Removing RGB jail bars idea?

[MEGADRIVE Jeroi]

OH my, 30 posts and now you find that subcarrier cause noise in image? I already tried to help you to add more filtering for the subcarrier to reduce vdp noise. It's the vdp noise in Subcarrier that causes the jailbars what I think. So I see that my s-video encoder thread can make perfect s-video as the chip dosent need subcarrier only horsontal ja vertical sync and NTSC klock signal to generate S-video from RGB signal.

[Drakon]

OH my, 30 posts and now you find that subcarrier cause noise in image? I already tried to help you to add more filtering for the subcarrier to reduce vdp noise. It's the vdp noise in Subcarrier that causes the jailbars what I think.

I already knew it was the subcarrier and I tried messing with the subcarrier circuit. I posted that ages ago.

[lumclaw]

OH my, 30 posts and now you find that subcarrier cause noise in image? I already tried to help you to add more filtering for the subcarrier to reduce vdp noise. It's the vdp noise in Subcarrier that causes the jailbars what I think. So I see that my s-video encoder thread can make perfect s-video as the chip dosent need subcarrier only horsontal ja vertical sync and NTSC klock signal to generate S-video from RGB signal.

Wasn't I the one who brought up the worst case possibility all hardware based on Megadrive may be affected by this? Then people began expanding upon my idea, to bring it closer to the realm of plausible. Instead of being just a newbie's rambling.

[TrekkiesUnite118]

Yeah we definitely know it's some sort of a subcarrier thing. What I need is someone with a more sensitive tv to try this trick out. Just as long as that person doesn't mind losing the ability to have s-video, composite, and RF. Since the subcarrier comes right from the vdp and disconnecting the subcarrier at the point of the VDP fixes it...we can just assume that the vdp is the source.

Well I wouldn't say VDP is the source. The Source the way I see it seems to be more or less the use of the CXA video encoders and how their pins are arranged. Basically all Genesis systems use a CXA1145/1645 encoder or a third party clone of them. And they all have a pinout that puts RGB right next to the subcarrier. So essentially all Genesis systems have to eventually put the Subcarrier next to the RGB lines, thus creating interference.

If someone had the parts and time to try it I'd suggest trying a video encoder that doesn't have the subcarrier near the RGB lines and hooking everything up to that, subcarrier and all. Then make sure you pull the RGB signals from a point before they get close together and see what happens. One could try replacing the subcarrier trace with a sheilded wire as omp suggested and see if that helps. The only issue I see with that is we don't know how well the inside of the CXA chips are shielded. So we may prevent it outside the chip but inside the chip the interference may still happen and that's beyond our control.

[MEGADRIVE Jeroi]

Dragon, did you dry to use EMI filter on subcarrier? On second taugh, Trekkies have some good ideas there. How ever my S-video sircuit is just like that, take a look. No CLK no H or V sync near the rgb. When I get the parts will test this circuit a lot.

[omp]

I shall have a look at that as well, when I tested the console last night that was free of the 32x. I will also set the tv right in normal mode, no widescreen/full bs. I use 2 scart cables a model 1 type for the sms2 and model 2 type for my mega drive, 32x, cd combo.

I have a model 1 genesis, japanese model 1 and a few stock aussie megadrives. Also got a few genesis 2's, I can have a look at rgb to component on my lcd tv which is a sony xbr40 & report back if you are keen?

[MEGADRIVE Jeroi]

... & report back if you are keen?

Sure would be nice.

[Drakon]

The cxa2075 has a pin that does nothing between rgb and subcarrier. Anyway I wired up my ad724 to the rgb coming out of the cxa1145 because the ad724 needs the rgb to be amped. The s-video coming out of the ad724 is messed up looking (colours all screwed up) but it's sharp enough to see if there's jailbars or not. I ran the system with pin 50 of the vdp lifted and still got scrolling jailbars from the s-video coming out of the ad724. So with no subcarrier connected to any part of the genesis pcb I was still getting scrolling jailbars. To me this tells us two things....1: The ad724 isn't a fix for the jailbars, 2: All video encoders (not just sony) give you scrolling jailbars when they have their own colour subcarrier. The ad724 circuit I'm using is driven by its own crystal for a subcarrier.

....and just for the sake of hilarity:



That's what happens when you wire up an ad724 into a genesis....yeeeeouch.

[TrekkiesUnite118]

Just to be sure, it is hooked up correctly right? It looks like you may have possibly swapped some of the RGB lines.

[Drakon]

Just to be sure, it is hooked up correctly right? It looks like you may have possibly swapped some of the RGB lines.

Haha I know it does look like that but trust me it's correct. There's something about the signals this encoder just doesn't like. After the tmss shows up the s-video just disappears for a few seconds and then comes back so something is really funky..... Also I've wired this encoder into my neo geo mvs it works fine with that. This encoder actually doesn't like a lot of systems I have certain arcade pcbs where I get the same crazy effect going on. I am really glad that my component encoder is jailbar free when I lift vdp pin 50. I also found out that the 32x s-video has slight jailbars but they only showed up when I used one s-video cord. I changed s-video cords and the jailbars went completely away again haha. Anyway the jrok is a bit overpriced and I've got a much cheaper component encoder on the way. Once I get that encoder and test that it works lumclaw or someone with a more sensitive tv could buy it and give it a try. The 50-60$ component encoder even has a scart port right on it so it should just be plug and play for anybody who gets rgb from a scart cable.

[TrekkiesUnite118]

Someone at Sonic Retro should make a Sonic 2 hack that uses those colors and call it "Sonic Acid Trip Edition".

[Drakon]

Someone at Sonic Retro should make a Sonic 2 hack that uses those colors and call it "Sonic Acid Trip Edition".

haha agreed.

Also if it were wired wrong the number "2" would be solid red. As all the background that's supposed to be blue is now turned red. The ad724 simply just doesn't like rgb signals from certain systems which is probably why I've never found one of these encoders in a console.

[Drakon]

Well I know I said I couldn't take screen captures of component video....that was only half true:



Dayum...that's one jailbar free image. Let's take a look at the luma from component video coming from the genesis with vdp pin 50 disconnected:



Not much difference in luma. So I guess the jailbars I was seeing on my component tv were mostly present in the colour lines. These screen captures were taken from component video generated by the rgb coming out of the 32x. But people say the rgb isn't effected by the 32x so *shrug*

[Drakon]

My second rgb to component encoder came in and I wired it up. Much to my surprise the picture was FAR sharper than the jrok so now the jailbars were VERY noticable! I once again disconnected the subcarrier line and the jailbars went away. I did a test. I ran the system with the subcarrier attached to a wire. If I brought the wire close to the cxa1145 without making contact I could indeed see the jailbars become stronger on my tv! I seems I owe trekkies an apology as surprisingly...the colour subcarrier does seem to be "radiating" interference. The picture looks amazing on my component crt but unfortunately that tv the colour green looks off and no matter how much I adjust the tint I can't get it looking right. So I'm still living with s-video. Once my component capture device shows up I'll take lots of before and after pictures so you can see the jailbar difference. The vdp pin 50 trick on assembergames totally works! And it seems lifting the pin of the vdp right from the pcb is the best idea as it's a leaky signal line. I was so shocked when I wired up my new component encoder and saw VERY noticable jailbars...coming right from the rgb. I don't know how tiido gets a jailbar free image, I guess it depends how sensitive your setup is? Like my jrok rgb to component encoder doesn't pickup the jailbars as well but the image isn't nearly as sharp looking as my new encoder. Anyway lifting the colour subcarrier right at the source seems to be the best and probably the only solution for cleaning up the jailbars. But hey..it looks perfect on my pixel perfect setup once lifting the pin is done so it's good enough for me. I mean if I'm like directly infront of the screen squinting hard I can see an exceptionally faint interference that doesn't even look like a jailbar...it's certainly perfect enough.

[TrekkiesUnite118]

Is the new encoder the BA6592F? If so try hooking up the subcarrier to it where it should go on that and see if the Jailbars return. It would be interesting to see if just using an encoder with a different layout that keeps the subcarrier separated would eliminate the jailbars or if the interference is that strong that even that wouldn't be far enough away to get rid of it.

On the BA6592F the subcarrier pin is on the complete opposite side of the chip from where the RGB in and component out lines are if I remember correctly.

Another interesting thing to try would be to possibly weaken the subcarrier signal. Again my electronics knowledge isn't that great so I'm not sure what component it would be but if you could cut the trace and put something in the middle to reconnect it that would weaken the signal just enough to remove the jailbars but still keep it strong enough for Composite and S-video to still work you could have a possible fix.