Trouble with 555 to make negative recovery delay, need advice.

[manulonch]

Hi,

I have a little problem with a 555 circuit,

here is the schematic



in fact the problem is simple,

when I do a short press on the switch, the relay remains activated during desired delay,

by cons if I press longer on the switch, longer than the time delay, the relay shut off as soon as I release this switch.

In fact, the delay begins when the switch is pressed, on the contrary with this table on right of the schematic.


Does someone have an idea ??

Thank you in advance.


Power supply is 12volts,
already try several 555

Schematic come from here : **broken link removed**


( sorry for my English, I'm French)

 

[betwixt]

I suspect the problem is the switch contacts 'bouncing'. I think what you are seeing is the time interval starting as soon as the switch is opened not when it is closed. If the switch is held closed, the timer will not run (as you observed) but when a normal switch is pushed the contacts open and close a few times as they collide and bounce off each other. The bounce will be so fast that you do not observe the effect yourself but the electronics of the timer notice them and will start a timing cycle.

Brian.

 

[esp1]

hi,
I would agree with Brian it could be switch bounce, it works as expected in simulation.
E

 

[manulonch]

Hi, Thank you for help !!

I make some test, apply directly GND on the Trigger (2) with a test probe, sometime the problem don't appear and it work normaly !!

Then, I try to debounce like this, with some different cap value, but the problemm still.


I forget to say, the control switch is a contact relay, not a simple switch,
the output go to a resistor then a transistor.

 

[betwixt]

Adding a capacitor in that position will change the timing anyway. In that circuit the point where the switch is connected has to be 'floating' (not connected to anything else) for it to work properly.

You might be able to use a transistor instead of the switch and filter it's bias current to clean up the switch edge. Try using an NPN small signal transistor with C connected to where the switch top is and E connected to ground. Wire your switch between +V and the base of the transistor with two resistors of about 47K in series. You can then experiment with capacitors between the junction of the two 47K resistors and -V to find a value that filters the switch bounce sufficiently. I would guess a value of around 100nF would be a good starting place. There is a drawback to this method which is it introduces a small delay between the switch being operated and the timer starting but it should be small enough that you wouldn't notice it.

Brian.

 

[manulonch]

Thank you Betwixt, seems to wok fine now !!!

Then from this answer become a new question ( sorry )

It work with the transistor (bc547), but I try with an optocoupler (H11A1) and the trouble comeback,
I suppose I just have to make another deboucing circuit on the input of this optocoupler...
Question -> what the best way to debounce this input ??

Thank a lot for help !!!

- - - Updated - - -

Little edit, I just have to raise the cap to 470nf to have a good work with several switchs.
An other stange stuff, the delay is shorter after a long "on state" switch vs a short impulsion.