PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
OverUnity Research > Electrical / Electronic Theory and Learning Center > Joule Thief Circuits > Lawrence Tseung sent a Prototype to test... any comments?
2024-11-26, 18:39:14
News: Check out the Benches; a place for people to moderate their own thread and document their builds and data.
If you would like your own Bench, please PM an Admin.
Most Benches are visible only to members.

Pages: 1 2 3 4 5 6 7 8 9 [10] 11 12 13 14 15 16 17 18 19 20 21 22
Author Topic: Lawrence Tseung sent a Prototype to test... any comments?  (Read 342859 times)
Topic Options img
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #225 on: 2011-01-18, 05:34:51 »
Group: Professor
Hero Member
*****

Posts: 3017
Thanks for the encouraging comments.  I'll tell you -- some effects with this Tseung device surprise me.

Milehigh wrote in reply 210:

Quote
The oscillating frequency (in Hz) is approximately equal to the battery (or source) voltage (in volts) multiplied by the battery's Thévenin-equivalent resistance (or output impedance) (in ohms), divided by the transformer's mutual inductance (in Henrys).

Frequency of oscillation is approximately equal to:  V_battery x R_battery_output_impedance / Transformer_mutual_inductance

In the above example, given a 1.5v battery with a 1 ohm output impedance, the transformer's mutual inductance needs to be approximately 14 uH for the circuit to oscillate at 50 kHz. The oscillating frequency is principally a function of these three parameters, and is approximately linear, although this equation is only an approximation.

  I decided to run the Tseung DUT with a power supply -- the output values were much the same as running with a battery.  I had thought several days ago to vary the input voltage to see what would happen, and did this today.  Very exciting response, really.

 Now, MH -- you say that with your equation (approximation, granted) that decreased V_battery will decrease the frequency.  This is the OPPOSITE  to what I observed today when I lowered the voltage from the power supply (PS).   Here are observed data (I feel I need to re-emphasize that experimental data trump theory):

 Input voltage (PS)  Frequency   Input LED  Output LED
   1.40                        64 KHz      Bright          Bright
    0.70                     112 KHz      Dimmer        Bright
   0.52                      177 KHz      Dim             Bright, somewhat dimmer
   0.46                      259 KHz      Almost out.     Bright still, though somewhat dimmer to the eye.  I should like to have a measurement of the lumens...
   0.39     Oscillations die; both LED's out.

Above with "constant voltage" on the PS.  When I tried to limit the current, at about 1.4 volts roughly, the DUT would not oscillate stably (tried it once or twice).

Now, we could talk about the input and output voltages if any are interested  -- and is there a way to post the oscilloscope traces here?  Very interesting! particularly at the end, at 0.46 volts  It really made my day!  and I hope you folks will not tease Lawrence.  This is quite a little device, I think -- very instructive.

OK -- Mile High, you (or your equation) say that the frequency should go DOWN as the voltage goes DOWN -- that is shown to be 100% wrong by the experimental evidence.  Further, the effect is non-linear -- again contrary to your equation.  I'm losing confidence in your equation, and therefore in your "theory."   Explain the empirical results (if you can).  

That said, I think it is useful to bring up an equation as you did even when it is shot down by the data -- love when that happens actually.  Makes it fun.

It is obvious  to me from the LED's that the input power is going down and yet the output power remains sufficient to keep the output LED brightly lit.

  
« Last Edit: 2011-01-18, 15:42:23 by PhysicsProf »
   
img  MileHigh
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #226 on: 2011-01-18, 06:46:09 »
Group: Guest
Hi PhysicsProf:

Great to see that you are getting your hands dirty!  lol

For starters, I was simply quoting from this Wikipedia link:

http://en.wikipedia.org/wiki/Joule_thief

I can't be 100% certain that the Wikipedia entry is correct, but it looks legit.  Note that I was quoting a description that describes the operation of a generic Joule Thief, and your circuit is not a generic Joule Thief.  Also, if we assume that your power supply voltage could not be pinched off by the Joule Thief transistor, then it could imply that your Joule Thief may have been oscillating in a different operating mode as compared to connecting a battery to it.

So I think it's too soon to say anything definitive because it looks like you were comparing proverbial apples and oranges.

It would be awesome if you could post some oscilloscope traces.  I believe that all that you have to do is upload your jpegs to a file sharing website like rapidshare.com.  Then you can use the URLs for the images and post them here.  There is an "attach image" icon on the toolbar.  Please resize the images first if they are very high resolution.

If I may offer you some advice a helpful tool would be to probe around your circuit and then draw up a timing diagram for it.  It should make it easier to explain your voltage vs. frequency observations.

MileHigh
   
img  ltseung888
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #227 on: 2011-01-18, 07:15:44 »
Group: Guest
Dear PhysicsProf,

If the file size of the photos or screen shots are too large, I normally use either the Paint program or the PowerPoint program to reduce its size.  Most of my pictures posted in the overunity and in this forum are done that way.  Some resolution will be lost but you always have the original.

Also, if you lower the Input Voltage, you are likely to get a higher COP!

Continue your great work.  See if you can have two oscilloscopes to do the conclusive Output and Input Power tests.  Meanwhile share some of your exciting screen shots or photos.

Lawrence
   
img  poynt99
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #228 on: 2011-01-18, 13:39:49 »

Group: Administrator
Hero Member
*****

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

When you are making a post, there is an "Additional Options..." link at the lower left of the message window. Click on that and you can browse your computer to attach photos and files.

Any luck with the second scope and/or probes?

.99


---------------------------
"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #229 on: 2011-01-18, 16:01:34 »
Group: Professor
Hero Member
*****

Posts: 3017
Thanks for the comments... Will see what I can do re: attachments as time permits.

@Lawrence:  yes, clearly the FLEET index increases as the input voltage is lowered.  There is a point near 0.40 volts for this DUT that the input LED is evidently off, yet we can still see the ringing in the system using the oscillosope and the output LED glows quite brightly.  As one lowers the voltage to about 0.39 volts, the system goes off, stops ringing.  Bringing the voltage back up, it does not re-start ringing until about 0.5 volts.  Fascinating.
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #230 on: 2011-01-18, 16:27:19 »
Group: Professor
Hero Member
*****

Posts: 3017
  I should add that as the input voltage is lowered to the critical point, the current in both input and output circuits goes down.  The output P-P voltage is still about 4.1 volts when the input voltage is 0.46 volts from the PS, so this seems to be what is driving the output LED to glow brightly (input LED hardly visible at all at this point).
 
  But currents in both input and output circuits are low, in the several milliamps range (PS at 0.46 V).  I'm thinking I'd like to lower impedance in the input circuit, to get more current flowing in both input and output circuits even at low input voltage...  Any ideas how to do this, while keeping the windings in the ferritic-core toroid?  That little 2n2222 transistor is a champ, by the way, to keep the system oscillatiing at 0.40 input voltage...
   
img  ltseung888
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #231 on: 2011-01-18, 17:57:13 »
Group: Guest
Quote from: PhysicsProf on 2011-01-18, 16:27:19
  I should add that as the input voltage is lowered to the critical point, the current in both input and output circuits goes down.  The output P-P voltage is still about 4.1 volts when the input voltage is 0.46 volts from the PS, so this seems to be what is driving the output LED to glow brightly (input LED hardly visible at all at this point).
 
  But currents in both input and output circuits are low, in the several milliamps range (PS at 0.46 V).  I'm thinking I'd like to lower impedance in the input circuit, to get more current flowing in both input and output circuits even at low input voltage...  Any ideas how to do this, while keeping the windings in the ferritic-core toroid?  That little 2n2222 transistor is a champ, by the way, to keep the system oscillatiing at 0.40 input voltage...

Increase the 1K ohm resistor to 33K ohm.  That was one of our successful attempts.  You can try other values.  You can also try to put a capacitor in parallel with the 1K resistor.  We had to try different capacitor values.  Some suggestions are available at the overuntiy.com Joule Thief thread.  You will probably find that the Input Current will be much lower but the Output current will not drop that much.  That is another trick to achieve high COP.


At this stage, focus on getting a value  of COP >1.

Looking forward to your success.
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #232 on: 2011-01-18, 19:43:38 »
Group: Professor
Hero Member
*****

Posts: 3017
    Per .99 suggestion, trying a graph attachment.  This is the input to Lawrence's prototype A, with power supply set to 1.4 volts -- the input signal is chopped and modified by the JT circuit:
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #233 on: 2011-01-18, 19:47:46 »
Group: Professor
Hero Member
*****

Posts: 3017
  Note that I did a screen shot to get a .png -- hope it's sufficiently legible.  One can clearly see the waveform I think.

Next the voltage drop on the 1 ohm INPUT resistor (i.e., the current):
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #234 on: 2011-01-18, 19:53:11 »
Group: Professor
Hero Member
*****

Posts: 3017
  Next the output voltage across the secondary-winding, with 1 ohm resistor in series with the output LED on the output side -- again, with 1.4 V in from the PS:
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #235 on: 2011-01-18, 19:56:37 »
Group: Professor
Hero Member
*****

Posts: 3017
Finally, the voltage drop across the 1 ohm resistor in series with the output LED on the output side ("output current") -- again, with 1.4 V in from the PS:
   
img  humbugger
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #236 on: 2011-01-18, 20:22:29 »
Group: Guest
Quote from: PhysicsProf on 2011-01-18, 19:43:38
   Per .99 suggestion, trying a graph attachment.  This is the input to Lawrence's prototype A, with power supply set to 1.4 volts -- the input signal is chopped and modified by the JT circuit:





This is the voltage measure across the input power supply?  I cannot see the scale factors at all, but it looks like your supply is very high impedance at high frequencies.  Can you point out the zro line and the nominal supply DC voltage line here?

It sure looks like your probe, including the ground lead, is picking up those spikes or your P.S. is high impedance (not a good voltage source).  I find it extremely odd that only this measurement shows any big spikes.  If there are no spikes in any of the other waveforms, where do these come from?  All the other waveforms look quite smooth and clean, but this one has lots of different levels and huge spikes.  Quite strange.  If you pinch your probe ground lead between your fingers and move it around in different positions and those spikes change, then you are not set up right and not getting a true "pictue" of the voltage dynamics.
« Last Edit: 2011-01-18, 21:11:10 by humbugger »
   
img  ltseung888
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #237 on: 2011-01-18, 20:56:59 »
Group: Guest
Quote from: PhysicsProf on 2011-01-18, 19:43:38
    Per .99 suggestion, trying a graph attachment.  This is the input to Lawrence's prototype A, with power supply set to 1.4 volts -- the input signal is chopped and modified by the JT circuit:



Dear PhysicsProf:

Pictures are worth a thousand words.  When you have digested the set up of the good oscilloscope, please display the Instantaneous Input Power Curve using the multiply function.  Use the “mean function” to produce a running Input Power value.  I believe that you can display them on the same screen shot.

Repeat that for the Output Power.  If you have only one oscilloscope, you may do them separately.  Critics will say that the configuration may have changed during the reconnection.

The best scenario is to have another identical oscilloscope and display both the Input and Output power side-by-side.

We all look forward to your mastering the multiply and mean functions of your oscilloscopes.  The screen shots will be the conclusive proof of COP > 1.

Prototype A has the lowest Tseung FLEET Comparison Index.  Other better prototypes will be even more convincing.  If possible, display them all as comparisons.

If you have two oscilloscopes for the simultaneous display, you can also show the effect of tuning via a video.  Move the wires or place an object near the prototype.  Both the frequency and the Power waveforms will change.  You would have then mastered the secret of building and verifying the FLEET prototypes.

Welcome to the Party. 

Lawrence
   
img  wattsup
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #238 on: 2011-01-18, 21:01:46 »
Hero Member
*****

Posts: 520
@PhysicsProf

Have you noticed when you lowered the applied voltage and saw the same brightness, while seeing the frequency increase, did you notice anything regarding the pulse width?

The fact that the output LED is still "Bright" when lowering the voltage tells me this DUT may be a good candidate for the cap/cap test method since a lowering of the voltage is what one would expect with a cap feed and the DUT is still providing a reasonable output level. If there was a germanium diode and an output capacitor instead of the output LED, you can then take the final measurements and "maybe" add 0.2 volts to the final output cap reading to compensate for the germanium diode. This should at least provide some idea on the OU status.

I am still asking if there is a build spec anywhere on this DUT.

wattsup


---------------------------
http://etherimpress.com
   
img  MileHigh
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #239 on: 2011-01-18, 22:21:45 »
Group: Guest
PhysicsProf:

Permit me to make a few suggestions about improving the information conveyed in your scope shots.

1.  The first go round was good, but you really need to eliminate the whole desktop and give us a larger view of the scope display only.  Hopefully we will be able to see the ground level, the voltage scale, the time base, and the virtual graticule.  You must have some sort of a "select a rectangle" tool for grabbing an image of your desktop?

2.  I hope that was "Plan B" and "Plan A" is to take pics of your real scope display and upload those pics.  As a bare minimum you might have to annotate the pictures of the real scope display to show the ground reference if it is not visible.

3.  You need to upload a schematic that shows where your test points are.  For each test point you have to show exactly what two points in the schematic your scope probes are measuring across.  This may seem 'silly' but I don't take anything for granted here.

4.  The best thing would be to pick a waveform that you think is a good reference waveform.  For example the waveform in post #234, the voltage across the secondary, is a good reference waveform.  Then whenever you take a pic of a new waveform, you always include the reference waveform in the picture.  In other words, always use both channels of the scope and always show the "new" waveform relative to the reference waveform.  The reference waveform can be set up to be much smaller than the "new" waveform if you want.

I realize that this is really my personal "wish list" and it takes some effort to do but the rewards are there in that the data presented will be clearer for everyone, the most important person being you.

MileHigh
   
img  ltseung888
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #240 on: 2011-01-18, 22:30:54 »
Group: Guest
Quote from: wattsup on 2011-01-18, 21:01:46

I am still asking if there is a build spec anywhere on this DUT.

wattsup


Dear wattsup,

I deliberately do not provide a build spec.

The reason is that this is a pseudo resonance device.  Small changes can affect the results tremendously.  PhysicsProf demonstrated in his post citing the drop of the Tseung FLEET Comparison Index from over 20 to less than 1 just by moving the components to a breadboard.

In Hong Kong, at least one Company was interested in “mass producing” an educational prototype initially.  They would do the necessary quality control, testing and have certificates, pictures and videos of the waveforms.  They hoped to charge US$1,000 for each guaranteed prototype.  They were willing to exchange guaranteeing that the prototype will have COP > 1..

I do not know what happened to their plans.  But the PhysicsProf followed my post on reply 80 and managed to create his own FLEET – much better than Prototype A.  At least thirteen groups in Hong Kong and China have done that.  Many of them are experimenting with different transistors, different toroids, different LCR values and a few are rumored to have achieved Output Power greater than 1 KW.  At least one told me that they will not reveal any information until they have products ready for market.

You can act on existing information and share the inventor’s joy in playing with your own FLEET prototype or wait to buy a working product.
   
img  humbugger
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #241 on: 2011-01-18, 23:16:27 »
Group: Guest
I don't think MH was asking for "a build spec".  A schematic with no part values or component details would suffice.  It seems odd, though, that you purport to be an open source save the world promoter and a humble old guy eager to give all credit to God, yet you tell folks to just watch and wait to buy your gizmo.  Seems like a contradiction.

What is the point of having the good professor present his testing here if no one can tell just what and how he is testing? The refusal to release even a blank schematic seems a bit constipated.  Maybe you need a FLEET.  (If you don't know what this is, check with your pharmacist at your local drug store).
   
img  catlady
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #242 on: 2011-01-18, 23:35:06 »
Group: Guest
http://www.fleetlabs.com/

Here is hoping more Forum members don't waste their precious time.


 ;D
Catlady
« Last Edit: 2011-01-19, 04:19:47 by catlady »
   
img  humbugger
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #243 on: 2011-01-18, 23:49:29 »
Group: Guest
Quote from: ltseung888 on 2011-01-18, 22:30:54
Dear wattsup,

I deliberately do not provide a build spec.

The reason is that this is a pseudo resonance device.  Small changes can affect the results tremendously.  PhysicsProf demonstrated in his post citing the drop of the Tseung FLEET Comparison Index from over 20 to less than 1 just by moving the components to a breadboard.



This is why some of us with a lot of bench testing experience are wondering if things are "kosher" on the professor's bench, Lawrence!
   
img  muDped
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #244 on: 2011-01-19, 00:02:33 »

Group: Tinkerer
Hero Member
*****

Posts: 3055
Quote from: humbugger on 2011-01-18, 23:16:27

...A schematic with no part values or component details would suffice.  It seems odd, though, that you purport to be an open source save the world promoter and a humble old guy eager to give all credit to God, yet you tell folks to just watch and wait to buy your gizmo.  Seems like a contradiction.
...

There should be no need for anyone to "wait to buy
[a] gizmo."

The circuit under discussion is amazingly simple.  Even
the several variants are simple and require few components.

The average person should be able to fabricate a suitable
equivalent for less than $5.00 (U.S.) even with new parts.

The tiny new "supertransistors" which exhibit very low Vce(sat)
(BISS) do work exceptionally well in this circuit.

Any who may have some old germanium transistors will have
great fun.

A scope is a necessity for performing the "fine tuning" in order
to achieve maximum output with minimum input current.

For higher powered versions a MOSFET is advantageous.


---------------------------
For there is nothing hidden that will not be disclosed, and nothing concealed that will not be known or brought out into the open.
   
img  humbugger
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #245 on: 2011-01-19, 00:16:27 »
Group: Guest
Quote from: Dumped on 2011-01-19, 00:02:33
There should be no need for anyone to "wait to buy
[a] gizmo."

The circuit under discussion is amazingly simple.  Even
the several variants are simple and require few components.

The average person should be able to fabricate a suitable
equivalent for less than $5.00 (U.S.) even with new parts.


My comment was in regard to MH's request that we see a diagram showing where exactly in the circuit the test probes were being placed by the professor.  That's not likely to happen without a schematic involved.  Lawrence answered by saying (to Wattsup) he would not allow publication of his "build spec".  If, as Lawrence says, the circuit and test lashup is so finicky that COP (excuse me...FLEET index) abruptly changes from below 1 to over 20 by moving the components to a breadboard, then any sharing of test specifics with those not sitting next to the professor at his bench will require at least some information about where the probes and ground clips are.  If the scope probe ground wire is laying across the toroid, for instance, it will show a remarkably different waveform than if it was short and far away from the toroid.  There are many things that can give false pictures on a scope.
   
img  muDped
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #246 on: 2011-01-19, 00:34:35 »

Group: Tinkerer
Hero Member
*****

Posts: 3055
Quote from: humbugger on 2011-01-19, 00:16:27
...
If, as Lawrence says, the circuit and test lashup is so finicky that COP (excuse me...FLEET index) abruptly changes from below 1 to over 20 by moving the components to a breadboard...
...


Discovering why this is so and what causes the apparent
"change" in circuit operation is a fascinating exercise.
Some have already hinted at what may be the cause.
Additional verification would be helpful.


Quote from: humbugger on 2011-01-19, 00:16:27
...
There are many things that can give false pictures on a scope.

Then we all must learn to use the scope in such a way that
any "false pictures" are precluded and their possible generation
is prevented.  Hopefully, progress is being made towards that
end.


---------------------------
For there is nothing hidden that will not be disclosed, and nothing concealed that will not be known or brought out into the open.
   
img  MileHigh
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #247 on: 2011-01-19, 01:26:50 »
Group: Guest
Personally I don't see why having the schematic should be an issue.  It's evident that it's a Joule Thief type of circuit with an extra secondary winding around the toroid driving two resistors and an LED.  There are also photographs of it provided by Lawrence and he stated the component values!  There are Joule Thief threads on OU and the EF that are hundreds and hundreds of pages long.  On OU the Joule Thief crowd have been playing with adding extra secondary windings for a long time.  People have also used Joule Thief circuits based on larger toroidal cores that can light up CFL's with the guts stripped out of them.

I can't tell anybody what to do but it looks like we would all benefit by having the proper information about this particular version of a well known circuit.  The two challenges are to understand how it works and to make measurements to check for over unity.  It's a mistake to presume anything at this point.

MileHigh
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #248 on: 2011-01-19, 05:34:15 »
Group: Professor
Hero Member
*****

Posts: 3017
  Lots of good comments.  I'm climbing the learning curve regarding all this -- Joule-Thief circuits, oscilloscopes with fancy math, displaying data intelligibly... 
Thanks for your patience.


And I'd really like to spend more time working with the DUT -and with a good scope (lots of other duties in life despite -- or perhaps because of -- being early-retired)...

  One thing -- I see Lawrence as out to seek confirmation/checking -- not the big bucks.  I don't think that is his motivation.  And it certainly isn't mine.
Let me be clear:  I am willing to fight one more battle to help a nascent science/technology get off the ground and hopefully into the hands of the millions worldwide, without being stopped by the big-oil boys.  Was I too frank?

  I should note that I now live about seventy miles from my university, and getting access to the Tek 3032's on campus takes arranging and travel.  Its a LONG commute, 140 miles round-trip.  The BitScope that my colleague and I are using (posted displays) is his, on his little PC lap-top...

If I had my own wish list (MH), I would get a lap-top (Mac) based oscilloscope with quadruple inputs and the ability to multiply V and I traces and calculate power, and from that, the mean power.  Will settle for 2 lap-top based units with dual inputs and fancy math (multiplying traces).   Does such a measurement tool exist?  I have some savings...    ;)
   
img  PhysicsProf
Re: Lawrence Tseung sent a Prototype to test... any comments? « Reply #249 on: 2011-01-19, 05:44:04 »
Group: Professor
Hero Member
*****

Posts: 3017
  PS -- I plan to get a decent scope-shot of the input V tomorrow for you all, using a battery input to compare with the result from the PS.  All the spikes on the input V are strange, I agree...

And the photo from Lawrence (which I displayed earlier in this thread) IS labeled -- I think this gets you started on your own build... but yes, will see if we can generate a schematic for you.  Not that all devices will work the same! I hope that's clear...  So in that sense, we're all climbing a learning curve with this amplified JT circuit.
   
Pages: 1 2 3 4 5 6 7 8 9 [10] 11 12 13 14 15 16 17 18 19 20 21 22
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-26, 18:39:14