz on 2011.06.21, 05:15:28
...
"...When there is inductive reactance present in the circuit, the phase of the current will be shifted so that its peaks and valleys do not occur at the same time as those of the voltage...."
Can this happens to a DC source with phase angle is always 0 degree?
...
Oh my God! An "always 0 degree" phase angle that could be shifted? J.W. Gibbs and H.v. Helmholtz must be turning in their grave with such a question. C.C
Quote from: GibbsHelmholtz on 2011.06.21, 05:15:28
From the link you provided:
"...When there is inductive reactance present in the circuit, the phase of the current will be shifted so that its peaks and valleys do not occur at the same time as those of the voltage...."
Can this happens to a DC source with phase angle is always 0 degree? Or are you referring to a special case where DC is pulsing?
Gibbs,
They are referring to the element that exhibits the inductance. So if the CSR has some associated inductance, the voltage across the CSR, and current through it will not be in phase. Since we are using the CSR to determine the battery current, in such a case the measurement will be in error.
.99
Quote from: poynt99 on 2011.06.19, 03:21:08
Gibbs,
If your input source is DC, the phase angle between the input voltage and current is always 0º.
.99
Thank you for your reply, Poynt... and your comment, Ex.
Professor, here is my result.
1.37V - 1.10V
Using coil ends for red LED, then 1N4148, then without both
1.37V - 1.10V, red LED, 3m54s, 6.70uW
Cap leak 1.370V-1.360V , cap leak power .274uW
Compensated power, 6.426uW
1.37V - 1.10V, 1N4148, 6m14s , 4.19uW
Cap leak 1.370V-1.357V, cap leak power .223uW
Compensated power, 3.967uW
Additional run with out both, 5m58s
Also cap leak is less when capacitor is exercised, maybe memory recharge.
Quote from: GibbsHelmholtz on 2011.06.21, 13:45:32
Professor, here is my result.
1.37V - 1.10V
Using coil ends for red LED, then 1N4148, then without both
1.37V - 1.10V, red LED, 3m54s, 6.70uW
Cap leak 1.370V-1.360V , cap leak power .274uW
Compensated power, 6.426uW
1.37V - 1.10V, 1N4148, 6m14s , 4.19uW
Cap leak 1.370V-1.357V, cap leak power .223uW
Compensated power, 3.967uW
Additional run with out both, 5m58s
Also cap leak is less when capacitor is exercised, maybe memory recharge.
Very impressive results, Gibbs! O0 4 uW is in line with what Xee2 and nul-pts are getting...
Now, can you tell us what circuit you used? A schematic would of course be great.
Circuit info:
PNP 3906
4700 uF input
1 MOhm pot
.01uF
28 turns 24AWG magnet wire bifilar on 5/8'' radioshack toroid (roughly measured with a ruler)
Gibbs:
Looks like a very cheap and easy circuit to replicate, and interestingly simple. I just picked up the a package of 15 Radio Shack PNP 2n3906 Transistors (part # 276-1604), for super cheap. For just a few bucks, 15 of them.
Would be interested in more info on the Radio Shack toroid(s), that you use, core type, specs. etz... as there are those stores here in Costa Rica. I'll also check on Radio Shack for toroids, since I have not been able to get any ferrite cores here, so far. NickZ
Hi, Nick.
This is essentially the same as the sj1 circuit, with PNP replaced the NPN, and taking some R's out as I have also mentioned doing.
Anyway, that doesn't really matter -- great work, Gibbs.
Could you tell us -- for your ~4uW run, what resistance did you have on the pot? and where did you connect the LED?
Hi NickZ,
I didn't think the toroid would be that important. I also use air core.
Professor,
I connected the LED to coil ends and not emitter-collector to compare with the 1N4148. The pot resistance was 1MOhm on full.
Right now I'm testing on connecting a red LED from base to emitter and an additional pot from collector to coil for tuning. If I record correctly this is a looping setup, but have to test some more with different tuning and coils.
GH
Gibbs,
Your circuit is different than what the Professor is currently testing, as the reactance circuit is in the collector, vs. in the emitter. Yours is truer to the original JT circuit, other than it uses a PNP instead of a NPN. Other than that, it's pretty much identical to the original JT.
.99
Quote from: GibbsHelmholtz on 2011.06.22, 13:29:20
Hi NickZ,
I didn't think the toroid would be that important. I also use air core.
Professor,
I connected the LED to coil ends and not emitter-collector to compare with the 1N4148. The pot resistance was 1MOhm on full.
Right now I'm testing on connecting a red LED from base to emitter and an additional pot from collector to coil for tuning. If I record correctly this is a looping setup, but have to test some more with different tuning and coils.
GH
Thanks for the responses, GH. Do you see the LED go out or dim, then re-brighten with lower Pinput, at a certain voltage, with this variation?
and -- How does the air core compare with the toroid?
Please continue your investigations!
Guys:
Radio Shack does not have any toroids for sale that I can find in their web site.
Gibbs: Yes, the coil is one of the most important and variable components. The more common Iron Powder cores will work but will not give the best results. I understand that the Goldmine 1" ferrite cores are out of stock, so now what to we do? What kind of results did you have with the air cores?
I do notice the led will go out at almost no resistance level on the pot. I use 1k, 2k, and 5k pots mostly. Each pots acts differently, so the led light will also act differently when using different pots, or differnt leds, and also when the voltage drops. That also affects the output coming from the possible sweet spot. That's why its important to use a steady non variable voltage source, or the tuning of the sweet spot will be lost, as resonance will depend on a having a steady voltage.
JouleSeeker has just confirmed that his two day test run using a single AA on his device resulted in no voltage drop, or hardly any. So, that to me is even more important than the cap test he is doing. Cap tests have only shown discharging of the capacitor, while the two day AA test shows hardly no loss, at all. Both results are interesting.
Below are some of the circuits that I've been playing with lately. The first was a Kooler replication, using a tiny core with a secondary to light three leds, although I've taken one of them out. All the different cores give quite different results, using otherwise the same components. My circuits are mostly Kooler replications, but I'll try a couple circuits using the 2n3906 trans. and also some air cores, as I can't get any ferrites here in Costa Rica.
Gibbs -- I got your circuit running here, dimmer than the others perhaps but LED is lit.
Conditions:
PNP 3906
10000 uF input
390Kohm
.01uF
20 turns 22AWG magnet wire bifilar on 1" e-goldmine toroid
Runs at ~310 Hz here.
Nick -- email me your snail-mail address and I'd be glad to send you a 1" e-goldmine toroid; I have an extra.
Gotta run - my daughter is having a baby! Hurray! :)
Quote from: GibbsHelmholtz on 2011.06.22, 02:16:14
Circuit info:
PNP 3906
4700 uF input
1 MOhm pot
.01uF
28 turns 24AWG magnet wire bifilar on 5/8'' radioshack toroid (roughly measured with a ruler)
Gibbs,
Thanks for the nice circuit. As I was comparing yours with Professor's, I figured out another way to use PNP transistor in the circuit. Please see below for the simulation and circuit.
Quote from: GibbsHelmholtz on 2011.06.18, 16:45:32
Lane,
I was very impressed with your looping technique. I had a lot of fun trying to looped the JT past weeks with your technique. This is why I know COP>1 is real, but it won't be in the form of charge. One can see if the circuit reached COP>1 easily with both cap/stopwatch and dual DMM method! The cap/stopwatch method is always be less than or equal to Dual DMM method. The energy consuming from the capacitor is a real component, there is no if, and, or but about it. The dual DMM method is not totally invalid. We can use it to determine the phase angle and calculate the ratio of heat&radiation: energy dissipate from cap. It is now important to get this thing in a chamber where not only heat, but radiation can be efficiently captured.
Gibbs, Thanks a lot for sharing your experiments with this looping technique. Here again I am putting up another (simpler) looping circuit. Please see the attached. From the trace plot you see that the battery is getting charged in each cycle. Similar modification can be made to the PNP transistor based circuit to make it self-charging. You can replace a LED in place of the 1N4148 diode.
Professor,
Thank you for checking the circuit out. Congratulation on your good news.
Thanks Lane. Your PNP circuit is the professor's design. Mine is Xee's circuit in PNP.
About the looping, I am still tuning it. My experience is that there are two types of loop. One with capacitor and one without the capacitor. The one without capacitor is at much lower efficiency than the cap loop. On my circuit, after putting the LED from base to emitter, the 1 Mohm is just for show(jump start). One can remove it and let the circuit self oscillate. Red LED is very bright. The smaller cap value, the smaller the input. I've also notice the frequency is locked on the coil. Big coils low frequency, seems like it has auto coil frequency locked down function lol (not too surprised with the JTs) My 14AWG air core shows 2+ Mhz on the meter.
Quote from: GibbsHelmholtz on 2011.06.24, 03:06:39
Professor,
Thank you for checking the circuit out. Congratulation on your good news.
Thanks Lane. Your PNP circuit is the professor's design. Mine is Xee's circuit in PNP.
About the looping, I am still tuning it. My experience is that there are two types of loop. One with capacitor and one without the capacitor. The one without capacitor is at much lower efficiency than the cap loop. On my circuit, after putting the LED from base to emitter, the 1 Mohm is just for show(jump start). One can remove it and let the circuit self oscillate. Red LED is very bright. The smaller cap value, the smaller the input. I've also notice the frequency is locked on the coil. Big coils low frequency, seems like it has auto coil frequency locked down function lol (not too surprised with the JTs) My 14AWG air core shows 2+ Mhz on the meter.
Thanks, Gibbs, but we're still awaiting the arrival of the little grandchild. Another false alarm yesterday, but that is to be "expected" I guess.
I'm intrigued that you're able to to remove the 1Mohm and let the circuit self oscillate, and also that the LED is very bright. I can't accomplish either of those with my build of your design -- do you have any suggestions?? Also, what sort of Pinput are you measuring for your DUT?
Wow... interesting. I thought it was so simple to get it self oscillate. I can understand the problem between successful experiment and replication now. Let's diagnose it then.
I use a 1.2V battery. It considered dead because brand new is 1.5V+ .
Try to maintain the 1 Mohm connected to see if the LED light up.
Use big Capacitor first. Too small cap won't work.
Too big coil won't work either. I think your 20 turns is fine.
Dial the 1 Mohm down to see if LED light up.
Try low LED forward voltage drop.
Make sure you have LED connected correctly. Current should flow from base back to emitter (battery).
I was however to get my build of Xee2's circuit (NPN) to self-oscillate with the LED glowing brightly (fun!). This was running with 2 Mohms to the base, and with your suggestion, I simply disconnected this Rbase, and the DUT continued to operate unabated, LED bright.
edit: just saw your response -- thanks. will play with my build of your PNP approach a bit more...
Lasersaber has done a build of this circuit with a few mods and experimental tests. Very interesting IMO:
http://www.youtube.com/watch?v=PS2W_48_QKQ
Some time ago Kooler posted these circuits (http://www.overunityresearch.com/index.php?topic=853.msg14148#msg14148)
as variants of the "Joule Thief"
Diagram Attached to Original Post (http://www.overunityresearch.com/index.php?action=dlattach;topic=853.0;attach=5064)
Upon analysis of the circuit diagrams it appeared
that the two circuits shown at the right of the
graphic were incorrect as there is no current
path through the transistors which includes the
transformer primary winding.
Kooler, in response to a question, indicated that
he believed those drawings were indeed correct
even though they looked weird.
It seems that the negative terminal of the power
sources in those drawings is connected to the
wrong end of the primary winding. As shown,
the negative terminal of the sources are connected
to the end of the primary winding which is attached
to the transistor emitter. These connections result
in a non-functional circuits.
By moving the negative terminal of the source to
the opposite end of the primary winding (where
the feedback capacitor is connected) the transistor
current paths will then include the primary winding
and the circuits will function.
Has anyone else who's tried those circuits had
similar experiences?
My friend Nul-points has results worth noting.
He started with the sj1 circuit and made it into an apparent "self-runner" -- he has results going for over a YEAR that show a rise in voltage of the battery while at the same time flashing an LED over the course of the year+.
http://docsfreelunch.blogspot.fr/
As I've previously noted, my own interest has moved towards devices with greater output power and thus of greater use to people in developing countries and world-wide -- but I wish to sincerely congratulate Nul-points for his continued effort and successful conversion of the simple circuit to a possible "self-runner" -- although this remains to be verified/replicated.
Quote from: PhysicsProf on 2013.01.21, 13:42:57
My friend Nul-points has results worth noting.
He started with the sj1 circuit and made it into an apparent "self-runner"
...
Certainly not, no OU, not even an apparence.
The blog is entitled "informal experiments with energy anomalies", but there is no energy measurement!
How much energy in the cells and the capacitors at the begining?
How much energy in the cells and the capacitors at the end?
Is there something depleted?
Nobody knows...
I listen to the radio every day. My receiver is powered with C batteries for 3 years. 3 years is a lot. So I have an apparent "self-runner"! C.C
Fortunately among my physics professors when I was student, not one led me to a so low level. I thank them for having demanded of me much more intellectual requirement.
.
Quote from: exnihiloest
Fortunately among my physics professors when I was student, not one led me to a so low level. I thank them for having demanded of me much more intellectual requirement.
Every discussion has a beginning which consists of
certain observations. Then, as it progresses it
becomes more detailed.
With continued analysis this discussion too will
become more detailed.
Patience. In due time. Rome wasn't built in a day...
It's possible that your present level of expertise
took a rather long time to manifest as well.
Quote from: Dumped on 2013.03.16, 18:16:51
Every discussion has a beginning which consists of certain observations. Then, as it progresses it becomes more detailed.
With continued analysis this discussion too will become more detailed.
Patience. In due time. Rome wasn't built in a day...
It's possible that your present level of expertise took a rather long time to manifest as well.
You speak of "observations", therefore at none time in the process, the discussion needs affirmations that are lies, like "apparent self-runner". We don't observe "apparent self-runner". False assertion, absence of facts, real wishful thinking, unacceptable misleading speech.
We observe strictly nothing new in this setup. There is nothing to expect neither in a day nor in a century, because there is nothing to observe, and we can't build anything from the nothingness.
We shall see. The Truth Squad will assure that
any and all corrections are timely made.
All pertinent facts shall be revealed as
rigorous testing and analysis is carried out.
It may well be that there will be nothing new
in the final determination. But, then again,
who can say conclusively?
"who can say Conclusively"?
Why Ex Of course !!
It will not happen here nor there ,it will not happen anywhere ...
It will not happen on this day Nor will it be next Centuray..
It will not happen In a pulse.in a Coil nor in a boil,It will not happen
how hard you toil....
Ex does not like to make OUnity a claim he makes with complete ImPUnity...
Thx
Chet
PS
And yes Exnihiloest,I know its not completely true....You see ...
You actually do believe
@Chet! Nice Dr Zeus!