Joule Thief: PCBJT Offset Current, Frequency Dependence on Input Voltage
I’m varying the input voltage to the PCB JT and monitoring the input current on the Tek 2213a scope and the oscillation frequency on the Philips PM6676 frequency counter. Power is supplied by the little HP 721a PSU and the input voltage monitored on the ElCheepo Cen-Tech DMM.
Errata: At a couple of places I say “milliOhms” where I should have said “milliAmps”. The current viewing resistor is 1.0 Ohms, so a voltage drop of 50 milliVolts indicates a current of 50 milliAmps through the resistor.
A reminder: The circuit requires a “reversed” orientation of the input current monitoring scope probe wrt its “ground” or reference. The probe tip is at the circuit point called “A4” — the negative battery terminal — and the probe reference is at “A2” on the circuit side of the input CVR. This means that a negative reading on the probe indicates ordinary current flow, out of the battery and dissipating power in the circuit. A positive current reading here would indicate a reversed current flow, presumably returning some energy to the battery.
So, is there some “sweet spot” where the behaviour changes or is there a smooth continuum without any reversed current anywhere? My control of the input voltage using the system shown here isn’t precise enough to tell. I would like to examine this issue further so I’ll be building a simple adjustable voltage divider that should give me more stable low voltage adjustment.
At one point earlier I had reported getting a current trace that crossed the baseline and showed a view just like LTseung’s traces. But I’m now unable to repeat that…. except by switching the channel input coupling to “AC”. Perhaps I did this inadvertently when making the earlier measurement; it’s certainly a possibility.
For the newcomers: The circuit and the “back story” can be found in my earlier JT videos, just look in my channel and you’ll find them. This is the same circuit that LTseung is sending out to businesses and universities and individuals around the world, except I’m using a crude PCB instead of his point-to-point wiring on a pad-per-hole prefab PC breadboard. The degree of “crudeness” is comparable, but my layout is tighter with less stray inductance likely.
Does overunity behaviour of a 2n2222 JT depend on stray inductances? It won’t be the first time.