The various JT threads on this forum document much more efficient JT circuits than this one here. The circuit under test here is the "JTBasic" circuit using a toroid and a 2n2222 transistor. Other transistors in this same circuit work "better" in the sense of being more efficient still, and slight modifications to the circuit, such as placing a 70 nF capacitor in parallel with the base resistor, make it even more efficient. If you get into optimizing the transformer, you can reach incredibly low levels of current input while still making appreciable light in a low voltage LED JT, and I've shown several times my HVJT which lights up six 90-volt NE-2 neons in series, from a single AAA battery, or even wirelessly without an onboard battery at all.

You can look thru the forum for Pirate's and Groundloop's and LaserSaber's and some other circuits for JTs that are amazingly efficient and powerful. There are many variations depending on what you want to test. My personal favorite is still my HVJT. It's not especially efficient, I optimized for high voltage, but it's cute, you have to admit that much.

Thanks for the info. That's a nice looking build you made there. I was looking for the most optimized/recent version of the JT circuit that Lawrence is saying he believes is over unity. I wanted to test with Lawrence's exact schematic and torroid winding method, if possible. I am inclined to think that it is not over unity, but want to build it and do some measurements to see how it performs. Lawrence, do you have a link to your most recent schematic?

The following was sent to Mr. Zhou and the Atten Technical Support.  I may have missed some firmware updates etc.
 
Both scopes showed crossing 0 ref line characteristics.  But the values were very different.  Both Scopes were re-calibrated according to the Instruction manual.  (The comparison may not be fair as I used CH1 Vrms as the common point.  The frequencies were clearly different).
 
That is why I prefer to wait for the Tektronics before doing more DSO analysis.  My advice to researchers is - get access to a high end DSO.  If you cannot have one in your lab, make sure you have acess to one somewhere. 
 
God Bless.

@TK
 
On reply 771, both scopes showed CH1 Vrms = 500mV because I adjusted the DC Power Supply so that such values are shown.  The moment I touched the DC Power Supply, I would have most probably changed the actual Input Power going into the circuit.
 
I and other researchers in Hong Kong have come to the view that spending money on the 4-Ch Tektronics is worthwhile.  We may not be able to have one each.  But having one to share is worthwhile.  TK, try to have access to one.  Spending 4 hours with you old scope may not be the best use of your time and energy.  (The getting wealthier Chinese Attitude is - if money can solve the problem, spend it.  The Government is printing money to match the rate of USA!)
 
Mr. Zhou will be happy if we order the Tektronics from him.  I and my group of researchers bought 7 Atten from him.  The cost of all of them add together is less than Poynt99's scope!

 
On reply 771, both scopes showed CH1 Vrms = 500mV because I adjusted the DC Power Supply so that such values are shown. 

I see. Thank you for the explanation. But do you have any independent confirmation that you were actually supplying 500.0 mV from the power supply?
I have found that the meters on power supplies are generally not all that accurate and should be supplemented by DMMs right at the input to the device under testing.
But since you are no longer relying on your Attens, the question is moot, I suppose. Still, it's curious that the DC signal from the power supply gave readings that agree, but the JT readings disagree so much.

You should rent a scope from Mr. Bahng's company before you buy one. Why throw good money after bad, even in China? You may find that the rental gives you all the data you need, and you won't have money tied up in an expensive scope that sits on the shelf and gets very little use after this.

The various JT threads on this forum document much more efficient JT circuits than this one here. The circuit under test here is the "JTBasic" circuit using a toroid and a 2n2222 transistor. Other transistors in this same circuit work "better" in the sense of being more efficient still, and slight modifications to the circuit, such as placing a 70 nF capacitor in parallel with the base resistor, make it even more efficient. If you get into optimizing the transformer, you can reach incredibly low levels of current input while still making appreciable light in a low voltage LED JT, and I've shown several times my HVJT which lights up six 90-volt NE-2 neons in series, from a single AAA battery, or even wirelessly without an onboard battery at all.

You can look thru the forum for Pirate's and Groundloop's and LaserSaber's and some other circuits for JTs that are amazingly efficient and powerful. There are many variations depending on what you want to test. My personal favorite is still my HVJT. It's not especially efficient, I optimized for high voltage, but it's cute, you have to admit that much.

TK is correct.  Gadgetmall has a JT circuit that will light an led off an AA battery for over a year.  So does Lasersaber.  Nothing against Lawrence's efforts here but his JT's are at least 2-3 years behind the work shown in the main JT topic area.  Oh, and those are not claimed to be O.U.

Nice circuit TK.

Bill

1. You have no instrumental measurement of light output. Your subjective evaluations of "brightness" do not equal an output power measurement.

2. You have no assurance that the batteries had the same amount of energy in them to begin with. Voltage is not energy; the starting energy content is not known. You should repeat this experiment with capacitors as the power sources.

3. Just as your oscilloscopes have been shown to be inaccurate, it is possible that your DMMS are as well. You need a voltage standard for comparison. I have shown you how to obtain a reasonably cheap (65 dollars US, iirc) standard of high accuracy delivering 5 voltages of your choice. Do you understand the process of instrument calibration? One plots the instrument reading (the dependent variable, vertical axis) against the actual supplied voltage from the standard (the independent variable, horizontal axis). Five points allow you to interpolate a line, and then when you read your instrument you can look up the corresponding _actual_ voltage on the graph.

4. You have not given any assurance that your timer is not supplying power to the circuit under test. I have asked you for a schematic and other technical details of the timer so that this matter can be evaluated. A _tiny_ amount of power leakage from the timer could be more than enough to fool you into thinking you have some advantage. This issue must be addressed, not ignored.

5. Finally (for now) if your timer is necessary to achieve the effect.... then you must include the power _supplied TO_ the timer as part of your total input power to the system, whether or not any of this leaks through to power the LED.


Since no seeds that you have sown have borne sweet fruit, it is important to find out why, don't you think? We already have identified major problems with the Atten scopes that would seem to poison all fruit harvested with their aid. The Divine Wine is going off, turning to vinegar, and your seeds are wasting the Gardener's time and effort. You need to concentrate on pulling up the weeds before you can tell if your fruit is going to be sweet or sour.

Directly across the street from where I live there is a large orange tree. Every year this tree is heavily laden with beautiful large oranges; they fall to the ground and the children use them as balls, or throw them at each other in mock combat. Nobody can eat these large, beautiful oranges, because they are as sour as lemons and as juicy as a dry sponge.  Fruit is not always what it appears to be, as the taster often finds out to his chagrin.

Why not ask Lawrence to send you one of his boards for testing? One that he already has tested and found to be OU.

I didn't get any reply from Lawrence about whether he has a link to schematic or not, so I will probably just experiment with some other JT circuits instead. I do already have the schematic for the basic 2N2222 JT circuit.

I didn't get any reply from Lawrence about whether he has a link to schematic or not, so I will probably just experiment with some other JT circuits instead. I do already have the schematic for the basic 2N2222 JT circuit.

The basic schematic is what is used. If you look back in the thread you'll find the last time Lawrence posted a schematic, and IIRC it will give the number of turns on the toroid, and also maybe the toroid material. The locations of the current-viewing resistors and probe locations are also given in the schematic.

One simple variation that has given me good results is to place a 70 nF capacitor in parallel with the base resistor. I'm sorry that I can't remember who suggested this to me, perhaps it was Gadgetmall or Conradelectro. There is a wealth of information earlier in this thread (before LTseung started posting) and also in the other JT threads here.