2 Stage Switch - No Uc, Any ideas?

Hi all,

I’m not sure if this is entirely possible, and if it is how simple it would be to implement, but I’m trying to ‘develop’ (I use that term lightly) an electronic switch of sorts, but without the use of a microcontroller. It should be obvious pretty quickly if this just can’t happen.

Basically I have one power source (in this case a lithium primary cell – 3V new) and would like to figure out a way that when it is connected power flows along one path, and when it is disconnected momentarily and then reconnected (less than 1 second interval) an alternate path (or both) is followed.

This is where my lack of knowledge comes into play. I have drawn the following diagram:



It would require logic of some degree. The ‘device’ in the above diagram requires an input charge on input 1 when a charge in input 2 is detected in order to output any power. If no charge is detected on input 2 when input 1 is connected, nothing happens. Said in another way, input 2 must be on already when input 1 is started, for an output to occur.

When power is initially applied there will be a charge on input 1, but none on input 2, meaning no output on path 2. This needs to remain the case even after the capacitor has become fully charged and starts discharging into input 2. Now, if the battery was disconnected quickly and then reconnected before the capacitor has been able to fully discharge, the device will have power at both inputs simultaneously, meaning power will flow along path 2.

This is a low power circuit. A few hundred mA (though possibly as high as 1800mA) and at most 3V. I have no idea what could be used to make the ‘device’ but I figure a MOSFET could be used for the actual switching.

Any ideas? All suggestions welcome. If it can’t be done, then it simply can’t be done. I’m just trying to think outside the box (or maybe just my box lol).

Thanks,

- Matt

hmmm ...
from what you describe i take it as an "almost" battery backup scenario .
is the "less than one second" bit a requirement or could it be (say) 10 seconds or longer ?

with simple battery backup circuits a steering/isolation diode is used .
if the main power is removed or drops momentarily , power from the back up battery is "steered" into the circuit (becomes forward biased) .

i am not sure if this is what you are describing by other words .

Sorry, I should be more specific. Everything I do is related to lights, and LEDs in particular. My particular expertise is in physical design (housing etc) and for the most part electronics made by others does the trick. For some of my own private stuff however off the shelf stuff just does not work - mostly in regards to space restrictions.

Basically put I am direct driving an LED straight off a CR123 cell. There are two paths to the LED from the battery, one with no resistance and one with added resistance to reduce the current available to the emitter.

When the light is first turned on (battery connected/switch on) current needs to flow through the higher resistance path thus providing a low mode. When the battery is disconnected and reconnected (switch off/on) quickly, I want the current to flow through both paths (combined resistance is heavily reduced and I get a high mode).

I can easily do this with a microcontroller, however I am going to the trouble to direct drive the LED (for circuit simplicity and space saving) so it would make no sense to then stick an MCU and its external components on the board. It increases costs and in my opinion decreases the reliability. Of course this other solution may do the same!

I can achieve my goals using a physical solution but the machining costs may out weigh the benefits. It is however my fallback should I fail to achieve a suitable circuit design.

- Matt

that is much clearer ...
have you thought about using an "off-on-on" single pole "click switch" ?
this way there would be no time constraints involved .
it would be a cheap , simple and elegant solution .
these types of switches are used in those combination led illumination/laser pointer units .

Or a small slide switch...

Ah, but it's not that simple lol.

The light is a 'twisty'. There is no physical switch. When the battery tube is loosened the battery does not make contact with the positive terminal and the light is off. When the tube is tightened the battery completes the circuit and the light turns on.

It's another space saving mechanism.

Look up Flip Flop circuits with transistors.. quick and easy with 5 parts..

--Edit---

Found this one that should help... Using a RC pair instead of the switch should be easy..

Thanks gremlinsa...I'll have a look though I'll probably be back here asking about something I don't understand no doubt

OK, so I finally got the time to sit down and look at this 'flip-flop' circuit. I did a bit of research and have (naturally) a few questions...

1) There are 3 types of multivibrators.

Astable and Monosable circuits will either switch between on and off constantly (astable) or after a period of time (monostable). I want a circuit that remains in a desired state indefinitely or until I make it change. Therefore I think I need to use a bi-stable multivibrator circuit.

2) I believe the use of a multivibrator will simplify my circuit somewhat.

As the flip-flop circuit has two outputs, I can use one output for high, and one output for low. The question is how can I apply this to achieve my desired behavior? I want to apply power to the circuit so that it turns on in one state (which raises another question, see below), and then cut the power for a short period and re-apply it to switch to the other state. Every example circuit I have found either has two switches (one for the set line, and one for the reset line - one for each side of the circuit) or a SPDT switch. In my application I really only have a SPST switch (power on or power off).

3) How can I ensure it will always turn on in the right mode?

Turn the circuit on from fully off and it should start in low mode. After a short power cut, the circuit should start in the other mode? Simply put how can I ensure it starts in low mode, and then after a quick off/on (say less than 1s) turn on in high mode. Anything longer than 1s will result in the first mode being activated.

4) How can I ensure the circuit can handle the power requirements?

This circuit will run off 3V and pull at most 1800mA. I'm assuming the flip-flop circuit itself could not be used as a sole power transmission path. If it can then awesome, but if not should the flip-flop circuit be used instead to control a MOSFET? Use the two states of the flop-flop (on/off) to turn a MOSFET on or off?

That's all I have for now. I'll keep reading and do some drawings to see if I ca work this out

- Matt