Testing the TK Tar Baby

[TinselKoala]

TK,

That variable oscillator I designed around the TC4426 doesn't have a very wide range of frequency or duty cycle adjustment. It's just enough to play with things a bit and provide an adequate drive to the MOSFET. Feel free to use your FG or 555 there instead if you do build this.

I just took another look at your schematic and it looks like I might be able to build an oscillator around the 4426 driver chip (excellent choice btw, makes a killer H-bridge SSTC driver with its complement 4427.)
I don't have any on hand at the moment but I do have some Intersil mosfet H-bridge driver chips all on one chip, 16 pin DIP IIRC. But I'm about to go to the store so if I'm lucky I'll pick up some of the 4426s. Perhaps using the 555 to clock the 4426 might work to give the wide range and still drive the mosfet "properly".

[TinselKoala]

Tar Baby likes the 10R in series with the 555 pin 3 output-- thanks PW---. And my inductance meter finally arrived.

Revision B:

[TinselKoala]

Can anyone tell me what are these "Phase Shifts" that suddenly have appeared in Rosemary's imaginings...er.... data? Is she talking about Paper2Fig8 below?

If she is talking about the phase relationship between the current (voltage drop) in the CVR , and the "battery" trace (or more useful for power readings, the common drain trace) oscillations at that frequency, my demonstration of that is being uploaded now.

And of course there is a "phase shift" in these signals... it's AC at RF frequencies, and besides... current and voltage are expected to be out of phase, or rather at a basic 180 degrees difference,  plus or minus , right?

[picowatt]

Tar Baby likes the 10R in series with the 555 pin 3 output-- thanks PW---. And my inductance meter finally arrived.

Revision B:

TK,

Did you take the time to measure the drain current versus temp before and after the 10R?  Just curious, that, to me, would at least be a bit of "fun". 

.99,

I am not sure why the switching circuit and switched mosfet in the source leg is required at all.  Why not just put a 50R in the source leg instead of the switched mosfet and let it run?  The 50R would emulate RA's FG out and you an set the bias current by adjusting the applied gate voltage.  Alternately, of course, the gate voltage could be fixed and Rsource adjusted to set Ibias.

PW

[fuzzytomcat]

Hi Tk,

Here is another 555 timer circuit that was hidden in plain sight on the internet .... with a 5% to 95% duty cycle .... don't tell everyone where it is. 

Cheers,
Fuzzy