Thread: Timer Switch
View Single Post
  #15   Report Post  
John Fields
 
Posts: n/a
Default

On Sun, 31 Jul 2005 23:14:58 +1200, Jasen Betts
wrote:

In article , John Fields wrote:
On Sat, 30 Jul 2005 12:49:55 +1200, Jasen Betts
wrote:

In article ,
wrote:
Hi,

I have a camera hooked up to a 9v battery and want to be able to use a
'timer switch' to turn it on at a specific time ( like the ones you
plug into a wall socket and then plug in your light socket and get it
to switch your light on automatically at a certain time when your away
)

Is there such a thing that can do this with a battery as the power
source?

a battery powered alarm clock with a few modifications could do that.

what you'd need to do is remove the speaker and put a little amplifier
in there to drive a relay.

.
. something like this maybe to supply +ve

---
Redrawing your circuit for clarity, here's what you're advocating:


actually no.. the 100k goes between the series diodes and the +V
like this:

+V

. R1 |
. +--[100K]----------+--------+
. | |K |
. | [1N4001] [COIL]
. | | |
. | +--------+
| |
| C
IN---[100pF]--+----|----+--[10K]-------------B (NPN)
| | |K E
[100K] | [1N914] |
| | | |
IN------------+----+----+--[1N914]--[1N914]--+
|
GND


which makes no sense, to me. Would you please explain how it's
supposed to work?


the two 914 diodes in series bias the transistor right to the edege
of its conducion region,

the capacitor and the third 1N914 diode form a diiode pump that'll
push it over the edge into conduction in the presence of an small-ish
AC signal of a few hundered milivolts.. like a battery clock would
use to power a piezo chime...

that's my thinking anyway,

the 1N4001 is a freewheeling diode to protect the transistor.
on second though that top 100K should probably be a 10K


---
1.
Since this thing is supposed to be battery operated, I'd be a little
reticent to use your biasing scheme since it throws away power.

Using a 2N4001 for the transistor and 100K for R1, the quiescent
current being drawn from a 9V supply is 1.05mA. Reducing R1 to 10K
increases the quiescent current to 5.38mA.

2.
Since you have nothing to store charge, the 100pF cap is going to
act like a differentiator and push charge into the transistor's base
on the rising edge of the alarm signal but, on the falling edge,
it'll pull the base as negative as the third diode will let it go,
so the relay may never get a chance to make. A conventional charge
pump would look something like this:


IN--[C]--+--[DIODE]--+----DC TO LOAD
|K |+
[DIODE] [CAP]
| |
IN-------+------------+


3.
Assuming that the clock's alarm signal is a 1V, 1000Hz square wave
with pretty crispy edges means that, if it's feeding a 10000 ohm
base resistor, the signal going into the base will be a
positive-going spike with a time constant of about 1 microsecond.
That means that if the spike rises to 1V initially, 1µs later it'll
be at about 330mV. Since you have the transistor conducting
slightly with your bias network and the alarm signal cap-coupled,
you'll probably be able to pump a little current into the base even
ten time consytants downstream, but that still leaves you with no
signal going into the base for 90% of the time, so there's no way
the relay's going to be made looking at that kind of a duty cycle.

As if that wasn't bad enough, there's the problem of drive. If we
assume that we can get 1V excursions from the alarm, we'll be
driving 1V into the 10k ohm base resistor, which means a base
current of 100µA on top of the 5µA or so that's already flowing in
there from the bias network, so even if you use a sensitive reed
relay that only needs 10mA of current to make it work, the
transistor would need to have a beta of 100 or so. If it was a
steady DC it might work, but I'd prefer to force the transistor to a
beta of ten to make sure that everything is nice and saturated.

So what's the solution?

It's hard to say because the OP hasn't stated whether he wants to
take a single picture or several over a period of time, but I've
posted a general scheme which might work to abse under "Alarm clock
interface"

For battery use I'd switch the comparators to micropower and change
some component values and, depending on what the OP wants, maybe
make some circuit mods.

--
John Fields
Professional Circuit Designer