How to get one-shot pulse when circuit receives power and another pulse when power is lost?

Thread Starter

oh2th

Joined Jan 18, 2020
5
Ideally using a 556 chip I'm trying to setup a circuit that:

a) when circuit is powered on it triggers a one-shot pulse on the 556's A side.
b) when the same circuit looses power the 556's B side is triggered.

Circuit voltage should be 5 Vdc (optionally 12 Vdc) and pulse length should be adjustable (1-15 seconds).

Regards,
Tapio
 

ericgibbs

Joined Jan 29, 2010
10,469
hi oh2th,
Welcome to AAC.
Do you have a sketch of the project you could post, so that we can see what you have so far.?
E
 

Thread Starter

oh2th

Joined Jan 18, 2020
5
Hi,

The power-on sketch is simple one-shot, what I can't get over my head is how to do the part when power is lost. And yes power loss also on the 555/556-chip. I.e. need to power the chip also enough so that the pulse gets triggered from the B-timer.
 

Thread Starter

oh2th

Joined Jan 18, 2020
5
First circuit for A-timer (one-shot pulse when powered on) works as intended.
Trying to figure out how to send pulse with the second B-timer when power has been lost.
 

Sensacell

Joined Jun 19, 2012
2,587
Then what gets pulsed when the power goes off?

"Send a pulse" to what?

Two different signals? or the same signal?

Context is needed here to really assist you with this.
 

peterdeco

Joined Oct 8, 2019
202
You really have to be more specific. Is power consumption an issue? An energized SPDT relay will release when power is lost and can use a charged capacitor to send a pulse.
 

Thread Starter

oh2th

Joined Jan 18, 2020
5
The original need is to be able to pulse signal a device that has "UP" - "common" - "DOWN" connections. Idea was to have remote control setup so that when remote device is been power on, the "UP" and "common" is connected with a pulse. And when the remote device is switched off, the "common" and "DOWN" is connected with a short pulse. I can get 5Vdc USB power from the remote device when it is turned on.

Even a simple relay setup could work, power consumption within USB 5Vdc (100mA).
 

Sensacell

Joined Jun 19, 2012
2,587
How fast does the power go up? and come down? (volts / Usec for example)
How long does the output pulse have to be?

How long does the power stay on and off? - the minimum - when its in the static state?
 

Sensacell

Joined Jun 19, 2012
2,587
Here is a really dumb simple way - assuming that the power goes on and off quickly and stays stable for a few seconds.
It also assumes that the "common" is at ground potential, the inputs needing to be pulled to ground to activate.power_pulser.jpg
 
Last edited:

AnalogKid

Joined Aug 1, 2013
8,536
The standard way is to have a diode and large capacitor form a power source for the OFF circuit, and the OFF circuit is triggered by the power source (in front of the diode) going away and appearing as a negative-going edge to the Trigger input. Sensa has taken a quasi-similar approach that capitalizes on the fact that your timing requirements are so loose that 555's are not required (unless you say so). What would be the power storage cap now is the timing component for both the ON and OFF pulses, and the optocoupler LEDs are their own isolation diodes.

Nice.

At 20 mA peak LED current (2.9 V / 150 ohms), the pulses will be shorter than required. For a 5000 uF cap, and 150 ohm resistors, I get approx. 0.75 s.

What is the purpose of the 1 K resistor? (the only one that doesn't need a reference designator)

Note: Even with the current limiting resistors, I would change the diode to a 1N400x for extra beefiness.

ak
 
Last edited:

Sensacell

Joined Jun 19, 2012
2,587
What is the purpose of the 1 K resistor? (the only one that doesn't need a reference designator)
That 1K resistor allows the cap to charge up to almost 4.4 V (5 minus a diode drop) to keep the pulses symmetric.
Otherwise, you lose an extra ~1.3 V due to the LED's forward voltage.
 

Thread Starter

oh2th

Joined Jan 18, 2020
5
Here is a really dumb simple way - assuming that the power goes on and off quickly and stays stable for a few seconds.
It also assumes that the "common" is at ground potential, the inputs needing to be pulled to ground to activate.
This looks pretty much to what I was looking for, need to test it.
Thank you for your kind help.
 

danadak

Joined Mar 10, 2018
4,057
Do you need a "glitchless" design, one that does not generate any
false pulses when generation circuits power turns on or off ?


Rewgards, Dana.
 

Janis59

Joined Aug 21, 2017
1,200
Simplest is exclusive NOR logics cell where one input is Vcc and other is the same Vcc via diode and larger capacitor.
At startup R1C=tau(0-1) wil give a time when Vcc already is on but V2 is still in the charge-up process, thus the output is 1. When boths inputs are charged up - it is 0. When switch off process is on, then Vcc fall fast, but V2=CR(2)=tau(1-0) slow, thus again in the output may be seen the pulse with a length of tau(1-0). Logically, the cell`s Vcc must be fed from rather large capacitor buffer.
 

danadak

Joined Mar 10, 2018
4,057
Be careful about hanging large caps off CMOS inputs due to power
cycling issues.

Data sheet has input current limits. If a power supply collapses to ground
with a low Z then a large cap will discharge thru input protection diode
possibly causing hot spot die melting, blown bond wire, poly R changes....

So between cap and CMOS input place a R to limit discharge current.


Regards, Dana.
 
Last edited:
Top