Power LED on .33hz square wave signal

[FractalAGF]

Hi all,

I'm a mechanical engineer and know a bit of electronic but not that much.

I need to power on a LED when it receives a .33hz pulsing square wave (normally open, 200ms closed every 3s). I understand how to do this with an arduino and stuff but is there an easier way to do this with basic components?

It is for a basic prototype and the signal is coming from a grease pump, it indicates a low-level of grease in the tank.

[cowana]

How critical is the timing - are you just looking for a 'yes it's pulsing' indication, or does it need to check for that exact timing?

If it's the former, a 555 monostable generating a 3 second pulse would be a neat solution - as long as the grease pump was generating pulses, it would keep the LED on.

[FractalAGF]

The timing is not critical at all, it's a case of either it is pulsing or it is not. Also, the LED can flicker or go on/off, it is not important. It is simply to give a hand to the people that will be operating the prototype in the coming year. We will either integrate the signal in our PLC (we don't have any input left at the moment) and put an alarm on our standard display or have a more complete controller developed for the greasing system.

If I understand correctly, the idea is to use a 555 to power the LED 3 second every time it receives a pulse? Can I set the circuit for longer to be sure I receive another pulse within the 3s? The pulse rate is driven by the paddle mixing the grease in the tank, hence is mechanically driven and I guess will vary slightly (+/- .5s?!?).

[CatalinaWOW]

Even simpler in concept, though possibly larger and more expensive in implementation is a diode followed by an resistor and capacitor.  The resistor should be about 1/3 or less of the current limiting resistor for your LED.  It may not be necessary, it limits the current sourced by whatever is feeding the .33Hz signal.  The capacitor is chosen to give a time constant of at least 1 second with the current limiting resistor.  The diode just keeps the possibly low shunt impedance of the short from dumping the charge out of the capacitor.

This concept will keep the LED on, though the brightness will vary.  Brightness variation can be reduced by making the capacitor larger.  Now days you can buy Farads in reasonably sized packages.   You don't mention the amplitude of your .33Hz signal, but if it is large enough you can totally eliminate brightness variation with a zener diode (and probably one more resistor).

If you go the 555 route you will have to do something similar to provide power, or provide a separate power supply (battery, wall wart or whatever), so it is likely that this approach will be appealing.

[Ice-Tea]

You haven't said much about the source... Amplitude? Source internal resistance?

[TheDane]

It's possible to get an integrated solution - the LED is flashing all on it's own.
http://www.ebay.com/bhp/blinking-led

Different colors, sizes, flash rates, voltage supplies, etc.
Could be an easy solution.

Blink = Power on led
No blink = Power off led

<edit< - Sorry, Just noticed it was the other way around.
What is the signal like when outside the [blink on]
If it's 0 volts, a capacitor and a high impedance FET/Transistor can drive the LED, and the right RC value can be calculated to turn on the LED permanently

[FractalAGF]

You haven't said much about the source... Amplitude? Source internal resistance?

Yeah, I was mistaken while reading the documentation of the greaser manufacturer.

It is not a signal per se, simply a switch being triggered 200ms every 3s

[Ice-Tea]

Page 15 would be interesting as well, but if you're OK with your light/LED following the same duty cycle (3s off, 200ms on) you could probably just wire 5V or so to the switch and have your series resistor and a LED on the other pin off the LED and you're good to go.

A simple FET with an RC at the gate would stay on during the off time as well.

[Electro707]

I am not sure how you build this, but you could try a transistor oscillator(it would be cheaper)

[Brumby]

I'd try starting off with something out of the TI application notes for the 555 timer...

www.ti.com/lit/ds/symlink/ne555.pdf

Section 9.2.1  -  Missing Pulse Detector

[FractalAGF]

Page 15 would be interesting as well, but if you're OK with your light/LED following the same duty cycle (3s off, 200ms on) you could probably just wire 5V or so to the switch and have your series resistor and a LED on the other pin off the LED and you're good to go.

A simple FET with an RC at the gate would stay on during the off time as well.

Here is page 15. I am planning to use pin 1 and 2 to bring +12v and ground to a remote box containing a keyed switch and that circuit xyz driving the LED.

I was thinking about about letting the LED flash however it is fairly slow and the operator can miss it. Because we will be using the greaser for a short amount of time, I'd rather have it not flash at all.

You guys gave me some great idea and I need to explore them at the moment. Is there anyway way to meet the other requirements (3 signal within a minute before opening the LED) without a controller/PLC or whatever? I'm starting to think it is fairly complex...

[grifftech]

use a capicitor

[Ice-Tea]

You could use an RC circuit on the gate of a  FET so that 2 x 200ms of charging does not suffise to charge it beyond the gate threshold voltage.

Sure there are other prettier ways to do it but this seems about the simplest