[ QUOTE ]
JSWrightOC said:
A MOSFET, when operated correctly, has extremely little leakage current. In fact, I would be more concerned about the self discharge of the batteries than the leakage current through the MOSFET.
MrAl, when driving any kind of a field-effect transistor with a contact closure, you need a pull up or pull down resistor (depending on if you are using an enhancement mode (N-type) or depletion mode (P-type) transistor) to keep the gate from floating around and causing erratic operation when it's supposed to be off. This resistor can be quite large (1Mohm, even) and will not pass any current when the contact is open--only when it is closed (the light is operating anyway--the resistor's current of 6 microamps is relatively insignificant!).
As for a latching circuit, a debounced momentary switch driving a CMOS J-K flip-flop would do the trick--but good luck fitting something like that in a size that small. There are several micro-mini slide and toggle switches that are available from Digi-Key, you can browse their PDF catalog and see illustrations and drawings with measurements.
[/ QUOTE ]
Hello there,
I agree that the leakage of the MOSFET wont be significant
but you dont really need a resistor to drive the MOSFET
gate if you use a SPDT switch. One contact to V+ and
one contact to V- drives the gate to fully on or
fully off with no chance of an inbetween gate state.
Im glad you mentioned it though, because im sure most
people will want to use a standard SPST switch to V+
with maybe a 10 megohm resistor to ground on the gate.
This allows a wider selection of switches to be used.
Also, while re-thinking the various solutions, i'd now
want to check the maximum gate current rating for
when the switch is first closed. The inrush current into
the equivalent gate capacitance could be quite high.
Perhaps a SM size 10 ohm resistor in series with the
switches' arm to limit the current spike. This wouldnt
add to the size much at all.
As for the latching mechanism, perhaps two small momentary
switches -- one to ground and one to V+ with a series
current limiting resistor, and a large cap connected to
the gate. One switch, when pressed, would charge the cap
turning the light on. When released, the cap would remain
charged for quite some time, untill the other switch is
closed, discharging it. This would take some testing
though to see if the cap ever reaches a very power consuming
inbetween state which could burn up the transistor.
Maybe i'll look into posting a schematic at some point.
Any other ideas or suggestions let me know.
Good luck with your LED circuits,
Al
