Joule Thief 101

[picowatt]

As far as frequency content (noise) go's,it has nothing to do with the frequency the circuit is running at,and to introduce such,is only adding to confusion that need not be there.
When we are talking about frequency,it is the frequency that the device is running at,or the signal it receives from the FG. If we rectify the output,then the frequency of the output will match that of the running or provided frequency--not the noise that exist within that frequency.

If my scope read's the same frequency as what is supplied by the FG,then that is the operating frequency of the device,and when we use these force switching method's in stead of running at a resonant frequency,then once again,noise that ride's on the back of this frequency has little to nothing to do with the way the circuit operates.

I chose to ignor it,because it has little to no impact to what we are doing here,and if you claim it dose-along with PW,then please feel free to show it.


Brad

Everything stated above could not be further from the truth.  It is very obvious that you do not get "it', and "it" is a very important and fundamental concept.  We are not discussing "noise"...

Your digital scope has a FFT mode to allow you to see the frequency content of a waveform (frequency domain), yet you apparently have not bothered to use it even as a learning tool (as I have suggested many times). 

When you are looking at your FG's output on your scope and you have the FG set to produce a 1KHz sine wave, both the FG and scope tell you it is 1KHz.  If you switch to a square wave, your FG and scope continue to say you are dealing with a 1Khz waveform.  But, it is obvious there is something very different because the sine wave now looks like a square wave.  The difference is that the square wave has had many harmonic frequencies (additional sine waves) added to the 1KHz sine wave (fundamental) to produce the waveshape you see as a square wave.

Read my previous post again, look at the Wiki reference (particularly the animated gif link) and try using your scope in its frequency domain mode (FFT).

PW

[EMJunkie]

You Blokes will still be arguing in a 100 Years...

PW is correct and Tinman is Correct, each from their own points of View...

PW is talking Total Wave Content, Tinman is not. Tinman is talking Fundamental Wave.

If memory serves, Sine is all Even Harmonics, Square Wave is Odd Harmonics and Triangle is all Harmonics. All Waves have a tone of real wave content that is either side Fundamental.

#59: Basics of a Square Wave signal's harmonic content

   Chris Sykes
       hyiq.org

P.S: You guys should really be working toward a common goal and not bickering over such thigns. After all, its just each others point of view. why not agree to dissagree?

[MileHigh]

Chris:

You are just "injecting noise" into the discussion and your posting should be completely ignored.  PW is dead-on 100% correct and Brad should undertake to learn and understand this fundamental concept.

MileHigh

[poynt99]

But i am a little confused about this turn off time,as the 1n4007 data sheet's all show a test frequency of 1MHz,

Indeed, but for what test is this 1MHz used for? The other test condition is a reverse voltage of 4V. Why would they do that? Why, to test C_T of course.

while here i was using only 10KHz. If it is suited to frequencies of 1MHz,then how can it not switch off fast enough for 10KHz ?.

Brad

Any assumption that the 1N4007 is suited for use at 1MHz is incorrect. It is a line frequency device. If you want a faster version, try the UF4007. It has only a slightly lower capacitance under the same test conditions, but its Trr is much faster.

As PW is saying, the frequency has little to do with how fast a device can switch. The real question is how fast can the device switch at the transition points? A 1N4007 with a 10Hz square wave (but with 10ns rise and fall times) is still not going to switch very fast from ON to OFF and back to ON, because it is limited by its own finite switching speed. Relative to the period of a 10Hz square wave, yes the 1N4007 will switch pretty fast, but relative to the rise and fall times at the transition points of this same 10Hz square wave, its switching time is comparatively slow.

The frequency will eventually catch up and become the limiting factor for all switching devices. If for eg, a diode requires 2us to recover (Trr), then the maximum frequency one should expect for it to fully switch would be about 250kHz (50% duty), but you are not going to see a nice square wave on its output; it will be somewhat sinusoidal.

The again, what do I know, and why am I posting? 

[picowatt]

Tinman was not correct...

His LED was being driven by the higher frequencies contained in his waveform well in excess of the 10KHz indicated on his scope.   

A pure "sine wave" contains no harmonics.  If it contains harmonics (as in distortion, THD, etc) then it is no longer just a sine wave and its shape will deviate (distort) from that of a pure sine.

A triangle wave, like a square wave, contains ODD harmonics, just as is indicated by the frequency domain plot EMJ posted (note there are no even harmonics displayed).  The difference being that the amplitude of the higher frequency harmonics are rolled off in the triangle wave.

PW