Schmitt Trigger Jitter

I have an ancient HP 185 sampling oscilloscope, 3 GHz bandwidth, something like 10 or so ps of RMS jitter. It's mostly tubes.

For really low jitter programmable-delay generators, I use RC ramps and expensive ECL comparators. But for less demanding stuff, an LVDS line receiver, like a DS90LV012A, makes a damned fine cheap almost-rri comparator, ballpark 10 ps RMS jitter.

Can you post your schematic? Any chance the current source is oscillating? Mine tend to do that. So I prefer an RC, with a fairly high voltage across the R, and do some polynomial correction for the modest curvature.

Fine, if you won't take my word for it...

I lied in that there's more to it, obviously, but as DTB6 is filtered, and the other edges are uncorrelated to the edge in question, it's not causative.

Tim

Like Wescott said, this is too terse to be properly understood by someone who hasn't seen the schematic or the graph.

I'll give my two cents: capacitors are not charged linearly. Instead, they are charged hyperbolically. The initial phase of the charge is steep, the later phase much leveler. This later phase is where cosmic radiation, "quantum fluctuations" and other niceties take effect and generate noise. The noise widens the voltage trace on your plot. So, the capacitor voltage does not cross the Schmitt trigger level in a point, but instead it has a chance of crossing it over a period of time. The more level your voltage rise, the longer the time period over which the Schmitt trigger will get - randomly - triggered.

This is the most probable cause for your jitter differences. Calculate to test the hypothesis. And post the results here. :)