Separate pulses from edges?

You could use a dual 555 timer, with one trigger driven by the signal directly, and one driven by the inverse.

It seems like that would recreate the original signal, or its inverse. Why would you do it?

-- john

Terry, Too late at night here to think through the details, but in general, a simple way to do what you want is as follows (example):

Feed the input signal through a resistor to the input of an inverter, with a cap to ground on the input pin of the inverter. The output of the inverter will then be a somewhat delayed, inverted form of the original input pulse. If you then take the original input signal and the output of the inverter, and connect them to the 2 inputs of an AND gate, then the output of the AND will be a positive pulse, with duration determined by the time constant (order of R*C) of the R and C you used.

The negative edge can also be obtained easily with a similar approach, just inverting the various signals as required. It can all be done with your

4001s.

Maybe someone where it isn't way after midnight might care to expand.

If you want to get a similar output regardless of the transition direction, delay the signal with an RC and combine it with the original with an exclusive or gate. The RC time constant determines the edge generated pulse width.

Of course, if you have no other use for the extra 3 gates in the exor package, you can use them to delay the signal, instead of the RC.

You could use a 4528 or 4538 dual monostable. One setup for pos edge trigger and the other for neg edge trigger. Set the pulse widths with

2 RC networks and select which outputs, Qs or not Qs, you want to use.

Mike

------------------------------------------------------------------------ The odds of a single modest length protein randomly forming is approx

1 in 10^260. By comparison the number atoms in the known universe has been estimated at 10^80 atoms. Figues are from the writings of Nobel Prize winner Francis Crick in 1981.

ms, not ns. (Would still trigger the 4013 though...)

Tim

--
Deep Fryer: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

Invert one and OR them? What kind of skew can you tolerate?

Thanks, Rich

I feel 555 can not be used in the monostable mode. It is agreed that the trigger by the falling edge will actuate the monostable but since the trigger pulse is in hours and we require output pulse width in msec, the trigger should have returned to its hight state before the

1.1x(RC) secs, which is the on time for the ouput pulse. If not the output pulse remains high. So for monostable operation the negative trigger pulse width should be less than 1.1x(RC) secs.

prateek

John O'Flaherty wrote:

On Sat, 24 Jun 2006 12:18:47 -0400, John Popelish wrote:

--- Unfortunately, since this is what he wants:

___________...___________ ________ IN____| |___________...___________|

_ _ OUTA__| |_________..._______________________...___________| |______

_ OUTB______________...___________| |_________...____________________

and what you suggested will give him this:

___________...___________ ________ IN____| |___________...___________|

_ _ _ OUTA__| |_________...___________| |_________...___________| |______

_ _ _ OUTB__| |_________...___________| |_________...___________| |______

your outputs won't set and reset his RS latch properly.

This seems to work:

Version 4 SHEET 1 880 680 WIRE -224 192 -224 -32 WIRE -224 256 -224 192 WIRE -224 320 -224 256 WIRE -224 464 -224 400 WIRE -224 512 -224 464 WIRE -176 -32 -224 -32 WIRE -176 256 -224 256 WIRE -80 256 -112 256 WIRE -80 320 -80 256 WIRE -80 464 -224 464 WIRE -80 464 -80 400 WIRE -64 -16 -112 -16 WIRE -64 -16 -64 -80 WIRE -48 192 -224 192 WIRE -48 256 -80 256 WIRE -16 -16 -64 -16 WIRE 80 -16 48 -16 WIRE 80 16 80 -16 WIRE 80 128 80 96 WIRE 112 -80 -64 -80 WIRE 112 -16 80 -16 WIRE 112 240 16 240 WIRE 224 -64 176 -64 WIRE 224 16 224 -64 WIRE 224 128 224 96 WIRE 224 240 176 240 WIRE 224 320 224 240 WIRE 224 464 -80 464 WIRE 224 464 224 400 WIRE 320 -64 224 -64 WIRE 320 240 224 240 FLAG -224 512 0 FLAG 80 128 0 FLAG 224 128 0 SYMBOL Digital\\\\and -16 160 R0 SYMATTR InstName A1 SYMATTR SpiceLine trise 1e-6s tfall 1e-6s vhigh 5v SYMBOL res -64 416 R180 WINDOW 0 -42 69 Left 0 WINDOW 3 -48 37 Left 0 SYMATTR InstName R1 SYMATTR Value 1E6 SYMBOL cap -112 240 R90 WINDOW 0 65 33 VBottom 0 WINDOW 3 67 34 VTop 0 SYMATTR InstName C1 SYMATTR Value .015E-6 SYMBOL voltage -224 304 R0 WINDOW 0 -49 10 Left 0 WINDOW 3 24 104 Invisible 0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value PULSE(0 5 0 1e-6 1e-6 5 10) SYMBOL Digital\\\\and 144 160 R0 SYMATTR InstName A2 SYMATTR SpiceLine trise 1e-6s tfall 1e-6s vhigh 5v SYMBOL res 208 304 R0 SYMATTR InstName R2 SYMATTR Value 10k SYMBOL Digital\\\\and -144 -96 R0 SYMATTR InstName A3 SYMATTR SpiceLine trise 1e-6s tfall 1e-6s vhigh 5v SYMBOL Digital\\\\and 144 -112 R0 SYMATTR InstName A4 SYMATTR SpiceLine trise 1e-6s tfall 1e-6s vhigh 5v SYMBOL cap 48 -32 R90 WINDOW 0 61 28 VBottom 0 WINDOW 3 65 30 VTop 0 SYMATTR InstName C2 SYMATTR Value .015E-6 SYMBOL res 96 112 R180 WINDOW 0 -43 70 Left 0 WINDOW 3 -47 43 Left 0 SYMATTR InstName R3 SYMATTR Value 1E6 SYMBOL res 208 0 R0 WINDOW 0 39 48 Left 0 SYMATTR InstName R4 SYMATTR Value 10k TEXT -328 448 Left 0 !.tran 30 TEXT 328 240 Left 0 ;OUTA TEXT 328 -64 Left 0 ;OUTB

-- John Fields Professional Circuit Designer

--- You misunderstood.

He has a signal which periodically changes state, and when it goes high he wants to SET what amounts to an RS flip-flop. Then when his signal goes low he wants to RESET the flip-flop.

All that's necessary to do to drive 555's with that is to invert the input signal, differentiate the falling edges of both the incident ind inverted signals, and use those low-going spikes to drive 555''s with 10ms timeouts. However, as you noted, the spikes will need to return to >= 2/3 Vcc before the 555's time out.

-- John Fields Professional Circuit Designer

I could not come up with an unambigious understanding of what he wants, so I prefaced my reply with "If". I am still not sure what he wants.

Thanks guys, much appreciate all the replies. Some neat ideas to follow up there.

But even while composing my OP it was nagging away at me that I'd asked essentially the same question before. Further digging this morning turned it up: news: snipped-for-privacy@4ax.com Subject: 2 pulses from 1 Date: Mon, 28 Feb 2005

Premature senility aside, I'm guessing that the reason it had slipped my mind was that, although I had two great solutions from Fred Bartoli and Fred Bloggs, I'd managed to come up with something trivially simple myself that seemed to do the job. In fact, then and now, I'd ignored the obvious: simply invert one of the inputs and use a vanilla RC differentiator on each.

This was the circuit I used then.

I'm pretty sure that was what I implemented, and the gadget has been working OK in the 16 months or so since.

The key difference *this* time is that I hope to feed these +ve pulses (or squared up versions of them) not to a 4001 bistable but to a 4013, as mentioned.

I haven't got around to that stage yet, so would welcome any feedback

- especially if my idea is a non-starter. I want the 4013 to achieve the following 3 things:

1. +ve pulse A to Set pin 8 always sets or keeps 4013 output high.
2. +ve pulse B to Reset pin 10 always resets or keeps 4013 output low.
3. A +ve pulse T (from another source) to Clock pin 3 always toggles the 4013 output.

For background, the practical application is a garden lights controller for my son. At present the lights have an on/off switch inconveniently located in his shed/garage at the bottom of the rear garden. My unit should

- Switch (or keep) the lights on at dusk

- Switch (or keep) the lights off at dawn

- Toggle the lights on or off at any time from a radio signal from the house, improvised from a wireless doorbell adapter.

--
Terry Pinnell
Hobbyist, West Sussex, UK

Um, if that's all that you want to do (turn the garden lights on at night, and turn them off at sunrise), wouldn't a simple photocell do the job? Something like this:

Good luck,

Michael

--
My take on it was that he wanted the rising edge of his input signal
to generate a spike which lasted for 10ms and was sent to the SET
input of his RS latch, and then, when his input signal went low, he
wanted that edge to generate a spike which would RESET his latch.

JF: Yep.

JP: I thought my carefully-worded opening sentence was clear enough: "I have a long, clean positive pulse (hours in duration). How can I get two *separate* 10 ms +ve pulses from the leading and trailing edges."

--
Terry Pinnell
Hobbyist, West Sussex, UK



please?

What about the third item of my spec? And its interaction with the other two?

--
Terry Pinnell
Hobbyist, West Sussex, UK

What about the third item of my spec? And its interaction with the other two?

--
Terry Pinnell
Hobbyist, West Sussex, UK

Hi Terry, I posted a simular question on 1-08-06 on this group. Under the subject >Generating start and end pulse< I put the circuit below together and it worked perfectly. Many thanks to Ken Smith. Sorry I don't recall component values, but an archive search should find some values. Mike

-----+---- ! ! !/ ! GND ---! ! !\\ e ! ! !/ ---!!--/\\/\\/--+--+---! ! !\\ e \\ ! / GND \\ ! GND