time delay circuit

[Scorp]

Hi Guys,

Hope someone can help me with a circuit I need to build. Here is my situation: I am building an interface between a clay target shooting range sequencer and the actual clay target traps. The sequencer has fifteen 12v outputs that I will be using to drive relays. The outputs will all stay at 0v until the sequencer wants to fire off a target, at which time the output will go to 12v for a short duration.

I will be turning on one of fifteen LED on the interface box whenever it's associated ouput goes high (which is easy enough to do..). however the duration of the 'pulse' is quite short and I'd like the LED to illuminate for a longer period of time. What I need to come up with is a circuit that takes the 12v 'pulse' and switches on an LED for approx 1 second. I've thought of something like an RC circuit powering the base of a transistor providing current to the LED, however off the top of my head I think it will just fade the LED as the capacitor discharges, rather than turning off cleanly.

Can someone offer some suggestions on the most efficient way to do this? I'd like to avoid using a micro to drive the LEDs, but will do so if there is not a more elegant solution.

Cheers!
Ben

[Simon]

Well a micro-controller will probably give you more outputs from 1 IC (you could use and addressed input system so that 15 inputs are represented by 15 3 bit codes leaving more I/O pins free for outputs to leds). you may be able to do it with a comparator although this is a classic for a 555 timer IC check out the 555 datasheet

[arcom]

You can also do it with a schmitt trigger IC, a diode, cap and a resistor. Less parts compared to 555 timer IC solution but less flexible.
Check out "Pulse Strechers" at the end of the page: http://www.musicfromouterspace.com/analogsynth/mmlogic.html

You can adjust pulse width by changing the resistor and/or capacitor value.

Second inverter is needed only to correct the polarity of the output pulse but if you need this only for an LED then you can omit the second inverter and connect the LED between the positive rail and output of the pulse strecher. 74HC14 is suitable for the job.

[rival_rage1]

link is broken now

[Rerouter]

you could always use a diode fed rc network, but then feed that into cheap shmitt trigger chips, being 6 of them to a standard chip you only need 3 total,

[codeboy2k]

Here's a simple, no programming suggestion. 

For each LED, use a single open-drain version of the 74HC595 serial-in 8-bit-parallel out shift register.  One open drain version is from NXP NPIC6C595-Q100.

The pulse will go into the SI (shift input). 

All 15 shift registers are continuously clocked by a 8 Hz RC clock.. nothing accurate needed here.

The 8 open-drain outputs of each shift register are wire-or tied together, then to the led, then to a current limiting resister then to your +12V line.

The pulse will shift into the register, and appear at each output, for 8 pulses, sinking current through the LED for 1 second.

For fewer parts and a bit of programming effort,  you use an MCU as others have said. But you need an MCU with 30 I/O's or use a few SPI port extenders, or use 2 eight bit shift registers to drive up to 16 LEDs (with timing in software), etc.

[rival_rage1]

so what would be a good method for a 10 second schmitt trigger?

[Rerouter]

using a 4584 hex schmitt trigger, have a schottky diode come off your switched output (assuming 5V), this then connects to (approximately) a 1 Meg and 1uF cap in parallel, to ground, and this then connects to your schmitt trigger input,

so following the turn on pulse, the cap charges up to about 4.6V through the diode, which turns on the schmitt trigger and then slowly begins to discharge,

somewhere between 7-13 seconds later (tolerances on the cap) the voltage will fall beneath 2.5 and will turn it off again,

[codeboy2k]

I was busy drawing a schematic (version 2 below) when Rerouter replied already

So I updated my schematic to include his version (version 1).

Version 1 has the advantage that it uses only 1 gate, but needs the initial pulse to be wide enough and supply enough current to charge up the capacitor. If the pulse is too short, or the source is high impedance, it cannot charge up quickly enough. In that case use a smaller capacitor and a larger resistor.  At some point you reach a limit, and version 2 should be considered.

Version 2 uses the initial short pulse to discharge the capacitor, which then charges up again through the power supply and the 15Meg resistor.

The actual values for R and C depend on the schmitt trigger's threshold and Vcc voltages used.

NOTE THAT the 74HC14 can only sink and source 4-5mA, not really enough for an LED.  The SN74LVC14 from Texas Instruments can sink and source up to 25-30mA, so the output can drive a LED directly and can be used to answer this thread's initial question. Also the MC14584 can sink and source up to 10mA, good enough for some modern LEDs.