Reduce input signal by 500ms

Thread Starter

XtinctElise

Joined Jan 21, 2012
10
I need a circuit to reduce the amount of time that a 12V relay is turned on by 0.5 seconds.

In this circuit if I have a 1 second input pulse, i need an output pulse of 0.5 seconds and if it is a 5 second input pulse, then i need a 4.5 second output pulse.

I would think this would be a rather simple circuit, but i have been away from circuit design a bit too long and need some help kicking my mind back in gear.

Can anyone supply a schematic or information on the design of such a circuit? Thanks in advance!
 

Brownout

Joined Jan 10, 2012
2,390
Sure. A simple timer set to .5s, and then logically inverted/anded with the original pulse. What happens then is nothing will happen for the first .5s, then the relay will activate until the pulse is gone. A 555 can be used for the timer. You'll have a bit of a race condition between the original pulse and the timer pulse, so you'll need a little delay, R*C, at the point where the signals hit the AND logic.
 

ErnieM

Joined Apr 24, 2011
8,377
As I'm kinda sure you don't want to do anything silly like violate causality then all you need is a 1/2 second delay. AND that with your turn on signal and drive the relay with that signal.

Basically it just subtracts the first 1/2 second from any pulse.
 

Thread Starter

XtinctElise

Joined Jan 21, 2012
10
Sure. A simple timer set to .5s, and then logically inverted/anded with the original pulse. What happens then is nothing will happen for the first .5s, then the relay will activate until the pulse is gone. A 555 can be used for the timer. You'll have a bit of a race condition between the original pulse and the timer pulse, so you'll need a little delay, R*C, at the point where the signals hit the AND logic.
Do you have a schematic for this layout? How much current will the 555 draw when it is not in use?
 

Brownout

Joined Jan 10, 2012
2,390
Do you have a schematic for this layout? How much current will the 555 draw when it is not in use?
Hi, no I don't have a schematic. But you'll want a one-shot, and you'll find many schematics for that with a simple search. Failing that, I can cobble someting up. But my time is limited, and you might wait for awhile. Some other members might have a diagram to post.
 

KMoffett

Joined Dec 19, 2007
2,918
Delaying the "pull-in" of the relay by 0.5 second assumes that there is no requirement that the start of the input pulse and the relay pull-in be coincidental.

XtinctElise,
Do you need the relay to pull-in at the start of the input pulse?

Ken

Ken
 

Adjuster

Joined Dec 26, 2010
2,148
Delaying the "pull-in" of the relay by 0.5 second assumes that there is no requirement that the start of the input pulse and the relay pull-in be coincidental.

XtinctElise,
Do you need the relay to pull-in at the start of the input pulse?

Ken

Ken
Note ErnieM's point about the violation of causality: it is a good one.

As I'm kinda sure you don't want to do anything silly like violate causality then all you need is a 1/2 second delay. AND that with your turn on signal and drive the relay with that signal.
Unless the pulses are repetitive AND we don't mind missing the first one, the 500ms has to come off the leading edge.
 

Brownout

Joined Jan 10, 2012
2,390
I got that, but does the OP understand "violation of causality"?

Ken
It doesn't matter if he understands it or not. The point is, there is really only one option.

If the two pulses need to fire at the same time, the original one can be delayed for .5S. If we can't delay it, we're screwed.
 

Brownout

Joined Jan 10, 2012
2,390
Do you have a schematic for this layout? How much current will the 555 draw when it is not in use?
Here is an adaptation of erineM's idea. Sorry I dont have the complete schematic, but this shows one way to delay the start of the pulse by .5s. Basically, the output of A1 ( a scmitt trigger device ) is connected to an AND gate, along with the original pulse. The output of the AND nees to feed a driver for the relay, typically a small transistor.
 

Attachments

Thread Starter

XtinctElise

Joined Jan 21, 2012
10
Delaying the "pull-in" of the relay by 0.5 second assumes that there is no requirement that the start of the input pulse and the relay pull-in be coincidental.

XtinctElise,
Do you need the relay to pull-in at the start of the input pulse?

Ken

Ken
Yes at the start of the input pulse. The pulses will be at least 500ms appart.
 

praondevou

Joined Jul 9, 2011
2,942
Yes at the start of the input pulse. The pulses will be at least 500ms appart.
As was mentioned, what you want to achieve is impossible, unless these are repetitive pulses.

If you trigger the relay when the input pulse starts you can not turn it off with exactly 500ms before the input pulses ends.

The end of the input pulse is unknown. You would want to foresee the future.
 
Top