Only-One-Shot Circuit

[ErnieF]

Need a circuit that, when triggered by a high-low transition, will issue one and ONLY one 1-second pulse. Would be reset by a pushbutton switch or by cycling power (preferred). Powered by batteries and 2950-type 5V regulator (slow power-on rise time.) Triggered by a proximity detector, Azoteq IQS127D-00000TSR. Suggestions?

 

[KlausST]

Hi,

Google for "one shot circuit"
Or use any small / cheap microcontroller..and a few lines of code.

Klaus

 

[dick_freebird]

You want a "non-retriggerable monostable
multivibrator" such as 74xx221. 74xx123 is a
retriggerable one, useful for things like
switch debounce.

 

[ErnieF]

Thanks KS and D. Have considered Arduino; too expensive and complex for my app. For D, my understanding of retriggerabls vs. non is that nonretriggerable cannot be retriggered DURING the output pulse, i.e. extend the pulse, but EITHER circuit CAN be triggered again after the output pulse has ended. If true, it is not a "one and ONLY one pulse" solution. Is my understanding wrong? Clarify? Thanks.

 

[BradtheRad]

I guess you know the popular 555 timer IC can do the job.
To stop it from triggering a second time, put a memory cell before it. This can be an RS flip-flop made from transistors or logic gates. It changes state one time when you trigger an input, then holds that state despite further triggers.

 

[ErnieF]

YES! Thanks B! My current design uses a TLC555 one shot and SN4027B JK flipflop in Toggle mode. The flipflop would control the Reset input of the 555 High and allow only one output pulse, i.e., the FF would be toggled by the 555 output pulse, setting the 555 Reset LOW, i.e, disabled. BUT, the flipflop state at powerup is random/unpredictable, but MUST be Qout High every time the circuit is powered, to enable the 555. The circuit is (must be) simple and unclocked. Have not found a way to do that.

 

[BradtheRad]

Does your flip-flop have a reset pin? This can be briefly activated at power-up by a capacitor-resistor network across the supply rails. Connect the reset to the join of the C & R.

--- Updated Mar 12, 2021 ---

Or consider adapting the RS flip-flop concept. If you make resistor values sufficiently lopsided, then output is always low at power-up. It goes high at the first throw of the switch, then is unaffected by further throws of the switch.

 

[FvM]

YES! Thanks B! My current design uses a TLC555 one shot and SN4027B JK flipflop in Toggle mode. The flipflop would control the Reset input of the 555 High and allow only one output pulse, i.e., the FF would be toggled by the 555 output pulse, setting the 555 Reset LOW, i.e, disabled. BUT, the flipflop state at powerup is random/unpredictable, but MUST be Qout High every time the circuit is powered, to enable the 555. The circuit is (must be) simple and unclocked. Have not found a way to do that.

Yes, the intended function refers to a combination of a bistable and a monostable circuit, can use discrete transistors or ICs. Defined power-up of the bistable circuit can be achieved by circuit dimensioning (asymmetrical resistor values in a discrete flip-flop) or a dedicated power on-reset, provided e.g. by a RC circuit.

 

[danadakk]

You can use block language to do an easy capable design with Nano board (xtal controls
timing accuracy to within ~ < .1%) or use an ATTINY85 8 pin AVR using internal osc which is
+/- 1% accurate user cal.

Fooling around with mBlock, here is a sequence of pulses easily coded (gui Scratch
like language, mBlock, drag and drop and configure, code then auto generated based
on blocks used and variable settings).

Reliably create fairly accurate pulses down to 1 mS, into years as well, and sequences,
repeating sequences, pulses depending on a V reading and/or triggers like external trigger.
With not only programmable width but inter-programmable delay.

Here the code created 5 pulses of varying widths and inter-pulse delays. Could have made
some/all of the pulses in this example triggered by conditionals, like a V or a T or incoming
pulse width or.......

Here is mBlock IDE, code automatically generated, shown in right hand window. Running on Arduino
UNO. I think I can squeeze this into an ATTINY as well.

Very complex sequences trivial to create, accurate (Arduino has xtal on board).
Note embedded triggers can be put in, like a voltage in a range or exceeding a
value. Or that AND a counter overflow, whatever. mBlock is free, Arduino UNO
or NANO < $ 5, easy to get started.

If instead of mBlock you use Snap4Arduino you can display interrogated Arduino
stuff like V measured or PWM / counter values on PC w/o having to write Python
for example.

Using ATTINY85 solution would look like this. Note I did nopt debounce switch in code,
you can easily do that but your delays effectively obviate need for this as solution
used "blocking" code, so bounce is long dissipated in delay time to set PW.

Note I forgot but could have eliminated need for pullup R by configuring the pin properly with its
input R pullup, juts forgot.

Note resetting thru another button, or the one button pressed twice quickly, or power cycling all
trivial.

Lastly pulses can come out on individual pin or all on same pin or a mix....


Regards, Dana.

 

[d123]

Hi ErnieF,

Your circuit requirement sounded lots of fun so I messed around simulating with BJTs and RC stuff, rather than with ICs - no 555 and no flip-flop. Here is a theoretical circuit that possibly could be used for a non-retriggerable one-shot, and that is only re-triggerable after power-cycling and that shouldn't be unpredictable at power-up. More a fun look at this hard to fulfil goal than a serious suggestion. If only it were the 1950's... Unfortunately, there is a slight delay from trigger to output high...

The circuit is good for comic relief, if not a particularly practical implementation of the design goals. I think timing via a capacitor would be messy this way, and messy for other reasons with BJTs (e.g. temperature range and Vbe). Component values are selected arbitrarily, not by calculation. And, I hold my head in shame to have to admit that I bungled/muddled my way through the design until the simulation tool showed the results I understood are the desired ones.

In my opinion, this rough draft of a discrete component version looks like a possible load of questionable rubbish (not breadboarded, I wouldn't trust much anything), but anyway, it might give you a (much) better idea of how to lock the TLC555 reset pin after it has triggered once until a POR event. As ever with simulations of rough sketches of ideas: 'caveat emptor'.

 

[d123]

Hi,

... A simulation is hazardous territory... Above BJT version truly awful on breadboard due to Vbe. Re-did a basic version with allegedly logic level MOSFETs - simplified with high-going input (add inverter NMOS before Q2 maybe) and haven't worked on non-retriggerable section yet which presumably is just another N +P latch to block input signal after first trigger event.

This works reliably with no POR false triggering (so far) on a breadboard with 22uF to 3,300uF delay and/or output pulse capacitors. Hobbyist circuit, re-inventing the millions of wheels already in existance like this... 3,300uF is about 120 seconds of delay and/or on-time, 22uF less than 1 second.

Just wanted to show Implementation, not sim-world dodgy concept that sims great but won't work in reality. It's nothing great, pretty pathetic, but can be done and my version could be improved massively by a non-hobbyist. Biasing on and off levels is the biggest faff, and MOSFETs with high gate capacitance seem to be a nuisance, useless if you are a current Scrooge (like I am) - wastes 3.6mA when off, 0.95mA during on-delay time. A pretty bad circuit really, nothing to be proud of, but it works in the real world.

Excuse the 100-metre long hand-drawn schematic, I have no electricity today as the weather is bad again.

I will never use a 9V battery ever again if I can avoid it, they are sooo feeble where even a few mA are involved.

Anyway, please show your version if you did the 555 + flip-flop circuit you mentioned, I'd like to see that approach.

 

[schmitt trigger]

I love schematics drawn with “hand-CAD”.
Specifically if drawn in cuadricule paper, like Forrest Mims would do on its popular books.