DESIGNING EXTERNAL RETRACE-BLANKING

Dear Analog Gurus,

I need some help/sanity-checks/ideas/more pairs of eyes:

I am trying to add a retrace-blanking circuit to my existing
curve tracer design, which uses a separate oscilloscope for
its display. The output of the blanking circuit will connect to
the scope's External Z-Axis Input, which needs >= 5v (into
10K, I think) to blank the trace.

What I have come up with is ALMOST seeming workable.
But I'm running into trouble, trying to make it work robustly
all the way from 60 Hz to 22 kHz.

I've uploaded LT-Spice files for what I have, so far, here:

http://www.fullnet.com/u/tomg/blanking.asc
http://www.fullnet.com/u/tomg/blanking.plt
http://www.fullnet.com/u/tomg/op275.sub
http://www.fullnet.com/u/tomg/op275.asy

[In LT-SPice, open and run the "blanking.asc" file. The other three
are then used automatically, as long as they're in the same folder.]
[Download LT-Spice (AKA SwitcherCad III), free, from
http://www.linear.com.]

The circuit is essentially just a buffered inverting differentiator
followed by a comparator, in order to generate a positive pulse
only during the fall-off at the end of the ramp, and zero elsewhere.

Included are two "artificial" voltage sources that closely approximate
the existing available sawtooth ramp sweep signal, at both 60 Hz
and 22 kHz. (This is actually just the available "raw" sweep signal,
from before it is set to the desired amplitude, offset-zeroed, and
power-amplified, which is used because it will always have
approximately the same amplitudes and offsets.) Just use the
"Move" mode to move the jumper ("X2"), to select one of the
sweep voltage sources.

The main problem it has is that when the differentiator is
made "fast-enough" for good timing at 22 kHz, its output at 60 Hz
has a way-too-low amplitude. It can be made to work, in
simulations, at both 60 Hz and 22 kHz. But when the 22 kHz
response is (almost) "good-enough", the 60 Hz response is
extremely dicey, relying on offsets and amplitudes being
within ranges on the order of a few hundred uV, which
seems unworkable outside the realm of simulation.

I would hate to have to try to squeeze one more 8-DIP onto
the existing board, beyond the two that this circuit will
already require. But, if I HAVE to, I guess I MIGHT be able to find a
way...

Any suggestions would be MOST welcome. I did also try using
a 555 as a one-shot, triggered with an existing end-of-ramp "sync"
pulse (normally used to sync the transitions of the staircase
waveform that's used for Base/Gate drive). But the differentiator
seems like a more straightforward way to get the beginning and
end of the pulse to be in the right places, when dealing with
varying ramp frequencies/fall-times.

Thanks!

Regards,

Tom Gootee, 15MAY2006

tomg at fullnet.com

------------------

.