The NEW Build Your Own Arcade Controls
Main => Monitor/Video Forum => Topic started by: Foxhole on January 21, 2017, 02:54:25 am
-
Hello.
I own a sony crt tv model: kv29ls30e, and it seems it doesn't support csync correctly.
When using csync the colors seem to be dull, red looking like brown, for example.
This doesn't happen with sync on composite video or luma.
To get around this problem i would just use consoles with rgb and sync on composite or luma, but i also wanna use my pc (crt emudriver) which is connected via vga to scart or via UMSA, which only offers csync as the sync, so i'm stumped as to what to do now.
Apparently, i'm not the only one with this problem: http://shmups.system11.org/viewtopic.php?f=6&t=58861 (http://shmups.system11.org/viewtopic.php?f=6&t=58861)
Is there a way to connect the pc using a different sync or some way to fix the tv itself to support csync?
Thanks.
-
Posted in the thread you linked.
What video card are you using for crt_emudriver, out of interest?
EDIT: if the UMSA is buffering the video, it could be that too, I guess. Probably not, but skipping it altogether might be worth a try.
-
I'm using hd7850 but tried with 5450 too and the same thing.
It's not specific to the pc, but rather every system that uses csync.
For example, i have two cables for my genesis, one wired for csync and one wired for sync on composite, and with the csync the colors get distorted just like with the pc or every other system that uses csync.
I'm also stumped as to why the sync is causing this but i have feeling it's a screw up by sony.
Any chance you know of a way to convert csync to something like sync on composite, or convert the H and V sync from the pc to something like sync on composite?
Thanks.
-
It's a weird issue. I'd love to hear if it ever gets sorted. I believe all a stripper does is pull the pulses, so to mux it back you could just try adding the two lines back together (mb with a diode from ground pointing to the signal, if you want)
For the PC, I breadboard out a whole load of stuff. You can just add small caps and resistors (for H-sync line, anything from 1K down, for Vsync line, anything from 3.3K down, in my experience) and see what happens. Try to keep at least one resistor *between* the V- and H-sync lines, so they can't feed one another directly. No way those cards won't be buffered, but best to be sure.
You can start by googling all all the sync combiners you can and seeing what comes up, starting with eviltim's on this page - http://members.optusnet.com.au/eviltim/vga2arc/vga2arc.htm. (http://members.optusnet.com.au/eviltim/vga2arc/vga2arc.htm.) A lot of them will be for other devices, TTL level stuff, old SUN Microsystem monitors, etc. But it's worth a look.
My preferred circuit for my older PVM's is a 1K resistor on the H-sync line, and a 3.3K resistor on the V-Sync line, then combine the two. I find that with a lower value on the vertical line, the flicker comes back pretty hard.
-
oh, and i should add that i have no idea whether trying this sort of thing will work, but i'd guess probably not. You need the help of someone like Eviltim, but everyone wants his help, and he only has so much time and patience to go around. Present him with a problem he wants to solve, and he probably will. But everyone's like that i guess :)
There's a thread on Shmups in the hardware section about an Ikegami monitor, and in it Tim goes into modifying sync timing. Which, again, may or may not be relevant, but it's a place to start. I can't seem to find it, but you might search his posts to find it.
You could also look into creating Sync-on-green, I suppose. A lot of the circuits on the net for that are ancient though, and use ancient and not-very-capable chips. There's some stuff in a thread on Shmups by a user called samiam about it. You'd need to build a cable driver amp and split it, once into the sync line, once into the green line. Unless the set takes SoG, which was something Sony alone pushed for as standard so there's a reasonable chance.
-
I'm sorry for necroposting but i think this is important for anyone with the same problem.
apparently RGB-PI managed to find the issue and fix it: https://github.com/mortaca/RGB-Pi/issues/12 (https://github.com/mortaca/RGB-Pi/issues/12)
unfortunately i don't have this model anymore, but i hope this helps to anyone who has.