Reducing Pulse Duration

Of course you can always shorten the pulse first, using a differentiating network (series-capacitor, resistor to ground), and then use your favorite timer.

Edge-triggered monostable multivibrators exist in the wellknown logic families (74XX series, CD4000 series), but it depends on your other requirements of these can be easily dropped in instead of a 555.

In the olden days I would have used a 74121. But it never made it into the newer families.

--
Reinhardt

The mickey-mouse method for shortening a positive-going pulse is to use two sections of a CD4093B Schmitt-trigger NAND gate plus an RC, like this:

in |----\

0--*------| - \ |----\ | | / }O----RRRR--*------------| - \ *------| - / | | / }O---- | |----/ CCC *------| - / | CCC | |----/ | | | | GND | *-------------------------------*

Of course this is liable to misbehave if your input isn't what you expect.

You can also do it with a CD4013B flip-flop: clock 1 into it on the rising edge of your pulse, and use the RC delay on the reset pin. That's liable to be less well behaved than the Schmitt method, though.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

Oh yes it did! The 'HC123A and 74HC4538 are both descendants, and very nice to use.

Cheers, James Arthur

Indeed, those are usable types. The CD4538 was of course the CD4000 series predecessor of the 74HC4538.

However, all of those have much less output drive capability than a 555. When you want to drive something, it may be necessary to add something like a BS170.

Or my favourite: the 2N7000.

That is about the same thing.

Which is why I suggested it. Just in case the OP has no BS170s hanging about. :-)

Anyway, these small MOSFETS are really convenient as drivers. Source pin interfaces directly to the logic output (no resistor required), switches very fast, almost no saturation voltage hence little loss and dissipation.

May not be the simplest way, but I would consider using an ARDUINO. Use one pin to sense the incoming pulse and another pin for the output pulse. You can make the output pulse pretty much what you want with one millisecond resolution.

This kind of assumes you have some other things you want and can use the arduino to accomplish those additional tasks.

Dan

@Rob:

1. How much output drive do you need from the 250ms source?

1. Do you need to trigger the device with a positive-going or with a negative going edge?

Rob??

It needs to be able to drive an IRF840 MOSFET preferably directly and this is an enhancement mode device:

Chris, sorry about that.

This should work for you:

dl=0

It does - only too well! According to the simulations I've run, the shrinkage is *too* effective and the output pulses too skinny. Which components on your schematic determine the output pulse duration, please? I shall need to tweak them. Many thanks.

--
The timing components are R4 and C2, and the formula for the output 
pulse's width is t = 1.1 R4 C2. 

John Fields

Excellent! Many thanks.

Try a one-transistor one-shot.

Actually, on the simulations I've run with different values (over several decades) for these components, I can't detect any change in the input/ output pulse duration ratio. But the 2n2222 input device is useful in itself; didn't think of trimming the pulse that way.

--
On the trigger side it's actually C1 which is doing the work by 
differentiating the negative-going pulse out of Q1 into a spike with 
way less ON time than the 555's output pulse. 

John Fields