Electric OU: Supplement: Driving a MOSFET by LOWERING SOURCE VOLTAGE Part 4
Please be sure to watch Parts 1, 2, and 3 first.
Here I’ve taken out the light bulb (a low resistance non-linear load) and put 10 ohms of wirewound power resistors in its place. I’ve also installed a Current Viewing Resistor (aka Shunt or Current Sense Resistor) in the location specified in the Ainslie schematic.
I’m sorry I didn’t have a schematic prepared for the video. It can be viewed at
The rest of this video is self-explanatory, I believe.
The Q2 transistors in Ainslie’s circuits are turned on by a bias current supply that is more negative than the “zero” or circuit negative rail. (Part 1.)
A Function Generator can be substituted for the battery and potentiometer arrangement used in Part 1 with the same results. This shows that the current return path must be through the FG, and also that the FG acts as a negative bias power supply. (Part 2).
A mosfet with long wire inductance and a battery stack with some added inductance, using the same mosfet part number as Ainslie’s, produces the same low frequency results. The gate drive is varied by amplitude, and also by offset, with radically different results at the “load” but with identical scope traces. Some small oscillations are demonstrated. The “reveal” shows that the schematic I am using is identical with Ainslie’s but with her Q1 removed. (Part 3)
And in this video Part 4, I install the CVR, change the light bulb for a load closer to the Ainslie R and L parameters…. and voila. The full behaviour of the Ainslie-Martin circuit in terms of Q2 oscillations is fully reproduced.
It cannot possibly be denied any longer: the FG provides the current return path to the battery, which is NEVER “disconnected” as Ainslie claims, even though the Source of Q2 is “floating” as she calls it (incorrectly). And the FG acts as the power source for the oscillations.
Parts 1, 2, 3, and 4 illustrate, demonstrate, and prove unequivocally that this is true and that Rosemary Ainslie’s conception of the Q2 behaviour is completely wrong.
Incidentally, the arcane process of determining a frequency of oscillation on an analog oscilloscipe is demonstrated, and the mathematical use of the term “PER” is once again shown to apply to a division operation.