Generating start and end pulse

You can make two differentiators, one goes to the transistor in question, the other goes to a PNP transistor that gets turned on by the negative-going pulse. Add a resistor from collector to the NPN's base and you've got your frequency 'doubler'.

Or you could use diodes to rectify the pulses, so you get a train of "up", rather than alternating "up" and "down" pulses.

Tim

-- Deep Fryer: a very philosophical monk. Website:

Do you have a schematic for that? Mike

negative-going

your

"up",

Hmm, on second thought, either the end circuit or the 555 timer will have to be floating, since it needs to develop that peak voltage across the rectifier to steer the up/down pulses correctly. Capacitance between the two circuits must be much less than the differentiator capacitor, say 5 times smaller, or better.

There's a circuit I made that does it with one transistor, it's not particularly elegant and probably won't work very well outside of the frequency range I was testing it in, also, it doesn't work well for frequencies other than what it's tuned to. I would suggest viewing the output pulse time somehow and adjusting values until it works right.

Ken's solution is interesting, low on parts, and a lot less ugly.

Tim

-- Deep Fryer: a very philosophical monk. Website:

I put your two transistor circuit together (as seen below), it works perfectly. The output pulses are not very pulse like, but it switches the (locate handset) button on the base station of a cordless phone and causes the remote to ring. Now from breadboard to real world. :-? Thanks for your help. Mike

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

--

-- snipped-for-privacy@rahul.net forging knowledge

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

The two pulses will be unequal, but since you didn't spec and matching of the times, I'll guess this is OK.

XOR gate? With a RC network delay the 4 sec pulse. Then gate the original and delayed one ,and voila! two pulses.

In article , Tim Williams wrote: [...]

Who's circuit are you calling "less ugly"! :>

How about this:

! / R1 \\ / \\ ! +---!!---GND ! ^V Triac ----- / ! ----!!---+---- ! !/ ! ---/\\/\\--! / R3 !\\ \\ ! / R2 GND \\ ! GND

So long as the current in R1 is less than the holding of the triac and the R3 current is more, the circuit should work.

Make a 3-second pulse, and trigger the second 1-second with its falling edge, and OR them together for the 4-second overall.

Good Luck! Rich