Joule Thief

[TinselKoala]

Now I've had a look at your .pdf "explanation". Lawrence, you are way off base and you are NOT using your oscilloscope correctly. You are misunderstanding the usage, operation and effect of the AC coupling / DC coupling feature of your scope.

AC coupling is MOST CERTAINLY NOT appropriate for the measurements you are making. The AC coupling feature takes the AVERAGE VALUE of a fluctuating signal and moves that to the ZERO REFERENCE LINE of your channel, removing and ignoring the DC component (offset) and thus destroying the numerical value (for your purpose of power calculation) of the data obtained . This is the explanation for your "sine wave" on the battery... which is really a 60 Hz signal from your mains, amplified greatly to show up, not coming from the battery at all, and your citation of a "peak to peak" reading on a DC signal, and your dropping to 220 mV p-p when you have your board hooked up: the mains signal is suppressed, the actual battery voltage DC value is removed by the AC coupling setting, and you are reading the "ripple" on the DC, which is not AC at all but merely a little fluctuation sitting on top of your 1.4 volt battery voltage.

In addition, this same phenomenon of moving the average value of the signal to the zero reference can clearly be seen in your Output scope traces, and is why you think there is "negative current" when there isn't any.

You are engaging in oscilloscope abuse! You are making conclusions based on incorrect data gathering methodology, and it's plain that you should read up on AC/DC coupling functions on oscilloscopes.

Please see my video here , to see what scopes do to accomplish the AC coupling function. It really has nothing to do with "measuring AC" at all... it is about COUPLING: a DC coupled input is a straight wire linkage, an AC coupled input goes through a capacitor. That is all !! Below, under the quote from your .pdf, I show the coupling selector switches on the front panel, and the backside, of my RM503. In a DSO this function is accomplished either by a relay switching a cap in or out, or digitally by signal analysis software. You may be able to hear a "click" sound inside the scope as you change coupling, if your scope does it by the relay/cap method.

[TinselKoala]

Lawrence inspired me to make a couple more video demonstrations. 

The first one just demonstrates that the NE-2 does require about 90 volts to light up, and the second one uses my LaserSaber variant NE-2 JT, running on a single AAA cell, to illustrate the difference between AC coupling and DC coupling on the oscilloscope, and what effect it has on masking or blocking the DC component of a signal.

It's good to have nice tools. But a micrometer caliper doesn't make a very good hammer, and a digital oscilloscope will only tell you the truth.... if you ask it the right questions.

http://www.youtube.com/watch?v=EVCXvX-uugA

http://www.youtube.com/watch?v=2TKEQwG-2gY

[ltseung888]

Now I've had a look at your .pdf "explanation". Lawrence, you are way off base and you are NOT using your oscilloscope correctly. You are misunderstanding the usage, operation and effect of the AC coupling / DC coupling feature of your scope.

AC coupling is MOST CERTAINLY NOT appropriate for the measurements you are making. The AC coupling feature takes the AVERAGE VALUE of a fluctuating signal and moves that to the ZERO REFERENCE LINE of your channel, removing and ignoring the DC component (offset) and thus destroying the numerical value (for your purpose of power calculation) of the data obtained . This is the explanation for your "sine wave" on the battery... which is really a 60 Hz signal from your mains, amplified greatly to show up, not coming from the battery at all, and your citation of a "peak to peak" reading on a DC signal, and your dropping to 220 mV p-p when you have your board hooked up: the mains signal is suppressed, the actual battery voltage DC value is removed by the AC coupling setting, and you are reading the "ripple" on the DC, which is not AC at all but merely a little fluctuation sitting on top of your 1.4 volt battery voltage.

In addition, this same phenomenon of moving the average value of the signal to the zero reference can clearly be seen in your Output scope traces, and is why you think there is "negative current" when there isn't any.

You are engaging in oscilloscope abuse! You are making conclusions based on incorrect data gathering methodology, and it's plain that you should read up on AC/DC coupling functions on oscilloscopes.

Dear TinselKoala,

Thank you for your reply and knowledge on oscilloscopes.

Since I shall not be in the BSI Energy Holdings Limited office for a few days, I cannot use the same FLEET for the following test.

I set the coupling to DC for the test.  Please see the attached photo.  The oscilloscope displays the Instantaneous Voltage no matter the DC or AC coupling setting.

[TinselKoala]

Please do what I do in the video: set the vertical trace position to put the _grounded_ signal, zero volts, at the center of the screen vertically,  put your scope probe on the battery, and then show two closeups of the screen, varying only the AC coupling - DC coupling setting. You will find that I am correct.

In the DC coupled condition, you will see the 220 or 260 millivolt p-p signal sitting on top of the 1 + volt battery voltage. In the AC coupled condition, you will see this same signal down oscillating around the zero baseline, and some of it will appear to be a negative voltage.

(ETA: It looks to me like this is in fact happening: I think I can just barely make out the channel "zero" indicator on the left of the screen near the center graticule marker, and the trace is well above that, but I can't tell what the channel volts/div setting is.)

Your "numbers in boxes" -- the signal parameter readouts -- will tell you the same thing in each case: you have a 220 or 260 mV p-p oscillation. The numbers in boxes do not tell you what the center voltage of this p-p oscillation is: this depends on the coupling setting. It is up to the user- YOU- to realise the true nature of this oscillation, and at what voltages it actually occurs, and to measure it correctly.

Please try to remember to display only two or three or four complete waveforms, unless you are trying to determine frequency from the peaks. This is particularly important because the "fences" and "combs" you show later in your .pdf are full of artefact and are impossible to interpret.
Also, when you are talking about voltages, it is critical that you include the channel volts per division setting and the probe attenuation setting, and the location of the channel zero reference level, as well as the coupling AC or DC. If you are talking about frequency then the horizontal timebase setting must also be included.

What you are doing is like when a pretty girl gives you her phone number.... but the three inner digits are missing.

Also.... how do you account for that huge, perfect sine wave displayed when you are scoping only the battery, in the picture I posted from your .pdf?

[TinselKoala]

OK... I made _yet another_ video, trying to get the point across about AC and DC coupling.

This time I used a circuit a lot more similar to Lawrence's FLEET. This is a standard JT circuit, I don't know why he calls it FLEET (which is the brand name of a popular enema product in the USA, not the most felicitous choice of acronyms IMHO ho hoho...)
I'm using my low-power JT test bed, with a 2n2222a metal can transistor, an inductor from a TV set on a rectangular ferrite core, with 1:1 turns ratio and perhaps 40 or 60 turns each side, a 1K base resistor, a 1R current viewing resistor on the negative battery terminal, and some dead AAA batteries. (for some reason in the video I say "600 ohms" for the base resistor... it's 1 K though. )

It runs a bit longer than usual for my demos.... sorry about that (and the light...     )

http://www.youtube.com/watch?v=dcMNM1TemT8