Composite sync stripper for the schematic illiterate

[HercTNT]

you get vertical lines down the screen. this image will give some idea.

http://www.google.com/imgres?q=rgb+jail+bars&num=10&hl=en&tbm=isch&tbnid=cd0c_xqmvzjJvM:&imgrefurl=http://www.sega-16.com/forum/showthread.php%3F21099-Removing-RGB-jail-bars-idea/page10&docid=x142Nc-Gyav2RM&imgurl=http://img.photobucket.com/albums/v54/theclaw/theclaw%252520sale%252520blowout/10-11/P0005928.jpg&w=1440&h=960&ei=hTCKUPLFBsXd0QGF94GQDg&zoom=1&iact=hc&vpx=811&vpy=210&dur=4166&hovh=183&hovw=275&tx=131&ty=124&sig=109559872676753118431&page=1&tbnh=143&tbnw=208&start=0&ndsp=26&ved=1t:429,r:15,s:0,i:116&biw=1600&bih=733

[NightWolve]

Ooooh, heh, that makes sense... Thanks. I did try googling before, but I could only find topics just talking about it.

[thesteve]

the parts on pin 6 are not needed for C-Sync out

[rCadeGaming]

A few things:

You can get raw c sync directly off of the PCE/TG16 graphics chip, so a sync separator is not necessary needed.

Some SCART TV's/monitors actually need full composite video as sync instead of raw c sync, so you may want to consider a selector switch if you might ever run into this.  However, if do you find something requiring composite, you can try a 1.2k resistor on the raw sync line to knock it down to .3Vpp as it would expect to find in composite.

For a few more dollars you can use an EL4583 to get real separate H and V sync instead of C and V.  Although, this is only needed if you have a picky piece of equipment, as C and V sometimes works fine for RGBHV (VGA).  The EL4583 isn't much more difficult to use though; keep in mind that the video filter is optional, and not needed for clean sources.

Using a pot in place of the rset resistor (680k resistor) will give you a good amount of horizontal position adjustment.  This is useful, as these old systems use different resolutions that can be all over the place.  This works with either the LM1881 or EL4583.

[rCadeGaming]

Also, the diagram here is missing something very important.  Using a decoupling capacitor will prevent and kind of shakyness/noise/whatever due to small fluctuations in the +5v line.  This means a 0.1uf capacitor connected between the +5v input on the chip and ground.  There's no reason not to do this with most logic chips, and it's pretty mandatory on a video circuit like this.

You can also try 220uf capacitors on the sync outputs (in line, not to ground), but results might vary depending on the rest of your setup.

[HercTNT]

well what does work and what should work are not always the same thing. I have tapped my Duo for raw sync when using my rgb and it did not work. Building the sync stripper was not only dirt cheap but effective. You can throw all the technobabble at it you want, but in the end, this simple cheap circuit worked just fine.

Edit: did not mean to make this sound snippy. what you said is fine and all, its just been my expirience, that things don't always go as they should. Sometimes the cheapest solution is the best one. In my case, for my application, the stripper was dirt cheap, easy to make, and solved the problem.

[thesteve]

the sync from the video chip is 0.3V, making it not acceptable for most RGBS monitors.
it can be amped and works fine, but this chip can be in your cable.
as for the coupling cap, its un needed if the chip is close (short wires) to the system, if sharing a ground with the video path it becomes a requirement

[rCadeGaming]

the sync from the video chip is 0.3V, making it not acceptable for most RGBS monitors.
it can be amped and works fine, but this chip can be in your cable.

Well of course.  Both the RGB and c sync signals coming from the chip should be amplified.  I thought this was a given for a standard RGB mod.  Sorry if it was unclear.  This is a proven amp circuit that I used:

http://www.gamesx.com/grafx/pce_rgb.jpg

2SC1815 transistors are hard to find these days; an NTE85 will work the same.

Of course if you skip amping the c sync, you'll never never need a 1.2k resistor to bring it back down to 0.3Vpp, as it's already there.

well what does work and what should work are not always the same thing. I have tapped my Duo for raw sync when using my rgb and it did not work. Building the sync stripper was not only dirt cheap but effective. You can throw all the technobabble at it you want, but in the end, this simple cheap circuit worked just fine.

Did you not amplify the c sync coming off the chip?  Whether you did or not, it probably wasn't the right level if it didn't work.  Or the monitor could have been really picky about the pulse width.  Both of these are set to a certain level with the LM1881, but it's possible you could gotten them to good values without it if you know what's going on.

Whatever the reason, there is an explanation that you would call "technobabble," that would be helpful to know instead of having to just go by trial and error.

[ApolloBoy]

2SC1815 transistors are hard to find these days

Not really... http://tinyurl.com/a3zvt6a

[HercTNT]

the sync from the video chip is 0.3V, making it not acceptable for most RGBS monitors.
it can be amped and works fine, but this chip can be in your cable.

Well of course.  Both the RGB and c sync signals coming from the chip should be amplified.  I thought this was a given for a standard RGB mod.  Sorry if it was unclear.  This is a proven amp circuit that I used:

http://www.gamesx.com/grafx/pce_rgb.jpg

2SC1815 transistors are hard to find these days; an NTE85 will work the same.

Of course if you skip amping the c sync, you'll never never need a 1.2k resistor to bring it back down to 0.3Vpp, as it's already there.

well what does work and what should work are not always the same thing. I have tapped my Duo for raw sync when using my rgb and it did not work. Building the sync stripper was not only dirt cheap but effective. You can throw all the technobabble at it you want, but in the end, this simple cheap circuit worked just fine.

Did you not amplify the c sync coming off the chip?  Whether you did or not, it probably wasn't the right level if it didn't work.  Or the monitor could have been really picky about the pulse width.  Both of these are set to a certain level with the LM1881, but it's possible you could gotten them to good values without it if you know what's going on.

Whatever the reason, there is an explanation that you would call "technobabble," that would be helpful to know instead of having to just go by trial and error.

agreed my problem stems from getting the most i can out of my systems for the least amount of money thesteve, dou-r, turbokon, and mike have been invaluable in achieving these goals. for the average use, the terms you guys throw out is technobabble no matter how affective it is. we just don't understand waht you saying. based on that, the simplest of solutions to fix the problem, become the best. mike has gone along way to help me solve problems on my own merit. otherwise. guys like thesteve, turbokon, and duo-r have filled the gap were i have failed.

[rCadeGaming]

2SC1815 transistors are hard to find these days

Not really... http://tinyurl.com/a3zvt6a

LOL, ok.  Well they didn't have them in stock with the vendors I normally use.  Just thought it could be useful to know that NTE85's are equivalent.

[thesteve]

just a note on the gamesx amp
its a current amp only (still outputs 0.3V)
my monitor and some others i have looked up require 4V sync
the 1881 provides 5V sync, that plugs directly into my monitor

[rCadeGaming]

Durp!  You're right, didn't look closely enough.

[bacteria]

A VGA sync needs V-Sync and H-Sync I believe, V-Sync is on the picture, how about H-Sync?

[vxbinaca]

Are there good images that demonstrate this "jailbar" effect ? I still don't know what it refers to...

Look into the NES-2. It's a classic example of out-of-the-box jailbars.