On Friday 22 October 2004 01:25 pm, Eisermann, Phil [Ridg/CO] wrote:
> piclist-bounces@mit.edu wrote:
> > On Friday 22 October 2004 05:25 am, Joe McCauley wrote:
> >> I need a delayed pulse output (delayed with respect to a trigger
> >> pulse). The delay should be variable between 5ns to 1000ns. This
> >> delay needs to be stable between pulses. The delay needs to be
> >> easily adjustable. The trigger pulse width will be in the region of
> >> 50ns wide. The output should be triggered from the leading edge of
> >> the trigger pulse. The output pulse width should be in the region of
> >> 5ns wide with with amplitudes of up to 10V. Does anyone have any
> >> clue how this might be achieved?
> >
> > For stability:
> > Basically, you want some form of 555 followed by another 555 timer
> > type circuit.  The 1st 555 has a variable R while the 2nd 555 has a
> > steady 5nS. If you want to go el-cheapo, you can build the circuit
> > using transistor logic.
>
> good luck getting a 555 operating at those speeds. The minimum pulse
> width in the hundreds of nanoseconds, the output rise/fall times at
> least 100ns, and the minimum duration of the monostable output is
> around 10usec

If you want everything served on a silver platter, than pay-up.  :-)
If you are willing to do some homework of your own, the idea should fly, 
after-all I did say "some form of 555".   ;-)
The original poster asked for a suggestion, and this seems like a fine 
suggestion... Only thing left is to work out the details.... which if 
willing, is his homework.  :-)

> a 5nsec delay implies the circuit must operate significantly faster
> than 200MHz.

Correct, and within the range of a transistor pulse circuit if you are 
willing to calculate the required R, C and choose the correct transistor.

> Unfortunately these speeds are out of the range of anything I have
> ever worked on, so am not sure where to start. If only there
> was an 800MHz PIC...

The PIC is acting as a variable R between +5v and the connection which you 
would normally use for your 555 RC circuit.

Now, for better idea of the variable R:
Suppose pin PORTB0 has R=100k and PORTB1 has R=50k
If PORTB0 = input & PORTB1 = 5v, then you hav an R between +5v and your RC 
connection of 50k because PORTB0 is effectively OPEN CIRCUIT.
If PORTB1 = input & PORTB0 = 5v, then you hav an R between +5v and your RC 
connection of 100k because PORTB1 is effectively OPEN CIRCUIT.
If PORTB0 & PORTB1 = 5v, then you hav an R between +5v and your RC connection 
of 33k3.

>From the above, you could see a variable R of 100k, 50k and 33k3  ;-)
For all intensive purposes, the PIC could be running on a 1hz clock and still 
do the job because the "speed" intensive stuff is done by the 555 
"equivalent" circuits....chips or transistor logic... your choice :-)

_______________________________________________
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist