shmups.system11.org

All of those options SHOULD be disabled on RGB. Enabling them would mean a useless colorspace conversion, inducing a little delay and more potential for quality loss. One of RGB's strengths is that colors wlil always maintain the correct relationship relative to one another. Introducing a saturation control means you're only making your signal shittier. An aperture control also doesn't make sense, as that's used to change filtration parameters for an NTSC signal to try and recover lost luma data. Again, not relevant for RGB.

As it goes, in the KV-27S42's jungle RGB inputs, brightness and contrast / picture (DC offset and Gain, respectively) are passed from the TV controls to the RGB, and the RGB signal is clamped. That is the processing performed, as it should be. The service menu's gamma controls also affect it, as they should.

----------

With regards to RGB to YPbPr, I would never want to use a converter box when tapping in RGB is available. For one, that's just another device to have eating up a spot on the power strip, and another device that must be purchased. More wires and bullshit behind the TV. The quality will be the same, in an ideal world in a vacuum with only perfect parts. The ones that cost $30 perform like they cost $30, and getting a perfect conversion out of them (with or without tweaking of many potentiometers) is asking a lot. Even then, with that conversion said and done, you're still vulnerable to the image processing I described above, which RGB conveniently bypasses - undersaturation, oversaturation, white balance issues, unecessary ringing from filters applied from sharpness control, signal noise from another amplification stage or two... YUV conversions are a chump solution when RGB is directly available. It's great to have, but not a substitute in my view.

They go inside the TV (no extra wires or power strips), and they can cost $30-40 because they don't need to cost any more than that. You're not paying for connectors, or a metal shell, or a power adapter, you're soldering leads directly to it. They do tend to need 3.3V or 5V power, though, but there is power available within the television that can either be used directly, or run through a 7805. HD Retrovision's cables, while retail pricing hasn't been announced, should cost something like that (they kickstarted at $35 a cable, if memory serves), and would seem to do a high quality RGB to YPbPr conversion.

If I'm already opening the TV, mounting a port, and soldering anything, why in the world would I pay for a converter board when the TV can accept RGB in any capacity?

You kind of missed the point of my original post, which was a television whose processor can accept a component image, but not an RGB image. In that case, it would be better to install a converter board rather than throw out the TV, which was the original sentiment.

My tv does already have component inputs. point is though, that the colors look washed out compared to my pvm and will never equal it when using the converter box.

if rgb was available i wouldnt think twice about using it. The benefit over the pvm is the screen size.

You're right, I missed your point. In such a situation, that would be a much more reasonable response.

No they shouldn't...

but if your tv has component inputs : wouldn't it be simpler to just get a Kramer FC-4?

Sharpness doesn't really have much of a place in RGB, not in the way it is usually applied to NTSC and NTSC's cousins. Contrast and brightness, as I mentioned before, are other words for gain and offset, and do have a place there.

Picked up the KV-27FS100 today. Not fun dragging this 109lb beast by myself! I have a few specific questions about the mod:

- why is a 6PDT switch used? Shouldn't a 4PDT suffice? I was under the impression that we're Hijacking R, G, B and injecting 5V. What are the other 2 switches for? Just to be safe, would it make sense to add a Diode on the 5V line to make sure that 5V doesn't feed into the regular RGB generator IC or will the switch suffice as protection?
- when the switch is on / feeding RGB via the SCART, I'll be feeding 5V from the console (that's how all my SCART cables are made). The TV will be set to one of the composite inputs, and the SCART connector will be feeding composite video into that connector (Yellow) + stereo audio (red/white) on the INSIDE of the TV. The TV theoretically will be using SYNC from composite video, and audio too. What happens if I feed it CSYNC into composite video input? Will it still work?
- any suggestion on what V rating the caps should be?

SAFETY QUESTION: What about ground? I don't believe there should be a potential difference between CRT chassis and console. But just to make sure, should GND wires from SCART all connect to GND on the chassis? Or am I asking for trouble here?

You're right about not needing a 6PDT switch, I used it only because it was on-hand. A 4PDT is fine. For switching the blank level, a high value resistor (10k) is fine as a pull-up resistor on the OSD input. 5V being fed to the input is fine, because that's what it expects. There's nowhere else for it to go. I got my 5V from the 5V line in the TV, but if you're providing it from SCART that's fine too. The resistor will make it safer if you fear a chance of making a mistake. As for making sure it isn't given 5V when you want the OSD instead, the resistor will make it a bit safer but the switch will cut it off completely. With the pull-up resistor it will at worst make it always active, but won't damage the MCU generating the OSD signal.

Giving CSYNC to the composite input will do just fine. There may be some sets that will be grumpy about this, but in my experience this has always worked well. If CSYNC is available, use it.

We're talking tiny ceramic 0.1uF caps, with very low signal voltage. The rating should not be anything to worry about.

For a safe ground for discharging the thing, the metal grounding strap that touches the outside of the CRT's aquadag coating is a good bet. For your signal ground, use the ground on the Jungle IC. You can probe around with a multimeter to verify that other locations are equivelant to that ground. On my set, the signal ground on the jungle IC and other control chips was common with the chassis ground, but be sure to check first using your multimeter.

I'm trying this on a Zenith B19A21D (Daewoo CN-220A chassis) and I'm having good results except that the image is much darker and less brightly colored than composite. At first I tried going direct from SCART RGB pins to the RGB pins on the LA76814 jungle IC, and then I tried it with almost the same circuit mikejmoffitt is using (75Ω to ground, 75Ω inline, and a 1.0U electrolyte cap because that's the only type I have). They both look the same, very dark and drab (but wired directly the jungle does a bit of seeking to find the proper input level). I do have 0.1 ceramic caps on order, how likely is that to solve the issue?

The really curious thing is this TV goes to all-white in composite mode, and with FB high (full screen RGB), the picture brightens up and looks great. I'm worried the CPU is controlling the OSD contrast and raising it only at that specific moment. Do any of you have ideas?

I never mentioned a 75ohm series resistor for the termination. That's why your picture is dimmed.

Take that out of the signal path, and use .1uf decoupling caps. Your Jungle I/C probably already has them in the circuit path on your RGB inputs.

I removed the series 75, and swapped the 1.0U lectro for .1U ceramic. It still looks exactly the same: very nice, but very dim.

Current path for each color is:


And it still looks excellent for 1 second after I push the power button off.

I'm going to hook the OSD back up and see if I can workaround the problem in the service menu by increasing the power on the R/G/B guns and then drop the contrast on the composite picture to match them back up.

Any more tips would be appreciated. If I can get a good result of this I'll post some kind of writeup or video about what I did. This Zenith is a fantastic TV even though it originally had no inputs whatsoever, only RF.

Try putting a 1K+ pullup on your low voltage source that feeds the blanking pin.


What's the part # of that jungle I/C. Although possible (though unlikely) it wants something like 1.0vpp or 1.2vpp video. Though, I find that highly unlikely.

I'm assuming that your RGB cable is wired correctly, yes?

The blanking pin on my jungle blanks on high, so a pullup wouldn't do me any good. Also I don't think there is any issue around the blanking because I can connect the OSD blanking and open the full screen menu and it looks the same as when I pull it high.



LA76814.



I have 2 retro_console_accessories cables with csync which I can trust are wired properly. And mostly I'm testing with a cheap chinese SCART cable for my SNES, which WAS wired incorrectly (with an open composite ground) but now is wired up correctly. The chinese cable is nice because it still has composite video so I can flip back and forth.

-
Edit: I think I may have solved it with a factory reset and a few other fiddly things... stand by.

I keep saying this in this thread. Did you look at the datasheet for your Jungle I/C? I take it you didn't...

Let's take a look at a few snippets of data.

Jungle I/C Pinout with application circuit (pay mind to pins 14-16):




Now, let's take a look at the wave-form / signal data for those RGB inputs.



What do you see? Right there's the answer.

Proper decoupling caps are already in the signal path, per the pinout. A nice integrated clamp there on the RGB inputs, too.

Looking at the wave-form data, the RGB inputs are designated for good 'ole fashioned standard video signal levels. .7vPP..

Terminate, terminate, terminate.

Apparently, my set (with its RF-only input) configures its own color settings automatically from the first channel it receives over RF after a factory reset. So, switching back to RF and doing a factory reset rebalanced some internal settings which are not accessible to the service menu, and made those settings which I can access such as RGB drive/bias and Subbrightness/Contrast become more effective. So, the wiring diagram I posted above is working great once I figured that bit out.

If anyone wants to mod this same chassis model (and I would highly recommend it) I can provide more detailed assistance on what I needed to do. I'm done for tonight but maybe soon I'll make a video or photo/write-up on this project.

https://pbs.twimg.com/media/CZT1hcJWwAA6SoA.jpg:large
https://pbs.twimg.com/media/CZT1hcJWQAAxw6b.jpg:large
https://pbs.twimg.com/media/CZT1hcJWcAASDO0.jpg:large

RF-only!? Go on Craigslist and you can find much newer CRTs for free.

Why? If he likes the set and managed to RGB mod it already kudos to him. No need to ditch it if he likes it and managed to get RGB out of it.

Giving a set like that RGB is probably the best possible future for it, really. It wouldn't be tremendously useful otherwise.

Funny story. I was forced to take the Zenith in order to get a Trinitron 27FS100* for free, they were a package deal. I'm glad I did! I tested out the Zenith with my Atari 7800 and was amazed how nice the picture was for an RF signal (AFAICT the Zenith doesn't have a 3D comb). I figured the TV would look even better with RGB SCART and made it my guinea pig, and it worked out great!

TVs like this are a great first set to practice RGB modding. If you FUBAR the motherboard by melting a bunch of components or traces, you've damaged a TV that wasn't valuable in the first place. If you succeed, you've turned a turd of a TV into a gem. And really - this thing has STEREO! What's stereo sound doing on an RF-only set...

* The 27FS100s are excellent TVs. They come with component, which you can just use an RGB-Component converter with.

For Atari stuff with the super-wide pixels, I'd prefer a non-3D comb filter. The tradeoff with that sort of comb filter is pretty visible dot interference along wide thin lines, vertically.

Since the flood gates have apparently been opened on this one...

I have a Toshiba MW24FN1R that i've been looking to do this mod on. Its one of those dvd/vcr combo sets. From what I can tell it has a separate chroma processor and IC, and the schematic shows a OSD_R/OSD_B/OSD_G between the 2. I also saw 0.1uf ceramic caps when i looked at the circuit. Thats about as far as i've gone with it. Am I wasting my time here?

service manual: https://www.scribd.com/doc/295769155/MW24FN1RSVM

Looks like a good one to look at. OSD R/G/B is your analogue signal (try the same 0.1uF + 75ohm termination) and the OSD Y signal will be your blanking signal.

It took about an hour or two, but I fixed purity / convergence on the set. I realigned the yoke and played the convergence-ring-balancing-act game and it turned out pretty well. At the very top, blue is SLIGHTLY lower than it should be, but this is leagues better than how it was:

looks gorgeous now

it was totally worth it to do these last ajustments

Nice. I couldn't ever fix the convergence on my Trinitrons.

Well, I gave this a try on my MW24FN1R and I couldn't get it to do much more than give me a blank screen. I lifted the RGB pins on the IC out of the circuit (not fun btw, very small pins) and did the 0.1uf cap/75 ohm resistor setup and +5v to blanking like Voultair suggested. I may try it with 10uf caps but im all out for now, will have to order more.

https://drive.google.com/file/d/0B0j8MA ... w?ths=true

OSD_Y is your blanking, that's what you should be tying your low voltage source to.