Battery conserving mechanism.

[N_Cook]

PP3 ,9V powered kit, 9V for brief power use and 3.3V via HT7133A-1 VR
for the uC. Simple 3 terminal device, In, out, gnd. Datasheet for the VR
says nothing about short circuit protection, in fact shows how to add a
discrete transistor for s/c protection. Normal use of this kit , for
manual switch off , not only the obvious isolation of the 9V but also
shorts across the 3.3V line to gnd.
The 9V lines are discrete gated in use , otherwise held off.
Requires something like 3.3R across the the VR output to make it drop
out, recoverably. The VR and micro resets itself and operates normally
after such a forced powerdown
This kit auto times out,via the uC, after 30 seconds of no use but I can
find no discrete item that crowbars across the Vdd line, to stop the
3.3V supply. Do they make uC (this one anonymous under black epoxy dome)
that can internally crowbar their own supply ? Just realised I've not
checked whether there is a 9V line to the micro, but still an internal
crowbar power option for a uC?

[Ian Malcolm]

N_Cook <div...@tcp.co.uk> wrote in news:m5f46e$a29$1...@dont-email.me:

Dont be daft. The regulator quiescent current is only a few uA so the
battery life with just the regulator as a load will be in excess of 10
years. Add a MCU with a micropower sleep mode and it probably halves
that, or worse, but if it can sit there for over a year in sleep and
still wake up and work, noone is likely to complain about short battery
life!

The main power switch crowbars the 3.3V rail as well as disconnecting the
9V supply to discharge the decoupling capacitance on the 3.3V rail and
ensure a clean reset of the MCU if its briefly switched off and on again.

--
Ian Malcolm. London, ENGLAND. (NEWSGROUP REPLY PREFERRED)
ianm[at]the[dash]malcolms[dot]freeserve[dot]co[dot]uk
[at]=@, [dash]=- & [dot]=. *Warning* HTML & >32K emails --> NUL

[N_Cook]

But something kills the 3.3V supply after the 30 seconds, of normal use
, ie not manually switching off, just not tapping any keys.
Surely if the uC goes into sleep mode and then no load, the VR would
still sit there with 3.3V on the output and 5uV standby consumption.
Its not just the uC there is some other SMD 16 pinner with some other
function associated with sensing or something. I think I'll cut the
output line and see if on no load it also goes to 0V.
How normal is it to short a VR and its smoothing electro in normal
manual switch off? regardless of any uC or whatever

[Ian Malcolm]

N_Cook <div...@tcp.co.uk> wrote in news:m5fgqb$p3b$1...@dont-email.me:

That is a horse of an entirely different colour. You didn't originally
mention the 3.3V rail goes away on auto-shutdown.

> Surely if the uC goes into sleep mode and then no load, the VR would
> still sit there with 3.3V on the output and 5uV standby consumption.

Yes.

> Its not just the uC there is some other SMD 16 pinner with some other
> function associated with sensing or something. I think I'll cut the
> output line and see if on no load it also goes to 0V.
> How normal is it to short a VR and its smoothing electro in normal
> manual switch off? regardless of any uC or whatever

I cant imagine any sane designer deliberatly shorting the rail
electronically for anything other than crowbar protection circuits.
Also, there's nothing in the regulator datasheet to indicate it will
latch off.

Shorting the rail via a mechanical power switch that also disconnects the
supply is a different matter. In addition to avoiding reset problems, it
prevents the MCU being powered up via its protection diodes if any inputs
are present when the power switch is off.