AW: Re: Pulse width circuit??????

[gstopp at fibermux.com]

     You got the idea... however no matter what chip I try in the future 
     I'll be looking at the scope as the second-to-the-last test. No matter 
     what the numbers are on the spec sheet, I want to look at the square 
     wave on the scope and see a pair of dashed lines going from left to 
     right, complimentary to each other, perfectly flat, with nothing in 
     between them, at all audio frequencies, into a 1K load to ground. No 
     ringing, overshoot, slews, slopes, asymmetries, or artifacts. If I see 
     that I'm happy.
     
     Ah heck another lab thing to do.
     
     - Gene


______________________________ Reply Separator _________________________________
Subject: Re: AW: Re: Pulse width circuit??????
Author:  ftom at netcom.com (Tom May) at ccrelayout
Date:    12/6/95 1:47 PM


My turn to apologize for extreme techno-babble.  Maybe this thread 
should park itself on synth-diy?
     
Doesn't ECL have the same problem with different rise and fall times? 
It's either that, or some kind of under or overshoot.  The ECL 
waveforms I remember were functional but not very pretty.  (I realize 
you didn't actually say "ECL", but "similar to ECL".)
     
Anyway, if my Fourier-series calculations are correct, a 20Hz +/- 15V 
square wave slewing at 8V/us has harmonics out to the 6,666th at 
133+kHz, give or take a few thousand harmonics, which should be enough 
for anybody.  Assuming the Fourier series is simply truncated after the 
6,666th harmonic, which is not true; if it were, you'd see the 9% 
overshoot associated with the Gibbs phenomenon on a scope.
     
Sounds like if anything is going to be a problem, it's the corners.
     
fTom.