PS2 31khz SoG Green Screen - Resolved

[mesmer]

Hi All,

I built a PS2 VGA adapter like this to connect my PS2 to my tri-sync rgb monitor (nanao ms2930). The differences I have are:
1. No 7404
2. Used electrolytic caps, with + site towards PS2
3. Used PS2 pins 1 and 3 for ground, not 8
4. Only connected CSYNC from lm1881, not VSYNC
5. Added a switch for the input to the lm1881, between green and composite video, for 15khz and 31khz operation

It works wonderfully in low res. 480i looks beautiful in 15khz, as long as the screen isn't moving!!

I then tried a 480p game. Flipping the switch to connect green to the lm1881 works correctly, the screen syncs. However, I get a green tint. The picture looks great, no interlacing, but green. I assume this is because the sync information is still in the green line.

I thought the video info was positive and sync negative, shouldn't the monitor ignore the signal below ground? Is there something wrong with my ground? Do I need to take the caps out of the rgb lines?

I'm afraid I will have to build a circuit to remove the signal below ground, to remove the sync info from the green line. I found this one that viletim posted on another forum. Will this work? Has anyone done this?

Thanks.

[antron]

any viletim circuit is worth a try.

I can recommend this to turn RGsB into RGBV(H):
http://media.extron.com/product/product ... type=1&s=4
looks brilliant

a current ebay:
http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... 602_263622

[Pocket_Man]

have you tried to turn on rgb instead of yuv in the system settings of your ps2?
maybe you will need an rgb fix on your ps2.

[dieKatze88]

have you tried to turn on rgb instead of yuv in the system settings of your ps2?
maybe you will need an rgb fix on your ps2.

This. YUV overdrives Green and sends changes down the other lines, so Green looks washed out on YUV. Try switching your Output mode to RGB.

[antron]

It works wonderfully in low res. 480i looks beautiful in 15khz, as long as the screen isn't moving!!

I think this means he has it set to RGB.

mesmer, note that some PS2 games (and most PS1) can output 240p

it goes green in 480p because the sync signals are still on the green line

[mesmer]

Yes, I have RGB set in the system menu. 15khz works great: converting cvideo -> csync with the lm1881.
It's 31khz that has the problem. 31khz is sync on green. Using the lm1881 to get csync out of the green line works great too. It's just that the picture is too green. I assume this is because the sync information is still in the green line.

Antron, cool I may give that a try. Yes, I will use the 240p patch for the low res games, thanks. But for the hi res ones, I would rather have 480p than 480i =)

[viletim]

There's a little space in the video signal right at the end (just before the start of a new line) called known as the 'back porch'. It's very narrow, but it's always at black potential. The monitor is supposed to use this area to clamp the video signal. But it's tricky to do, requires different timing every time the horizontal frequency changes. Some monitors cheat and instead of generating a back porch clamp signal, they clamp on the horizontal sync. This works perfectly so long as the videos signal remains at black level through the entire time there is no active video (blanking interval).

Putting sync on the video signal is not such a bad thing if you apply it to all three colours, but sync only on the green signal is a bastard of an idea. The only way to make it work with a sync-clamping monitor like mesmer's is to either remove the sync information from the green video or to add it to the red and blue video signals too. It's easier to operate on one signal than on two, so the usual approach is to fix the green.

The circuit linked to is one of mine, but it's very, very crude. Reduced to the bare minimum of cheap components. It should work, though it may need adjusting for each different video source. There's an app note by Intersil (http://www.intersil.com/data/an/an9514.pdf) which has a better sync remover circuit (designed with intersil parts, naturally) which would likely outperform my version.

That said, the box antron links to looks like a bargain.

[mesmer]

Excellent, thanks.

If this is any more evidence to confirm: The green goes away (no green at all) as I move the horizontal position pot way to one side or another. As I move the image to be horizontally centered, the green sort of "fills up" from the bottom, like a glass filling with water. With it being the highest when the image is properly centered, reaching almost the top.

[mesmer]

Looks like there is a simpler circuit with less features in the el8102 datasheet. I'll see if I can get a sample.

[RGB32E]

That said, the box antron links to looks like a bargain.

There's a reason why I've been hyping decomissioned commercial gear for years now...

[viletim]

Looks like there is a simpler circuit with less features in the el8102 datasheet. I'll see if I can get a sample.

This won't work with the PS2. It's only for direct coupled video where black is 0v and sync falls below.

[mesmer]

This won't work with the PS2. It's only for direct coupled video where black is 0v and sync falls below.

Ok, am I understanding this correctly?:

The PS2 RGB output is some unknown positive offset, black > 0v
Running through the 220uf coupling caps results in an "ac coupled" signal, where the positive peak and negative peak have the same absolute value?
The first part of the circuit in the app note is a "dc restore" circuit. Which results in a +0.7v -0.3v signal with black at 0v, this part needs the back porch signal from the lm1881 - I've seen this in other circuits.
The second part (also the one I screen-capped) will remove the negative parts of the signal, thereby removing the sync.

[viletim]

This won't work with the PS2. It's only for direct coupled video where black is 0v and sync falls below.

Ok, am I understanding this correctly?:

The PS2 RGB output is some unknown positive offset, black > 0v
Running through the 220uf coupling caps results in an "ac coupled" signal, where the positive peak and negative peak have the same absolute value?

More or less. In theory the absolute positive peak of the AC coupled signal is the bias voltage + the peak to peak value and the absolute negative peak is the bias voltage + the peak to peak value (ie. a 0.7Vp-p terminated video signal, when capacitvely coupled without bias will swing between +0.7V and -0.7V in absolute terms as the video varies between a single white dot in the centre of the line to a completely white line). How fast the DC level changes is dependent on the value of the coupling capacitor and the impedance of the circuit.

The first part of the circuit in the app note is a "dc restore" circuit. Which results in a +0.7v -0.3v signal with black at 0v, this part needs the back porch signal from the lm1881 - I've seen this in other circuits.

No, this circuit doesn't clamp the video signal at all. It's a sample and hold circuit. It's function is to store the current line's DC level of the back porch in capacitor C1. This DC voltage is then subtracted from the video signal passing through the line driver, IC3.

The second part (also the one I screen-capped) will remove the negative parts of the signal, thereby removing the sync.

Correct. It relies on biasing the video signal in such a way that the output clipps on the negative rail and cuts off the sync pulse. Not the most elegant way to do it, but it should work.

I can already see a problem with this circuit. There's no external bias on the input, only a 10k resistor to ground this means IC3 needs dual supplies, something like +/-2.5V. I can't see see any other way it can work. That means the output would below 0v which doesn't sound right...

It needs a bias network to run on a single 5v rail - Here's how I'd build it.

[mesmer]

Thanks alot for the explanation, it helps alot. I decided to get one of those Extron RGB 150xi units off of ebay. I could only find the el8102 for sale in SOT package, didn't sound fun to try to solder to. No SOICs were available. I found the Extron on ebay from a local liquidator, and ended up only paying $21 total!

I'll update you all when I get a chance to test it out.

[antron]

cool mesmer, note that the unit will only output a different signal type from the source using the bnc output, not the pass-through.

[mesmer]

Yeah, I just picked up a VGA - 5 BNC cable, and a VGA "extension" cable with male on one end and female on the other.

[mesmer]

Results: It works!

It took a little trial and error. The sync wasn't stable using RHBHV output with RGsB input (31khz PS2), but worked fine with RGBS output. With RGBS input (15khz PS2) both HV and S output worked fine. Also, my monitor works fine with 640x480 RGBHV from my PC. So, just using the 4 BNC connection for RGBS from the Extron to the monitor works great.

Here is another one from the same seller I got mine from.

Antron, Have you played with the 'DDSP' dip switch? I've left mine on, which seems to disable some extra sync processing. Here is the weird thing: With the DDSP switch off, and the PS2 in 15khz 480i, the extron seems to convert to 240p. Have you seen this behavior? I believe the sync processing is to ensure digital (lcd, dlp) display compatibility. Which makes sense because LCDs can't use an interlaced signal.

Big thanks tor Antron and Viletim

[antron]

Antron, Have you played with the 'DDSP' dip switch? I've left mine on, which seems to disable some extra sync processing. Here is the weird thing: With the DDSP switch off, and the PS2 in 15khz 480i, the extron seems to convert to 240p. Have you seen this behavior? I believe the sync processing is to ensure digital (lcd, dlp) display compatibility. Which makes sense because LCDs can't use an interlaced signal.

I had not idea it could accept interlaced low resolution signals. I can't wait to try, nice find! We need to try with an xbox 360!

Some LCD TVs can accept 240p

[mesmer]

Some more info

150 marketing info

Sync processing
is accomplished through Extron’s ADSP™ Advanced Digital Sync
Processing technology. ADSP provides all-digital processing of sync signals,
eliminating incompatibility issues encountered when using analog sync processing
with digital display devices (DLP, LCD, Plasma, etc.).

* Built-in Advanced Digital Sync Processing (ADSP™) – All
digital sync modification process insures flawless operation
with any LCD, DLP, plasma or other digital display device.

160 setup guide

Digital Display Sync Processing (DDSP™) — Set to off (down) for sync processing; set to on
(up) for no sync processing (may be needed for some LCD and DLP plasma displays) .

This seems like the most clear, non-marketing explanation

160 manual

ADSP™ (Advanced Digital Sync Processing) — Allows sync processing operations, such as horizontal centering, to occur without affecting the signal’s sync timing. This allows horizontal centering to be applied to signals that are output to digital display devices, such as LCD projectors, DLP (digital light processor) projectors, and plasma displays.


DDSP™ (digital display sync processing)
On — If this switch is set to on (up), the interface does not perform sync processing. This may be necessary for digital display devices, such as LCD (liquid crystal display), DLP (digital light processor) and plasma displays.
Off — If this switch is set to off (down), the interface performs sync processing operations, such as horizontal shift, using Extron’s ADSP™. Turning on the DDSP feature disables the horizontal shift control.

Marketing turning a disable switch into a "feature." =) My guess is the engineering name for DDSP was "Disable Digital Sync Processing"

From my testing, setting DDSP off (sync processing on) converts 15khz 480i -> 15khz 240p.

[skyline7]

From my testing, setting DDSP off (sync processing on) converts 15khz 480i -> 15khz 240p.

This is great news! I've been scouring these forums for an inexpensive way to pull 240p out of a 480i signal. Just won a 150xi on eBay, so I'll do some testing with PS2, Xbox 360, and Dreamcast once it shows up and let everyone know the results. Right now PS2 is my only real problem child (I have an Ultracade UVC hooked up for DC and 360 use), but I'm curious to see what it'll do to the 360 and DC.

[antron]

Hey mesmer, for the PS2 in 480i did you you strip out sync to feed to the 150ix?

And does anyone know if the 360 has c-sync on SCART?

[mesmer]

yeah, cvideo -> csync using LM1881

I got the parts from digi-key:


1 2 478-3192-ND CAP CER .10UF 50V 20% RADIAL 0 0.16000 $0.32
4 2 CF1/4680KJRCT-ND RES 680K OHM 1/4W 5% CARBON FILM 0 0.08000 $0.16
5 1 LM1881N-ND IC VIDEO SYNC SEPARATOR 8-DIP 0 3.28000 $3.28
12 3 P829-ND 220UF 6.3V MINI ALUM ELECT (KA) 0 0.15000 $0.45

2 2 17EHD-015-S-AA-0-00-ND CONN HD DB15 FEMALE GOLD FLASH 0 1.98000 $3.96
13 1 CP1-3543-ND CONN JACK STEREO R/A 3PIN 3.5MM 0 1.00000 $1.00

[dieKatze88]

I picked up a Extron unit. I'm waiting for BNC to VGA cables to show up this week so I can test it. I will DEFINITELY test the shit out of the idea that 480i becomes 240p.

[skyline7]

Just got mine in the mail as well. I'm hoping I have time to solder together a SCART to HD15 adapter today so I can test it with a PS2.

[dieKatze88]

Hooked it up.

Some Gradius V time? I think yes.

http://img205.imageshack.us/img205/4747 ... adison.jpg

240p, I'll be damned. Although the image is still "interlace jumpy" and its a little stretched from being displayed on a 1080p monitor (which is great for Otomedius... for what its worth)

This box works SO good though.

The spec sheet says it doesn't do YPbPr. well mien won't do RGsB from a PS2, it displays correctish colors from YPbPr though, so I'm not going to complain.

[antron]

dieKatze88, how are you running G5? With HDXploder (480p) or just the G5 disc by itself?

[dieKatze88]

dieKatze88, how are you running G5? With HDXploder (480p) or just the G5 disc by itself?

I usually play off of HDLoader (latest version) at 480i because I ripped the game myself from my legal copy.

I haven't yet made a hacked version of Gradius V that has the 480p patch and installed it to my HDD yet.'

Didn't answer your question, oops.

That night I was playing on my real copy, as my HDloader setup is acting up.

[antron]

dieKatze88, how are you running G5? With HDXploder (480p) or just the G5 disc by itself?

I usually play off of HDLoader (latest version) at 480i because I ripped the game myself from my legal copy.

I haven't yet made a hacked version of Gradius V that has the 480p patch and installed it to my HDD yet.

I see, why does your screen capture say VGA? Your TV takes a low-res signal on a PC connection? What does it say if you bypass the Extron? (and what model are you using?) Thanks for the tests.

[dieKatze88]

dieKatze88, how are you running G5? With HDXploder (480p) or just the G5 disc by itself?

I usually play off of HDLoader (latest version) at 480i because I ripped the game myself from my legal copy.

I haven't yet made a hacked version of Gradius V that has the 480p patch and installed it to my HDD yet.

I see, why does your screen capture say VGA? Your TV takes a low-res signal on a PC connection? What does it say if you bypass the Extron? (and what model are you using?) Thanks for the tests.

because it is vga. its out of the extron, I can do as low as 15khz on these monitors, it works with any 15khz signal, but it has to be RGB S or RGB HV or HDMI.

Thats a monitor, btw. A really nice one

[antron]

I think the proper term would be CGA.

So, you lose picture if you set the PS2 to RGB? I quess you don't have a sync splitter on the c-vid line.

But you get correct colors when you set to YPbPr? Wild. I guess it gets the sync from the "green" line, that makes sense. But how are the colors right!?