I was pleasantly surprised by this circuit and I never once thought it would get so high, and stable, in frequency. If I changed the parts I wonder how high I could get it?
If anyone knows why I can't get the trigger transformer to work please let me know as I have hit a brick wall with it.
http://www.youtube.com/watch?v=FQ5_6UXPN9I
Led was about half as bright but by adding a .005 ceramic cap in parallel with the base resistor I managed to increase the frequency of this oscillator.
Did you double-check your frequency counter reading by doing a manual computation on the scopetrace? What's the timebase setting on the scope?
.5uS I can't go any lower.
Quote from: Legalizeshemp420 on October 19, 2013, 01:08:49 AM
.5uS I can't go any lower.
OK. That is 0.0000005 second or 500 nanoseconds per division. Right?
So what is the period of a 17 MHz signal?
1/17,000,000 = 0.000000058 = 58 nanoseconds. So a 17 MHz signal should have 500/58 = almost nine peaks per horizontal division.
Right? Or did I drop a decimal point again?
Here is the signal unzoomed.
Your scope trace shows 7 peaks in 8 horizontal divisions. So that is 7 / (8 x 0.5 microseconds) = 7 / 4 microseconds = 7/0.000004 sec = 1.75 MHz.
It looks like your counter is the one misplacing decimals.
ETA: So you are using a 10x zoom in the first scopetrace? Why did you not say so when I asked you what the timebase was?
It's hard to discuss stuff, without a schematic or the correct information about your scope settings.
Quote from: TinselKoala on October 19, 2013, 01:27:43 AM
Your scope trace shows 7 peaks in 8 horizontal divisions. So that is 7 / (8 x 0.5 microseconds) = 7 / 4 microseconds = 7/0.000004 sec = 1.75 MHz.
It looks like your counter is the one misplacing decimals.
ETA: So you are using a 10x zoom in the first scopetrace? Why did you not say so when I asked you what the timebase was?
It's hard to discuss stuff, without a schematic or the correct information about your scope settings.
I didn't know what my scope was at until I just used it for the picture AND I counted almost 9 within a division.
5e-7/9=18000000 so that is pretty much what I am seeing on the scope.
Quote from: TinselKoala on October 19, 2013, 01:14:18 AM
OK. That is 0.0000005 second or 500 nanoseconds per division. Right?
So what is the period of a 17 MHz signal?
1/17,000,000 = 0.000000058 = 58 nanoseconds. So a 17 MHz signal should have 500/58 = almost nine peaks per horizontal division.
Right? Or did I drop a decimal point again?
You dropped a decimal point as 1/0.0000005 = 2mhz not 20mhz so just add one more zero 0.00000005 to equal 20mhz.
I could run up a schematic but I have no way of knowing what the inductance is of my coil. I use a toroid with 24ga (seen in the picture and video) wire wrapped twice. The toroid came from a 100watt equivalent CFL.
In my circuit you see on the video it is using a S8055 transistor but I just tried a BC337-25 and it was even better except at lower voltages where the S8055 beat it. Same beautiful waveform but about 200khz higher frequency using the BC337-25 at 1.2V
When I simulated this circuit it did not behave as irl.
I swapped R1 for a 10k pot and it seems the brightest spot is between 1.5k to 1.51k.
Interesting that I googled "Blocking Oscillator" and a bunch of results showed up, BUT, there was no cross-refference to the Joule Thief circuit. Not even on Wiki. I have no idea why this is. Do people not know that this is what the JT circuit really is? Check this for yourself.
Or, is the JT circuit different somehow? I am not arguing either way I just thought that at least Wiki would have said "See Joule Thief" when searching for "Blocking Oscillator".
I don't really care what we call it, I do love what it is able to do.
Bill
Have you looked at the current draw of that circuit?
Quote from: Legalizeshemp420 on October 19, 2013, 02:22:24 AM
You dropped a decimal point as 1/0.0000005 = 2mhz not 20mhz so just add one more zero 0.00000005 to equal 20mhz.
What?
I never performed that particular division so how could I have dropped a decimal there?
You said you had 0.5 microseconds per horizontal division, but in fact you were displaying 0.05 microseconds per division. My math is correct; you did not tell me you were using a 10x zoom setting on your scope.
Quote from: TinselKoala on October 19, 2013, 07:45:19 PM
What?
I never performed that particular division so how could I have dropped a decimal there?
You said you had 0.5 microseconds per horizontal division, but in fact you were displaying 0.05 microseconds per division. My math is correct; you did not tell me you were using a 10x zoom setting on your scope.
My point is in that picture there are 9 peaks within a division and that division is .5uS as the zoom was put back to 1x. That equals 18mhz but the peaks are not exactly 9 so the multifunction counter was correct.
Not a big deal and no egos need be bruised. Frankly I despise egos so I am only telling you what the scope settings were in the picture you see.