Two monostable multivibrators. The first detects the leading edge on the input pulse and outputs a pulse of the desired delay length (0.1mS - 1.0mS). The trailing edge on the first monostable's output triggers the second monostable, and that outputs a pulse with the same length as your original pulse. This assumes that your input pulse has a fixed width, even though the frequency varies (1Hz-400Hz).
Again...is the positive or negative period of the signal a fixed width (for example: 1.0mS positive for a any frequency between 1Hz and 400Hz.)? Or. is it like a square wave where the periods of both the positive and negative portions vary with the frequency. This makes a big difference!
OH I see...sorry... well, its from a signal pulse(from ignition pulser) so width of positive(or negative) varies ... pulse width will be shorter in duration at higher frequency(400Hz) ... its mainly a "positive" pulse...
EDIT: I have found such a circuit already (googled) 4538
thanks a lot for the help!!
actually, this is just PART of what I want to experiment on... as frequency INCREASES, the pulse delay would INCREASE and vice versa... need a way to inversely affect pulse delay timing resistor....
How will the 4538's delay time change with the change in input frequency?
But it might ba a job for a microcontroller like a PIC. The signal rate is slow enough that it could even be handled by an interpreted BASIC chip like a PICAXE. Is it possible for the system to delay the output one full pulse cycle before it outputs a delay, so it knows how long the pulse is, and adds a lookup-table defined delay to the beginning and end next pulse and subsequent pulses?