Splitting sync-on-green video signal
mike3
How could one make an electronic circuit that would split the green
signal from a "sync on green" source (like what comes out of some
types of computers) into the green, horizontal, and vertical sync
components individually?
a7yvm1...@netzero.com
mike3
Hmm. But how about not only extracting the sync but also removing
it from the green line to get a pure green signal? In other words,
have
something where the input is the green+sync, the outputs horizontal
and vertical sync and also pure green component (no sync)?
John Larkin
wrote:
If you really don't want the sync pulses, the thing to do is clip the
negative swing at the "black" level. But why?
John
who where
wrote:
oooh, no! That chip is downright evil.
[O/T] We used them in an olde (pre-GPS) frequency reference, to
extract the sync pulses from rubidium-locked TV transmissions. It
gave spurious pulses, and on further investigation this proved to be
unavoidable. We switched to the Gennum GS4981 and the problem was
history.
a7yvm1...@netzero.com
>
> Hmm. But how about not only extracting the sync but also removing
> it from the green line to get a pure green signal? In other words,
> have
> something where the input is the green+sync, the outputs horizontal
> and vertical sync and also pure green component (no sync)?
Clamp the black level and subtract it from the video, this will bring
the video down by the 0.3V (IIRC) of the usual black level, but like
John said, why?
It makes the circuit complex and you don't see the syncs anyways so
why remove them?
mike3
Because I need to use this with monitors that do not support sync on
green. None of the monitors I have are listed as supporting sync on
green, and I think there aren't many models of monitors that actually
do,
are there? I've got an HNC AH191A monitor and Acer X223W, neither
of which ring up anything when searching for them plus "sync on green"
on Google. I don't want to get rid of the sync, just split it off into
separate
lines, like the monitor requires, and send it a pure green signal down
its
green port, as it requires. That is, to make a thingy that takes in
the
combined green+sync signal in one end and yields horizontal, vertical,
and pure green component signals out the other.
mike3
As I've heard if you try and send a green with sync on green in it
through a monitor that doesn't support it, it'll fill the screen with
lots of green hue, or the monitor might not work right, etc.
a7yvm1...@netzero.com
I'm no expert on monitors, the last time I played with S.O.G. was with
an Amiga and a Commodore 1950 monitor...
Gives you an idea of the time frame.
I don't think a monitor would care if there are syncs or not on the
green. I think the "not supported" means the monitor can't extract
syncs from green (cost cutting), not that it will harm the monitor.
However you could look for an Extron box called a 118 on eBay, for
about 20$ it does all this stuff for you.
20$ or days of fudging around... You choose!
Glenn Gundlach
> On Sat, 9 Jan 2010 13:04:34 -0800 (PST), a7yvm109gf...@netzero.com
Nah, it's just crappy. They worked considerably better with bigger
coupling caps and smaller (?) resistor for the burst flag timing. I
haven't touched the HD version which might be better for the higher
resolutions.
G²
a7yvm1...@netzero.com
> As I've heard if you try and send a green with sync on green in it
> through a monitor that doesn't support it, it'll fill the screen with
> lots of green hue, or the monitor might not work right, etc.
That would make sense since having a sync means the entire black level
relationship is blown if the monitor expects a 0V reference.
It's weird since the Amiga 1950 monitor supports SOG but now that I
think about it there was a switch somewhere on the thing to select for
that mode.
Glenn Gundlach
You'll nead something like the LM1881, a DC restore amplifier to
assure the DC levels and then a 'clamp amp' (clipper) to remove the
sync currently on the green. The Analog Devices AD8036 is the best
clamp amp I ever used and would remove anything below the threshold
level and drive the coax dierctly. It isn't terribly difficult but
also not trivial. You'll also need 2 gain of 2 line drivers to
distribute the red and blue. Mis-terminated analog video looks very
bad.
G²
Glenn Gundlach
It depends on the specific monitor. Different models from the same
manufacturer may behave differently. Better to clip off the sync but
more expensive.
G²
a7yvm1...@netzero.com
Checking my 1950's manual (I'm a bit of a hoarder)
it states that it expects video 0.7V and sync on green is -0.3vpp.
Here's what the 118 manual has to say:
When setting the RGB 118 for Sync on Green, you must also
disconnect the BNC cable from the external “Sync” output of the
RGB 118 and RGB 118 PLUS. Otherwise, sync is output on both
green and the composite sync output which may cause
abnormal levels of green to appear on the presentation display.
That might explain the green you're seeing.
mike3
Why does more processing need to be done on the red and blue lines,
anyway?
mike3
I sounds like this is going to take quite a few chips and other
components. How much would it cost for all those components?
a7yvm1...@netzero.com
> Why does more processing need to be done on the red and blue lines,
> anyway?
Otherwise they arrive before the green to the monitor. (Propagation
delay through whatever circuitry for the green = time delay = shifted
green picture on screen => same delay for R and B to restore picture
quality)
a7yvm1...@netzero.com
>
> You'll nead something like the LM1881, a DC restore amplifier to
> assure the DC levels and then a 'clamp amp' (clipper) to remove the
> sync currently on the green. The Analog Devices AD8036 is the best
> clamp amp I ever used and would remove anything below the threshold
> level and drive the coax dierctly. It isn't terribly difficult but
> also not trivial. You'll also need 2 gain of 2 line drivers to
> distribute the red and blue. Mis-terminated analog video looks very
> bad.
>
> G²
Not to mention the time difference. Unless the OP needs a lot of
these, I'd say go buy an Extron 118 off eBay, it does all of this and
they go for cheap.
Glenn Gundlach
I got the mistaken(?) inpression you were sending to 2 monitors, not
just one. For multiple monitors you need distribution amps. For one
you don't need to do anything.
G²
Glenn Gundlach
If you can get into the source video and disable the sync insertion
onto the green, that would be VERY EASY and nearly no cost. You might
need a line driver(s) to send composite sync or H/V to the outside
world.
G²
Tim Williams
news:f7e8c633-672c-4233...@m3g2000yqf.googlegroups.com...
> Otherwise they arrive before the green to the monitor. (Propagation
> delay through whatever circuitry for the green = time delay = shifted
> green picture on screen => same delay for R and B to restore picture
> quality)
So throw in a few feet of coax. :^)
Tim
--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
mike3
How though would one determine the propagation delay?
whit3rd
Start by separating out a logic signal for 'sync'; a one-transistor
amplifier, with the base DC bias set by the average collector
voltage, will work. You know the sync duty cycle, just
ground an NPN's emitter, put a pullup resistor to +5 on the
collector, and AC-couple the video to the base, DC-couple
a divided-down collector voltage to the base.
Now run two PLLs on the sync signal, one for H, one for V.
Finally, use a video buffer amplifier biased to clip the low SYNC
level
but still have unity gain for the video signal range; it's do-able.
Hal Murray
>> Otherwise they arrive before the green to the monitor. (Propagation
>> green picture on screen =3D> same delay for R and B to restore picture
>> quality)
>How though would one determine the propagation delay?
Standard trick in the (old) digital days was to run both
signals through similar gates.
For example if you wanted to gate a clock pulse, you ran it through
a x00. You also ran all the other clocks (that were supposed to
happen at the same time) through a similar gate.
First chioice was to put all the related gates in the same chip.
Second choice was to use the same type of chip.
Third choice was to use a handy gate of the same logic family.
--
These are my opinions, not necessarily my employer's. I hate spam.
Glenn Gundlach
Um, you measure it with a scope and a test pattern that has white (R/G/
B) vertical lines and measure the delay. You can easily get to under a
nanosecond with a good scope. The delay in the green processing can be
made up by shortening the cable. Any decent analog video guy can do
that.
G²
JosephKK
Recover the separate sync and leave the green plus sync alone. It is NOT
during displayable time; it is "blacker than black" (gun cutoff). It is
not visible and will not hurt the monitor.