Post by thread:[EE:] Soft power switch circuit

Any comments on the circuit at http://www.nosreme.org/tmp/selfpwr.png? The intention is to run a PIC-based circuit (16F628) from a 9V battery, but reducing the power consumption to (virtually) zero when not in use. I also want to be able to read the status of the push-to-make switch while running. The switch turns on the MOSFET via NPN Q2, and the PIC can then keep the power on through Q1. It can read the status of the switch on another pin through R6. This is a one-off hobby project, and I already have these parts around, which is why I haven't just got a regulator with shutdown option. I have seen Roman Black's circuit at http://centauri.ezy.net.au/~fastvid/self_swi.htm, but as this circuit is likely to be running for a couple of hours at a time, I'd rather not have to sink 5-10ma the whole time. Thanks, Chris -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

At 01:21 PM 3/14/2004 +0000, you wrote: >Any comments on the circuit at http://www.nosreme.org/tmp/selfpwr.png? > >The intention is to run a PIC-based circuit (16F628) from a 9V battery, >but reducing the power consumption to (virtually) zero when not in use. >I also want to be able to read the status of the push-to-make switch >while running. > >The switch turns on the MOSFET via NPN Q2, and the PIC can then keep the >power on through Q1. It can read the status of the switch on another >pin through R6. Put another 10K resistor from the base of Q2 to ground. Use a TTL-level input on the PIC. Then it should work reliably. The 7805 wastes 5mA typically just by itself, quite a bit of drain for a 9V battery, you'd be better off with a lower-power regulator. The 78L05 and 78M05 are not that much better. If you do get a lower power regulator, you can save a bit of current by increasing R2 to 100K. >This is a one-off hobby project, and I already have these parts around, >which is why I haven't just got a regulator with shutdown option. > >I have seen Roman Black's circuit at >http://centauri.ezy.net.au/~fastvid/self_swi.htm, but as this circuit is >likely to be running for a couple of hours at a time, I'd rather not >have to sink 5-10ma the whole time. > >Thanks, Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" spam_OUTspeffTakeThisOuTinterlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

You can simplify the circuit a bit by eliminating Q1 and connecting the left end of R3 to the base of Q2. Also, put in that 10K base-to-ground resistor Spehro suggested so the circuit doesn't turn on due to transistor leakage. I've largely abandoned 7805 voltage regulators for my hobby stuff, especially for battery operation. They waste too much current and require several volts of input headroom. I now use the LP2951 low-dropout regulator from National Semiconductor; it draws only about a hundred microamps for itself (vs. several milliamps for a 7805), will regulate well with an input only a few hundred mV above the output, and it also has a handy ERR output which you can use as a PIC reset input. You can set it to 5 volts just by strapping a couple of pins together, or program it to a different voltage with a resistor voltage divider. It comes in an 8-pin DIP. IMO, it's the ideal hobby voltage regulator. HTH a bit... Dave D. Chris wrote... {Quote hidden}>Any comments on the circuit at http://www.nosreme.org/tmp/selfpwr.png? > >The intention is to run a PIC-based circuit (16F628) from a 9V battery, >but reducing the power consumption to (virtually) zero when not in use. >I also want to be able to read the status of the push-to-make switch >while running. > >The switch turns on the MOSFET via NPN Q2, and the PIC can then keep the >power on through Q1. It can read the status of the switch on another >pin through R6. > >This is a one-off hobby project, and I already have these parts around, >which is why I haven't just got a regulator with shutdown option. > >I have seen Roman Black's circuit at >http://centauri.ezy.net.au/~fastvid/self_swi.htm, but as this circuit is >likely to be running for a couple of hours at a time, I'd rather not >have to sink 5-10ma the whole time. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

At 02:49 PM 3/14/2004 +0000, you wrote: > I've largely abandoned 7805 voltage regulators for my hobby >stuff, especially for battery operation. They waste too much >current and require several volts of input headroom. I now use >the LP2951 low-dropout regulator from National Semiconductor; it >draws only about a hundred microamps for itself (vs. several >milliamps for a 7805), will regulate well with an input only a >few hundred mV above the output, and it also has a handy ERR >output which you can use as a PIC reset input. You can set it to >5 volts just by strapping a couple of pins together, or program >it to a different voltage with a resistor voltage divider. It >comes in an 8-pin DIP. IMO, it's the ideal hobby voltage >regulator. The LM2931 is kinda cool too, "load dump" (automotive transient) and reverse polarity (eg. battery) protection built in (also mirror-image insertion protection). Cheap like the LP2951. Iq is around 400uA typically, which < 1/10 that of a LM7805. All these bipolar LDO regulators are fussy about output capacitors compare the to the 78xxx regulators. They can't be too lousy *or* too good ESR, and have to be in the right general range of capacitance or you can get oscillation. They also tend to draw more current as the output current goes up (roughly 10% of output current gets wasted) because of the lousy characteristics of the lateral PNP on-chip pass transistors. Then you have the great new CMOS parts like this one: http://cache.national.com/ds/LP/LP8345.pdf Unfortunately, no reverse battery protection, and 10V maximum input. 8-( I doubt we'll see another real jellybean part like the 78xx regulators anytime soon, all the upsides seem to have downsides. Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" .....speffKILLspam@spam@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Thanks Dave and Spehro! On Sun, Mar 14, 2004 at 02:49:13PM +0000, Dave Dilatush wrote: > You can simplify the circuit a bit by eliminating Q1 and > connecting the left end of R3 to the base of Q2. I see. So even when the enable signal from the PIC is high, the base of Q2 is going to be at ~0.6V and the R4/R5 divider will still give a sensible voltage out. Thanks. > Also, put in that 10K base-to-ground resistor Spehro suggested so the > circuit doesn't turn on due to transistor leakage. Ah, so the transistors leak if the base is open-circuit? > I've largely abandoned 7805 voltage regulators for my hobby > stuff, especially for battery operation. They waste too much > current and require several volts of input headroom. I now use > the LP2951 low-dropout regulator from National Semiconductor; Thanks; I'll bear it in mind for future projects. > HTH a bit... Very much so, thanks. Chris -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email listservKILLspammitvma.mit.edu with SET PICList DIGEST in the body

Chris wrote... >On Sun, Mar 14, 2004 at 02:49:13PM +0000, Dave Dilatush wrote: > >> Also, put in that 10K base-to-ground resistor Spehro suggested so the >> circuit doesn't turn on due to transistor leakage. > >Ah, so the transistors leak if the base is open-circuit? Yep. Like all other PN junctions, the collector-base junction of a transistor has a certain leakage current (designated Icbo on the transistor data sheet). While it's usually pretty small at room temperature, it varies exponentially with temperature, approximately doubling with with every 10 degrees C increase. Without any external resistance bridging the base and emitter, this collector-base leakage just goes right on through the base into the emitter, causing an amplified current in the collector the same way a base signal current does; and at a high enough temperature, this can be enough to turn a transistor switch effectively ON. Hence the practice of putting a resistor between base and emitter; the exact resistance is not usually important, and values from 1X to 10X the base feed resistor are common. Dave D. -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email .....listservKILLspam.....mitvma.mit.edu with SET PICList DIGEST in the body

> >Ah, so the transistors leak if the base is open-circuit? > Yep. Like all other PN junctions, the collector-base junction of > a transistor has a certain leakage current (designated Icbo on > the transistor data sheet). > Hence the practice of putting a resistor between base and > emitter; the exact resistance is not usually important, and > values from 1X to 10X the base feed resistor are common. All good stuff. Just note that if you use a small value (eg 1X base feed resistor), you will halve the driving voltage available to turn the transistor on. This is usually unimportant but can make the difference between operation and no-operation when small drive voltages are used. (Proper design of the input circuit would make this obvious soon enough, but it's something that can be overlooked when just throwing things together). So eg for a 100k base drive resistor use of something like a 1M from base to ground would be advisable. Note that this tends to be relevant only when the driving circuit goes high impedance when off. If the drive circuit instead pulls down to ground when off then the base drive resistor provides a path for leakage current. RM -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email EraseMElistservspam_OUTTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

I use a lot of LDO's in set top box PSU's that I design.It is true that most of them are a bit of a problem as regards caps, specially ceramic caps across the output, but most designers do use ceramics across the power all over the place. The latest generation are much better. What to look for is the pass element feeding the output from the cillector or drain of a pnp or p channel device, though some like this are stable in all conditions. The high open loop o/p impedance causes an additional phase shift and the reg relies on the ESR of the output cap to stabilise the response. George {Original Message removed}