uxomm:
Normally you need 2 transistors (or MOSFETs).
This may work for low current needs (less than 100 mA):
Awesome circuit! A couple of suggestions:
- The BC557 has an ABSOLUTE MAX IC of 100ma and the datasheet specs the ß at -2ma, and the Static Characteristic curve diagram stops at 50ma, this really doesn't seem a suitable transistor for this application [unless current demand doesn't go above, say -35ma]. The PN2907 [or even 2N2907] seems a better choice, at least for higher currents up to 100ma.
- Since this is for a battery powered device, I suggest increasing the value of R1 a bit. One can argue that to turn a bipolar transistor "fully" on [i.e. saturated], one should set the base current in terms of ß=10. But, there's some wiggle room if one balances power loss, in the transistor, with current drain in the Emitter-Base leg. And, this is especially "wiggle happy" with cases of low power demand -- as is likely the case here. The reason for increasing the value of R1 is to reduce the current it drains away while this circuit is powered. But, if this is considered trivial, then no need to change it. Another way around this, is to use a P-Channel MOSFET, instead. The Source pin goes where the Emitter pin is, the Gate where the Base is and the Drain where the Collector is -- then add a 100k resistor from the Source to the Gate. Then there will be no wasteful current path. Plus, a MOSFET can be chosen to have FAR less voltage drop, and so it will not heat up at all. The SFT1342-W looks like a good choice. Relatively inexpensive, logic level [gate threshold of -1.2V to -2.6V], RDS(on) of 96mΩ at VGS = -4V [and even at VGS = -2.5V the VDS is -0.1V at a Drain current of half an amp! And at VGS = -5V, the VDS is more like 20mV at half an amp] and, its a 60V transistor, so plenty of headroom
True, it's a transistor designed for much higher currents [up to -12A], but it has a very low channel resistance, thus it will drop very little voltage at the small currents likely involved here. I looked at smaller, (TO92, for instance) MOSFETS but I kept seeing RDS(on) in the the 10Ω range. At 100ma, that's a voltage drop of 1V! And, I couldn't find anything that wasn't SMD in the 1Ωish range, that were truly Logic Level (with gate threshold below -4V). Since this one is less than a dollar, then why not? It probably has a higher gate capacitance, but switching speed is not a requirement -- thus no need to worry about high gate drive current -- and that 1k resistor in the Gate leg will fix the max current to 5ma.