super simple pulse stretcher

Great, thanks!

Cool!

Yep. Seat of the pants engineering at its finest >:-} ...Jim Thompson

In this case is that comment meant as a compliment or pejorative? Or some of both?

Both. Some of concept was on the right track, implementation doesn't work _at_all_!

Can you define "stretch duration" requirements? How fast might the input pulses be coming in?

Any reason you can't use a 555 timer ?:-) ...Jim Thompson

They're sync pulses from one of those old musical instrument sequencers of the type from like, the late 1970s, before MIDI was invented. The only way to interface gear was with a gate pulse and a control voltage.

I don't know, I guess a realistic upper bound for how fast the input pulses are coming in would be like, 32nd notes at 200-something beats per minute? Whatever that works out to...100 ms I guess? That might be a good stretch length

The 555 timer is really cheap in quantity but the bipolar version draws a lot of quiescent current and I'm running on batteries. The CMOS is more expensive. And I'll probably have free opamp sections.

Can you "see" 50ms on and 50ms off?

You say you have free OpAmp "sections" (plural?). How many sections might be used? 2-3 sections would make it a fairly easy task to get all your desired features. ...Jim Thompson

Why does it need pulse stretching? - wouldn't it be sufficient to drive enough current through the LED in 5mS to make it bright enough to see? Some phone camera flashguns are LED based these days.

Ah, yeah I slipped a decimal point there. At 10 ms that will be steady on for sure, but that's OK.

An ordinary pulse stretcher will produce flashes of width "T" until the incoming pulse spacing is less than "T", then steady.

Seems to me like the trivial solution would be a toggle-flop, driving the LED with Q.

Jazzier, combine toggle flop with stretcher so that pulse width is restricted to TAU, until Q "snuffs" off the width at high pulse rates. ...Jim Thompson

See BitrexStretcher.pdf on the S.E.D/Schematics Page of my website. ...Jim Thompson

If you insist on using discrete's, see ToggleFlopAncient.pdf on the same page >:-}

...Jim Thompson

This is pretty simple:

Optional R2 limits the peak cap current, and sets charging time. R1 sets LED current and decay time. Stretch ratios like 50:1 might be feasible. The power supplies need to be able to deliver the peak charging current.

LED brightness will sort of follow the input pattern.

The OA prototype monostable, let the OP figure out how to adapt it: Please view in a fixed-width font such as Courier.

. . . . . . -- . | Cc |\ --------- . | >-----||---+----------|-\ | . -- | | > ---+--> | . [R1] ---|+/ | --- . | | |/ | . | | | . --- +----||----- . com | C . [R2] T . | . | . | . --- . V- . . .

I've done two RC-driven LEDs, from both Q and /Q of a toggle flop, just so one would register close-spaced pulses. All we had in the drawer were 74HC74, but it's a lot easier wiring to a '74HC73 JK to do this task. Less fiddling with which inputs to strap high, and which low... they're all high.

For a while, in the 1980s, 'logic probe' high/low/pulse detector circuits were a rage. There was lots of cleverness expended on getting blinkie lights to divulge useful circuit info. That's about the time everyone started noticing the '555

Oops, for some reason I just noticed this now. Thanks!