🚀 go-pugleaf

In article <2820@td2cad.intel.com> mielke@td2cad.intel.com (Neal Mielke) writes:
>
>I'm looking for a curcuit to allow a PC to measure the edge-to-edge
>delay between two TTL input signals.  This delay will be in the range
>of 50 to 100 nsec, and I'd like an accuracy of 5 nsec.  Any ideas?
>Are there any generally-available IC's that could help?
>
	This is a problem dealt with often in nuclear instrumentation;
the LeCroy catalog gives a discussion of the common technique.  One
runs one-shots off of the input signals (up to the maximum time which
you wish to resolve, in this case 100 nsec), inverts one signal, and
ANDs the result.  This results in a square pulse whose width is
the required delay time.  This square pulse is integrated and (at the
turn-off time) represents the time delay as the voltage on a capacitor.
So, your time delay now is an analog (voltage) value; digitize it
or discharge the capacitor with a calibrated current and time the
discharge to get the delay.  
	Other techniques get better accuracy for shorter delays; LeCroy
offers modules that go down to 25 psec accuracy at 100 nsec range.
	
	LeCroy Corporation
	700 Chestnut Ridge Road
	Chestnut Ridge, New York 10977-6499
	(914) 425-2000

a discussion of the procedure starts on page 21 of their '89 
"Nuclear Products Catalog".  Their modules are not generally PC plugins,
but use other standards (NIM, CAMAC, FASTBUS).
	Good luck.

I am known for my brilliance,                   John Whitmore
 by those who do not know me well