delayed retriggering

[justinlw]

i'm looking to setup a timer circuit, that is triggered on the downswing of a pendulum. the pendulum induces voltage in a coil to trigger the timer on. the pendulum, however, passes over the coil twice per swing, once on the down swing, and once on the up. i only want it to trigger on the down swing.

would using a lm555 in a delayed retrigger setup, a al **broken link removed** , be a feasable set up for this?

(after the downswing trigger, i'd have the delay set to a time less than the time it takes for the bob to reach the top of its swing, and return for most of the down swing. the delay ends sometimeframe before the instant it retriggers.

 

[philba]

that certainly couild be made to work. You'll have to figure out how to trigger it as the current circuit uses s1 to pull the circuit low. maybe use an NPN transistor with the coil feeding the base (with a limiting resistor and diode in series). It's really hard to say with out more info about your coil.

If the coil is a sensor or the pendulum, there are simpler approaches. a hall-effect device would be my first choice though a photo transistor and led together could work well, too.

I'd probably use a flipflop or counter as a toggle to skip every other pulse. that way you aren't dependent on the pendulum frequency and ampllitude.

how are you determining which is the downswing?

 

[justinlw]

yeah, planned on using the coil to feed an npn.

the downswing will be the first swing, ie: the moment we start the pendulum in motion. we'd leave the timer's off untill the pendulum reachs the top of its swing. then turn the timer on. so the first swing the coil ever see's is a downswing.

since the freq/period of a pendulum doesnt change, i know exactly how long the positive and negative duty cycles must be. just need to keep the delay long enough, so the bob passes the oppsite side of the coil before the delay runs out. should actually keep the delay long enough so the bob is almost to the trigger spot again. that'll keep any outside sources from induceing thru the coil prematurly

 

[philba]

yeah, planned on using the coil to feed an npn.

the downswing will be the first swing, ie: the moment we start the pendulum in motion. we'd leave the timer's off untill the pendulum reachs the top of its swing. then turn the timer on. so the first swing the coil ever see's is a downswing.

since the freq/period of a pendulum doesnt change, i know exactly how long the positive and negative duty cycles must be. just need to keep the delay long enough, so the bob passes the oppsite side of the coil before the delay runs out. should actually keep the delay long enough so the bob is almost to the trigger spot again. that'll keep any outside sources from induceing thru the coil prematurly

Why not not just use a counter chip like the 74xx393 (though any counter would do). use a schmitt trigger to clean up the wave form and feed that into the counter. bit 1 (Qa on the 393) would be your 1st, 3rd, 5th, ... triggering. Then you don't need to worry about timing.

Its your call, of course. Your dual 555 scheme should work ok assuming you get the timing right. Your lockout time should be just a bit less than the period of the pendulum. still , that locks you into a specific pendulum frequency.

Note, that with a microcontroller, you can automatically determine which is the downswing. assuming the sensor location is asymetric, of course.

What are you doing this for?

 

[Nigel Goodwin]

You might check EPE magazine April 2006, that has an omni-directional pendulum project - it uses the two mono-stables in a 4098 CMOS chip.

 

[justinlw]

its for a foucault pendulum thats going in the lobby where i work.

but thanks for the ideas. i'll look into them.