HDMI to Old CRT

[Zebidee]

About that pencil note on your schematic - they wrote "RED | BLUE | GREEN | WHITE". White against BLK.

This made me think about your friend worrying about contrast levels and pull-ups.

Blanc and blanco etc, in many languages, mean white. In most video signals, there is a luma (composite, svideo, component all have it) which is, more or less, brightness and everything else except colour. This luma is carried on the same signal line as the sync. But not so for RGB

In RGB, all that luma data is rolled into the colour signals. The sync is just the sync The blanking is just a switching signal and nothing else, not "white". When the voltage is between X and Y, OSD RGB is switched on.

In terms of pull-ups: most times you'd input external RGB and not use the OSD; but leaving the OSD's RGB lines connected is important for pulling up the voltage. Between the OSD, the jungle IC and RGB inputs, the resistors form a voltage divider matrix, which ideally delivers 0.7vpp to those 104 capacitors and RGB pins.

[abstract3000]

Ok so the TV was altered to the suggestions made above but still no dice here is the current schematic of how it's set up


This is what I got:
I have made the suggested modifications and am even more convinced the issue is the jungle is looking for a TTL video signal.  If I run a CGA TTL signal in we get a perfect picture.  Keeping all the parameter the same and simply changing the signal to analog you get nothing.  Signal is being generated from a Sencore CM2125.

Adding the potentiometer on the blanking line does give me control of the blanking threshold and changing the resistor values I think made the picture crisper but still will not work with an analog signal.[ /i]

As you can see we just put a db15 plug on the TV and running VGA to vga cable from PC directly to TV.

I'm going to try and play with the CRT emudriver software some more tomorrow.

[Zebidee]

OK, that does look about right to me.

Make sure that the PC with CRT_emulator is outputing 15khz signal properly. Try testing it with another 15khz display/TV/monitor, if possible.

If no other display to test on, and if you have access to one, hook up an oscilloscope to check out the sync.

[abstract3000]

Ok so checked the Software and Still no dice, it is in the emulation mode giving out the 15.7, but as mentioned the Sencore CM2125 Machine has the settings 15.8 with the 640x 480 set signal sending the screen. When switching the machine to digital we get a Screen, but when we turn off the digital for analog it disappears. This has us convinced the signal is being locked out due to being Digital RGB, at this link https://crtdatabase.com/modding/rgb-via-mux you can see he states the following criteria is a "Must" for the RGB Mux Method:

• If your jungle does not have RGB inputs you cannot use this method.
• If your jungle has digital RGB inputs you cannot use this method.

we did in fact test the signal on it's way to the Jungle with an oscilloscope and this is what we got:

Digital input gave the following:
.978 (5.28pp)


Analog Gave the following:
.128 (2.3pp)


All throughout the SHUMPS Thread it is mentioned that if the Jungle is Digital RGB it's a no go. So this had us seriously discouraged. Though after reading for a few hours through everything I stumbled upon a post that discussed the Sanyo LA76843N Chip which you linked earlier "IX3354CE" chip is really a "X3354CE" chip which is really an "LA76843N" <- That was quoted from MarkOZLAD.

So the Thread I stumbled upon discussed how Voltage needed to be injected into the RGB Lines for this Chip to get the Signal. This explains in the first attempt of the mod, why when handling that little circuit board I had with both the auto on circuit and the RGB mod, when messing around with it I managed to get an image displayed as I must have been accidently introducing the 5V from the On/Off circuit to the RGB lines. You can see The picture here on the screen:


So It appeas from his workup on Stack Exchange here: https://electronics.stackexchange.com/questions/568045/single-supply-op-amp-non-inverting-level-shifter-for-ntsc-signal-processing-from/568154?noredirect=1#

He uses the THS7315 OP amp and drives it up to 2.5Vpp, though my friend suggests it might get better with 3.5Vpp. Not sure if you have any idea on this work around, or most likely being a one off. The issue is all the people involved in this model of TV MarkOZLAD, rx7turbo233, Syntax, and benyamin39 have not logged into the forums over there for months or in a couple cases years. The main thread has ran stagnent with the tail end just being numerous people asking for help with no response from those that once lead it. I understand the times change and people move on, but it would be great if I could at least get ahold of the Ben guy to see what the final results were.

I tries responding to his thread quoting it "Crossing my fingers" it triggers an Email and he checks, but luckily it appears to be allot of detailed info he left behind that at least we have something to work with.

[Zebidee]

Ok so checked the Software and Still no dice, it is in the emulation mode giving out the 15.7, but as mentioned the Sencore CM2125 Machine has the settings 15.8 with the 640x 480 set signal sending the screen. When switching the machine to digital we get a Screen, but when we turn off the digital for analog it disappears. This has us convinced the signal is being locked out due to being Digital RGB, at this link https://crtdatabase.com/modding/rgb-via-mux you can see he states the following criteria is a "Must" for the RGB Mux Method:

• If your jungle does not have RGB inputs you cannot use this method.
• If your jungle has digital RGB inputs you cannot use this method.

I get this now. The TV can show analog fine, but the digital OSD RGB inputs won't read analog.

we did in fact test the signal on it's way to the Jungle with an oscilloscope and this is what we got:
...
Analog Gave the following:
.128 (2.3pp)

I don't understand this, as you say you are measuring sync before it gets to the jungle IC. Are you measuring the analog sync at (or close to) the composite/AV input jack? Analog sync should still have the square waveforms, much like the digital input picture you showed. If there is no sync from the VGA card, might explain some things.

Check the sync both before and after any modifications (resistors etc).

So the Thread I stumbled upon discussed how Voltage needed to be injected into the RGB Lines for this Chip to get the Signal. This explains in the first attempt of the mod, why when handling that little circuit board I had with both the auto on circuit and the RGB mod, when messing around with it I managed to get an image displayed as I must have been accidently introducing the 5V from the On/Off circuit to the RGB lines. You can see The picture here on the screen:


So It appeas from his workup on Stack Exchange here: https://electronics.stackexchange.com/questions/568045/single-supply-op-amp-non-inverting-level-shifter-for-ntsc-signal-processing-from/568154?noredirect=1#

He uses the THS7315 OP amp and drives it up to 2.5Vpp, though my friend suggests it might get better with 3.5Vpp.

The blanking signal might have multiple states? Like, it might be one voltage level for digital, another for analog. When your finger touches the 5v for blanking, it may be pulling the voltage down (like a voltage divider, some of it being grounded through you), triggering analog. It's just a theory, but it is quite common for different blanking voltages to trigger different things. Could be worth putting that pot back on and going through the range while inputting analog.

You could try a video amplifier from Ultimarc. They are relatively cheap, a bit more powerful, and I think they supply a VGA cable with it.

What puzzles me about this is that if this amping idea works, I'd expect to see at least a faint image without it. Maybe there is a threshold.

Not sure if you have any idea on this work around, or most likely being a one off. The issue is all the people involved in this model of TV MarkOZLAD, rx7turbo233, Syntax, and benyamin39 have not logged into the forums over there for months or in a couple cases years. The main thread has ran stagnent with the tail end just being numerous people asking for help with no response from those that once lead it. I understand the times change and people move on, but it would be great if I could at least get ahold of the Ben guy to see what the final results were.

I tries responding to his thread quoting it "Crossing my fingers" it triggers an Email and he checks, but luckily it appears to be allot of detailed info he left behind that at least we have something to work with.

Those guys, especially MarkOzlad and Syntax, would probably know what is possible. They both post on the FB CRT groups. Mark is one of my friends and I've seen him active recently, at least on the Aussie CRT group. I'll PM some details to you so you can contact him.

[abstract3000]

Well, My friend told me had an Idea he wanted to try and I didn't hear back for a week it appears he ordered one of the Video amplifiers from Ultimarc as you suggested Zebidee, and in Positive news the Picture is now on the Screen (but appears to be inverted/ black/grey Scale with White Text) We are not understanding at what point the Video could have become inverted, but my friend wanted to give a shot at building an inverter to see if that fully corrects the issue. So waiting for those results.

Had I known this Sanyo Jungle was such a pain in the Ass I think I would have just spent more time trying to find a 27" CRT on FB marketplace to buy that either had the RGB inputs already or was something like Sony and well known to be capable of modding.

The theory we have with this TV is that the Jungle RGB Inputs are Digital therefore as speculated amongst the community is non moldable, but I believe we are bypassing that by raising the Analog Signal to the Same Voltage as Digital and fooling the chip to take it so to speak, but how good it will turn out at this point is anyone's guess :/

[Zebidee]

More options for you:

"Neckboard" mod

Hacking directly into the jungle's RGB output to the neckboard. You don't necessarily have to do it physically at the neckboard either, you can probably do it right on the main chassis, before (or even at) the neckboard connector.

I've never actually done this myself. I know it is possible though. The main trick is to figure out the voltage levels.

The schematic I have shows RGB out on pins 19,20,21. It shows expected voltage levels for those pins, at 2.3 2.4 2.5 respectively. About 2.5v then. The block diagram (click on the thumbnail above) show RGB Vcc at 9v, so seems it uses that to give the RGB outputs a fair boost to ~2.5v.


     


That ultimarc amp should boost the RGB voltage to approximately same level, so worth a go.

Still need to terminate each RGB input with 75R resistors, and after that a ceramic cap (100nF 104) and a 100R resistor in series.
Like this x3, hijacking the neckboard connector (should be 5 pins, RGB+GND+9v). Alternatively, tap into the base of the 100R resistor already there. If the RGB is too bright, try pots (low value, probably 100R or 200R range sufficient) in series before the 75R terminations.


RGB input------?POT?-------------------CAP 104------100R------neckboard connector
                                           |
                                         75R
                                           |
                                        GND



Input the sync via composite as usual.

Pot on blanking signal

With the existing mod, and if your mate hasn't already done this (not sure if he has?), put a 3-pin pot on the blanking with middle pin to ground, to make a voltage divider. Try the full pot range. Different voltages may trigger different result. That's my "Hail Mary" for the OSD approach. Otherwise I think try hacking the neckboard/jungle output.

Inverted picture

Regarding the "inverted" video, you can easily address that by flipping the yoke coil connector. If you mean upside-down, flip the vertical coil connector. If you mean mirrored, then flip the horizontal connector. Each coil has 2 pins, typically joined as a 4-pin yoke connector. You would need to cut that in the middle, between the pins, to separate the H and V coils. Flipping the coils won't help with correcting the colour though.

[abstract3000]

Thanks for the Info Zebidee, My friend Advised from the get go that hacking directly into the Neck was an option but one he would prefer to leave alone. It appears I spoke too soon in reference to what we thought was an inverted signal.

He got the picture coming through and in Color none the less, apparently the issue was the decoupling Caps, that were never present before, removed those and the Signal passes right through. Though 2 Things are a a bit off (the colors look really good an vibrant) but The screen resolution looks the same as it does on my PC monitor, so 640x480 is not what the TV is liking, so Going to have to find a way to Change the resolution it sends over.

The second thing is what we believe to be the Vertical Sync, there is some waviness of sorts around the screen, but I see in VMMaker adjustments can be made, I did play with it a bit when i was over there today and i saw it did change it, so will have to play with it a bit more.

[Zebidee]

Thanks for the Info Zebidee, My friend Advised from the get go that hacking directly into the Neck was an option but one he would prefer to leave alone. It appears I spoke too soon in reference to what we thought was an inverted signal. He got the picture coming through and in Color none the less, ...

Good news then! I was hopeful that you could do the input via OSD because you said you got a picture briefly, which means it must work somehow.

... apparently the issue was the decoupling Caps, that were never present before, removed those and the Signal passes right through.

Try different values for the decoupling caps, perhaps one or two orders of magnitude lower or higher.  The theory is that different capacitor/resistor combos resist different frequencies. High-pass or low-pass filters. I don't really want to go into theory I barely understand myself, but this is possibly what is blocking the signal.

Capacitors in parallel add total capacitance, so when testing it is easier to start small and go big. You just solder the higher capacitor onto the legs of the earlier one. If you used 104 (100nF) caps before, remove them and start with some 102 (1nF) or 103 (10nF) first. Then try some 105 (1uF) caps, even 10uF if you dare. Because 1.011uF is about the same as 1uF, you don't have to remove the smaller caps for this kind of testing.

Though 2 Things are a a bit off (the colors look really good an vibrant) but The screen resolution looks the same as it does on my PC monitor, so 640x480 is not what the TV is liking, so Going to have to find a way to Change the resolution it sends over.

Not sure I understand what you mean there. Pics? Even with CRT_emulator, there remains a "native" 640x480 that is a VGA resolution hardwired in by Windows (thus "native"). Go into Arcade_OSD and change your desktop to the "custom" 640x480 interlaced res. Or something else

I'm mostly using 720x480i@60hz on my TVs desktops now - it is essentially NTSC standard. A tiny bit of letterboxing, but despite that the pic looks a lot crisper (an extra 80 pixels on each line) and seems easier to fit most of the desktop on screen, with little overscan. TV seems designed for it! Also saves me that native/custom confusion. I just add it to my user_modes list.

The second thing is what we believe to be the Vertical Sync, there is some waviness of sorts around the screen, but I see in VMMaker adjustments can be made, I did play with it a bit when i was over there today and i saw it did change it, so will have to play with it a bit more.

There are all kinds of possible reasons for this. You sorted out that resistor on the sync input yeah? Maybe your mate could hook up the scope again and have a look, it looked wrong before (more like static). You need to zoom in to like 200us or less to really see it properly, because an entire line of sync is only ~64us long.

[abstract3000]

Try different values for the decoupling caps, perhaps one or two orders of magnitude lower or higher.  The theory is that different capacitor/resistor combos resist different frequencies. High-pass or low-pass filters. I don't really want to go into theory I barely understand myself, but this is possibly what is blocking the signal.

I discussed this with My friend he stated he did tried a range with the caps and got nothing, hence removing them he realized it most likely wasn't going to go with them.

Not sure I understand what you mean there. Pics? Even with CRT_emulator, there remains a "native" 640x480 that is a VGA resolution hardwired in by Windows (thus "native"). Go into Arcade_OSD and change your desktop to the "custom" 640x480 interlaced res. Or something else

I have gotten into the Arcade_OSD Swapped the Resolutions then put into the Emulation mode with multiple resolutions to no avail, I even tried changing the resolutions from the desktop while it was being emulated to the CRT with no Change.

See the Pic below, it's almost as if the Screen Area it's filling is larger than the actual "Viewing" area


I know this most likely has nothing to do with it, as they are separate but the LCD Monitor I'm using shows the Same damn thing when I put into Emulation mode


So the other issue I can't capture in a video or a picture is the "Flashing" maybe what I'm seeing is a Refresh issue? It occurs around Windows Boxes, menus etc, and the text in the Windows Explorer is pretty garbled, but when you put a Picture or a Video up, you don't get this quick "flashing" (excuse me I'm sure this has a proper term, but i don't know a better way to describe it)

There are all kinds of possible reasons for this. You sorted out that resistor on the sync input yeah? Maybe your mate could hook up the scope again and have a look, it looked wrong before (more like static). You need to zoom in to like 200us or less to really see it properly, because an entire line of sync is only ~64us long.

At this point I am on my own I have the TV back and he stated he did everything he could do with it. I appreciate all his efforts, but I know it was taking up space in his shop he needed back and he really went through a lot to get it finished.

He did however leave the micro pot on the board for me to fine tune the blanking signal in the event i want to use the Svideo Option instead, but in his belief he really thinks there is something to do with the Vertical Synch and hopefully the Software/Drivers I'm using Can adjust that. I think I wil try setting up MAME on the System now and loading a Game to see how it even looks for that matter, but was hoping I could do something to get the screen adjusted so you could see the entire Windows Menu Bar at the bottom

[Zebidee]

OK, looks like you've got it working anyway. It's hard to take good screenshots, but your picture looks pretty clear and you seem to have all the colours.

I have gotten into the Arcade_OSD Swapped the Resolutions then put into the Emulation mode with multiple resolutions to no avail, I even tried changing the resolutions from the desktop while it was being emulated to the CRT with no Change.

I don't understand what "emulation mode" is? You mean EDID emulation? I usually just leave that alone. Despite the name. it has nothing to do with actually using game emulators. It concerns fooling the computer into regarding your TV as a proper monitor. I'm not really an expert with CRT_emulator, but think I recall someone saying that whenever you click that, you should re-generate and re-install the video modes.

See the Pic below, it's almost as if the Screen Area it's filling is larger than the actual "Viewing" area

The pics you show actually look fairly normal. Overscan (the desktop hidden behind screen edges), especially in 640x480i, is typical with TVs.

Later, you can trim the frontporch/backporch of the CRT_emulator modes to minimise overscan, but it will probably never be entirely gone. I get less overscan with desktop set to 720x480i. You can also have a look around in the TV service mode geometry adjustments, but be careful as sometimes it is possible to set values for which TV cannot display an image! Then you'd be stuck.

So the other issue I can't capture in a video or a picture is the "Flashing" maybe what I'm seeing is a Refresh issue? It occurs around Windows Boxes, menus etc, and the text in the Windows Explorer is pretty garbled, but when you put a Picture or a Video up, you don't get this quick "flashing" (excuse me I'm sure this has a proper term, but i don't know a better way to describe it)

Do you mean that the picture does not stay on screen, comes and goes ("flashes") fairly quickly? Maybe rolling picture? But stabilises once you have a window open? This is a strong indication that there is no vertical sync coming through. Possibly no sync at all. This is because the TV, in a desperate move, attempts to guess where the sync is by identifying gaps in TV video signal input. The video signal is more distinct when you have windows on the desktop, making it easier for the TV to guess at beginning and end. You might lose sync for a moment if you try moving windows around on the desktop.

How did you end up doing the sync? You still using the Ultimarc video amp? That outputs composite sync already. Take that, run it through a resistor (depends on sync voltage, but try from 470R up to 1K) into the TV's composite/AV input.

Alternatively, your "flashing" around window edges might simply be that you are noticing interlacing in modes like 640x480i. Interlaced modes refresh lines at half frequency, so there is enough time to draw two lines "interlaced" with each other, in the same time it takes to draw one line in progressive modes like 320x240p. Total screen refresh rate remains the same, twice the lines but each is drawn half as often. The slower line refresh rate is much more obvious around straight lines.  Hard to notice with non-linear images, and you probably won't be able to notice it at all with video, which is why it works for normal TV.

[abstract3000]

but your picture looks pretty clear and you seem to have all the colours.

Yeah the picture is clear, yet not so sure about all the colors the Whites are "Really Really Bright" I also opened this Jpeg on the Monitor


And this is What I got

Do you mean that the picture does not stay on screen, comes and goes ("flashes") fairly quickly? Maybe rolling picture? But stabilises once you have a window open? This is a strong indication that there is no vertical sync coming through. Possibly no sync at all. This is because the TV, in a desperate move, attempts to guess where the sync is by identifying gaps in TV video signal input. The video signal is more distinct when you have windows on the desktop, making it easier for the TV to guess at beginning and end. You might lose sync for a moment if you try moving windows around on the desktop.

Ok so think of the Windows in the Explorer, and the black lines that make up the border or the windows or the Menus that pop up, now think of a fluorescent light bulb on its way out, how it starts doing that rapid flashing of light (dark bright, dark bright, in rapid recession) and around the desktop Icons themselves the screen is jittery/wavy so your stating refreshing is redrawing itself rapidly but not fast enough for my eyes not to see it I assume that's probably what I'm seeing? Though open a Normal Picture or play a video you don't see any of that (just the Windows components of the operating System)

How did you end up doing the sync? You still using the Ultimarc video amp? That outputs composite sync already. Take that, run it through a resistor (depends on sync voltage, but try from 470R up to 1K) into the TV's composite/AV input.

The synch was just piped on pin 13 through the DB15 last I knew to composite, yes the Ultimarc Video Amp is still in use but not enough power by itself so it is also connected to the PC for power as well.  If I unplug that from the PC power supply I get No screen, the instant i plug it In I have screen.

I tried the 720x480 I had available to me in the Arcade_OSD and the Icons dropped even further down in the overscan. The last strange issue I took a video of is when i grab a window and drag it around, when it gets like to the sides or most noticeably the bottom corner is bends the picture as if it's pushing what's there out of the way to make room (Like the clock in the bottom right corner) The clock doesn't disappear but rather bends/distorts slightly like it's making room for the incoming window and when you move it away it bounces right back to normal.

[Zebidee]

One last time: put the sync through a resistor 

Without a video, what you describe sounds like you are just noticing interlacing.

The last strange issue I took a video of is when i grab a window and drag it around, when it gets like to the sides or most noticeably the bottom corner is bends the picture as if it's pushing what's there out of the way to make room (Like the clock in the bottom right corner) The clock doesn't disappear but rather bends/distorts slightly like it's making room for the incoming window and when you move it away it bounces right back to normal.

This is "blooming", a very common thing with CRTs. As the image gets brighter, more voltage is being pumped at the CRT phosphors and the picture gets a bit bigger. Better CRTs do it less, but they all do it. Many games actually take advantage of it. For example, try loading up Golden Axe and let it run in demo mode - it takes advantage of blooming when the title comes up on screen.

From that first pic of the test colour field, looks like you may need to adjust the flyback, brightness and RGB cutoffs in the service menu. While you are there, you might find some geometry controls for V and H-size or similar.

You probably also want to play with the CRT_emulator monitor presets, and/or adjust modelines via arcadeOSD, to minimise overscan.

In general though, well done.

[abstract3000]

One last time: put the sync through a resistor

I sincerely apologize, I meant to comment on this with my last post but dropped the ball. As mentioned I am alone on this project at this point forward so I will need to source those resistors (470R up to 1K) and most likely a small breadboard, there's no Radio Shacks around so It will be a bit of a drive for me to get those. So that might be a bit before I can give that a shot.

In the meantime I am setting up MAME, the last set I had procured and finely tuned was .206 and I cannot find a version of GroovyMame that far back even through the archives, So transferring the base set now of .260 and when I get that all set up with CRT_Emudriver I want to give it a shot and see what I get on the Screen to give me an Idea of how much time and effort to keep dispensing on this, and fine tuning it out so I have some decisions to make I suppose.

I assume the MAME setup I should have done by tomorrow night, and will probably drag everything to the monitor out in the garage to screw around with it on Thursday some more and have a better Idea of where I really stand with the picture and the purpose.

There are 2 knobs on the back of the Flyback, I have considered looking into giving those some fine adjustments, and as in the Service menu you mean the typical menu you change brightness and Contrast levels? I was going to give that a Shot as well.

I will update you a bit more in a couple days where I'm at and will start looking for the Resistors, will just trace out pin 13, cut it between the small board and the TV Motherboard, plug both ends into a bread board and take turns swapping out the resistors to see if there is any change

[Zebidee]

One last time: put the sync through a resistor

I sincerely apologize, I meant to comment on this with my last post but dropped the ball. As mentioned I am alone on this project at this point forward so I will need to source those resistors (470R up to 1K) and most likely a small breadboard, there's no Radio Shacks around so It will be a bit of a drive for me to get those. So that might be a bit before I can give that a shot.
...
I will update you a bit more in a couple days where I'm at and will start looking for the Resistors, will just trace out pin 13, cut it between the small board and the TV Motherboard, plug both ends into a bread board and take turns swapping out the resistors to see if there is any change

You can buy resistor "kits" online, which have a bunch of different values, for a few dollars. 1/4W rating is plenty for most DIY low-voltage DC stuff that we do. This is really just for your TV's protection from over-voltage and ensuring there are no other issues going on -  but it seems to be working as-is for now.

While you are online, grab some heatshrink tubing. Also, make sure your soldering iron is ready to go   You should not need a breadboard for this - just cut the wire, splice the resistor in. When happy, slide some heat-shrink over the join.

In the meantime I am setting up MAME, the last set I had procured and finely tuned was .206 and I cannot find a version of GroovyMame that far back even through the archives, So transferring the base set now of .260 and when I get that all set up with CRT_Emudriver I want to give it a shot and see what I get on the Screen to give me an Idea of how much time and effort to keep dispensing on this, and fine tuning it out so I have some decisions to make I suppose.

I assume the MAME setup I should have done by tomorrow night, and will probably drag everything to the monitor out in the garage to screw around with it on Thursday some more and have a better Idea of where I really stand with the picture and the purpose.

You can download Groovymame 0.206 here

There are 2 knobs on the back of the Flyback, I have considered looking into giving those some fine adjustments, and as in the Service menu you mean the typical menu you change brightness and Contrast levels? I was going to give that a Shot as well.

The bottom knob is the screen voltage. Top knob is focus. Set brightness and contrast in neutral/middle setting, and put a dark screen or game on display (Galaga works OK). Now adjust screen voltage up a little bit, then down until the flyback retrace lines disappear completely. Flyback retrace lines are the faint and slightly diagonal lines you'll see, especially near the top, when screen voltage is high. Once you have the screen voltage right, adjust the focus until everything looks as sharp as possible.

Adjusting flyback is not the same as brightness/contrast, nor is it a substitute. It is more about getting the base level right, and you don't want to be adjusting flyback screen voltage every day. Your brightness/contrast are there for day-to-day fine-tuning.

When I was talking about RGB cutoffs, they are not the same as brightness/contrast either. Cutoffs are the minimum level at which the colours will be displayed. You might see them in the service menu as "RC", "GC" or "BC", or maybe something else. You will not see them in normal TV menu, they are not the same as brightness/contrast. Your display seems to be bright enough, but lacks colours at low brightness levels. That's why I ask about the RGB cutoffs.

[abstract3000]

Thank you so much for giving me the link to that .206 Version!

So as for the Sharp 27R-S100 Set that project I suppose is officially dead. :/ After grabbing the photos for you the other day and shutting everything down that was the last time That Set ever turned on. When I came out the next day to try MAME, the Set would no longer even power up, I checked all the solder points, tops of Caps, fuse etc. and have no idea. I txt'd my friend to ask him what his thoughts were and no Response, he is at the end of his rope with the TV and rightfully so.

I went out and purchased a Panasonic CT-27D11E, which has both S-Video and Component YPbPr Inputs, I bought a $15 S-Video Video card to fall back on. Though I am no longer interested in this RGB Mux mod stuff, It sounds awesome but between paying for the mods done on the last TV and buying the New TV I'm out $525 and nothing to show for it (back to square 1) Where i could have just simply purchased a $15 S-Video GPU, and had it running like it used to. So a bit bitter about the entire thing honestly and feeling defeated, just want a working monitor, and finding the Panasonic set was a diamond in the rough and required a bit of a drive. The other option was another of the Same set and an 8 hour round trip.

I do however greatly appreciate all the time and effort everyone has lent in helping me on this tragic journey

It seems though that you can use CRT_Emudriver with a transcoder to then filter to YPbPr and get great results, though I think you again need a transcoder. There is this one on eBay that states it handles 240p - 1080i
https://www.ebay.com/itm/174166352619

If thats the case, then I will just get that and call it a day, otherwise I might just plug S-Video back in to finally put a close on the project.

[Zebidee]

Thank you so much for giving me the link to that .206 Version!

No worries

So as for the Sharp 27R-S100 Set that project I suppose is officially dead. :/

I went out and purchased a Panasonic CT-27D11E, which has both S-Video and Component YPbPr Inputs,...
It seems though that you can use CRT_Emudriver with a transcoder to then filter to YPbPr and get great results, though I think you again need a transcoder.

You can get fantastic YPbPr results with a transcoder, which is exactly what my GreenAntz VGA->component transcoders do. Just now, I've been running two cabs side-by-side, one with a component-modded generic 21" TV + GreenAntz and the other with an RGB/SCART Panasonic 28" TV. The CRT_emulator setup in both is identical in each, no special setup required. In most cases I honestly can't tell the difference between YPbPr and RGB.

Sorry for hoisting my own flag, I'm between revisions on GreenAntz ATM so I couldn't sell you one right now anyway. Hopefully I'll have some to go before Christmas. However, if you can get one of JAM's quite decent wakabavideo ones, grab it because most of the time he's sold out too.

[abstract3000]

You can get fantastic YPbPr results with a transcoder, which is exactly what my GreenAntz VGA->component transcoders do. Just now, I've been running two cabs side-by-side, one with a component-modded generic 21" TV + GreenAntz and the other with an RGB/SCART Panasonic 28" TV. The CRT_emulator setup in both is identical in each, no special setup required. In most cases I honestly can't tell the difference between YPbPr and RGB.

After doing allot of reading over the last couple months on this, I had kind of gotten the idea that had i not had the S-Video but rather YPbPr Component, I would have been allot better off to begin with and never would have went through all this. I saw several screen shots of people using Transcoders to Component, and honestly I don't care of RGB is somehow better I don't even think I would have been able to figure it out and see up that closely. I never had an issue with the S-Video to begin with but these are the tolerance differences we are looking at here:

SVIDEO VS RGB

I Could go confront my friend on the dead CRT, but i feel even if he did manage to get it resurrected, we would be back to fine tuning that thing and hoping we could work through most of the obvious issues, always wondering in the back of my head when would it go dead. I know he has done allot and is tired of working on it, just as tired as I am hauling it back and forth and discussing it with him so I let the sleeping dog rest and figured learn from my mistake. It's almost meant to be with this new CT-27D11E, I Set my search radius a few hundred miles, I figured I'm on Vacation and have time to drive. Only 1 CRT I could find that matched the specifications 27", Curved Screen, YPbPr inputs. Crazy how scarce these are getting, but the guy was willing to negotiate and not as far as I expected.

So I went ahead and bought 2 Transcoders I bought this Cheap Chinese one off of Amazon because it will arrive tomorrow and I can continue to play with it while on vacation:


I read several reddit posts, a write up on the Shmups board on it, and watched a Video Review looking at the comments sections with several confirmations that running from PC with CRT Emudriver to Component worked great. I also spent the $80 (cringe) on wakabavideo's Transcoder but that wont get here till later next month i suspect coming all the way over from NZ. I saw it up for sale yesterday with 5 remaining for 530 originally available, when I logged in this morning 1 left. Figured it must be a good one, and will most likely last allot longer than the chinese one, but the amazon special will get me by to get everything setup.

Though once again I really appreciate all the help from you over the last couple months on the Sharp TV I will post some pictures when i get everything setup.

[Zebidee]

Don't give up on the Sharp TV, though may be wise to take a break from it. Hopefully it is a simple problem.

SVideo quality is quite OK, but for your mame setup will still look mediocre. The bigger issue is the video card will be outputting in a squished & interlaced video mode, either 640x480i or NTSC 60hz. You won't get any benefits from running games in original pixel-perfect video modes.

You'll have noticed the SCART input on the Chinese units - this is because they don't do anything to combine the separate H+V sync to composite. You have to do that yourself.

Those cheap Chinese ones ripped off an early GreenAntz design I published 5 years ago, on another forum. A mate published his version on Github. Chinese ones didn't exist until about 9-12 months afterwards. Some version of my design and now you buy them on Amazon & Aliexpress. Since publishing, I (and my mate) solved the sync and various other improvements. So, that is why the cheap Chinese ones don't do the sync. I also why, even though I believe in principles of open source sharing cooperation yada yada, I don't publish designs anymore.

With the China ones, you'll need to make or hack (or buy) a VGA-SCART cable. You'll also have to combine the sync somehow. Simplest way is to twist H+V together and run that in to SCART pin 20 through a 1K resistor. This is not perfect but probably sufficient for most TVs, though some may have trouble locking on and many will show artifacts, like a zig-zag at top of screen.

[EDIT] You can get decent composite sync via the CRT_emulator/VMMaker option. You will still need the resistor.

You don't have to worry about all that with a GreenAntz unit - just setup CRT_emulator and plug it in. Comes supplied with 4 standard PCB offset legs, making it super-easy to install in cabs (just screw it down!).

[abstract3000]

Well, the SHARP TV is up and running again, messed with it today and got it going. The picture is back up, tested with Games and my suspicions were right, the interlacing issue goes away and the games seem to display just fine. Though offset a bit which stated could be fixed with adjusting the horizontal and vertical porch settings though this brings up 2 new things today, I need to think about.
 
(though before that, yes I did order the resistors, though when I spend some time with the multimeter tomorrow on my friends mod I have a feeling I'm going to find a resistor on the synch as he is pretty wise individual, but If not I will add it.)

1st issue. I am convinced something is up with the EDID Emulation when sending the signal through, what you are calling the offset is identical on the monitor and the TV, what has me convinced of this is with the new Chinese adapter I got today I hooked that up to the Panasonic ran the signal through and it displays just fine but the image is offset identical on that one as to the monitor and the other CRT and even more to a degree. Why do I believe this offset is an issue? Well in the first images I showed early on getting the signal through the TV displayed the entire image of the desktop with the Task bar fully visible. So I'm guess with all my playing around with Arcade OSD and CRT_Emulator I messed something up. I get that the EDID emulation isn't the "important" part but you don't get a signal unless its turned on.

2nd issue is with colors, maybe you might have an idea what it means, I discovered today with the SHARP that the Green shows sometimes for some things but is missing in allot of other things and I almost believe if this issue is fixed that would be the end of most everything

Here is an example of Paperboy on the SHARP CRT


Now here's an image of Paperboy on the Panasonic CRT


In your expert opinion what does that mean? Or What would be the root cause of that?
You mentioned to go checkout the RGB cuttoffs, I got the service manual and studied it a bit, and found out how to get into the service menu. Here in the service menu I have these particular options:



I highlighted the most obvious, but which one of those would be the cutoffs?


oh and here is an image from the Panasonic connected to the cheap Chinese adapter, this is out of the box no adjustments, as well Groovymame has not been configured with CRT_Emudriver either, I think the colors look really good actually.

[Zebidee]

1st issue. I am convinced something is up with the EDID Emulation when sending the signal through, what you are calling the offset is identical on the monitor and the TV, what has me convinced of this is with the new Chinese adapter I got today I hooked that up to the Panasonic ran the signal through and it displays just fine but the image is offset identical on that one as to the monitor and the other CRT and even more to a degree. Why do I believe this offset is an issue? Well in the first images I showed early on getting the signal through the TV displayed the entire image of the desktop with the Task bar fully visible. So I'm guess with all my playing around with Arcade OSD and CRT_Emulator I messed something up. I get that the EDID emulation isn't the "important" part but you don't get a signal unless its turned on.

Errr...., "offset" is your term, the only time I've used the word here was regarding PCB offset feet which are used to physically screw down a GreenAntz unit.

As you've discovered, EDID emulation either works or it doesn't. It has nothing to do with your overscan issues. If it ain't broke leave it alone. I recall someone saying somewhere that every time you click EDID emulation on/off, you should re-generate and re-install the video modes. So keep that in mind.

2nd issue is with colors, maybe you might have an idea what it means, I discovered today with the SHARP that the Green shows sometimes for some things but is missing in allot of other things and I almost believe if this issue is fixed that would be the end of most everything

In your expert opinion what does that mean? Or What would be the root cause of that?

Yeah, your Sharp TV definitely lacks green, though not completely. Seems to be absent entirely at lower levels. However it seems to be present if the signal is strong enough. Which is why I think about RGB (green) cutoffs (not offsets, presets or porches...).

Does the Sharp TV lack green with other inputs? If so, maybe there is an issue with the green drive circuit or the cathode in the tube itself.

Another possibility is that with S-video (& YPbPr), the Y signal carries both sync and luma. Luma mostly concerns brightness, but is also important for determining green in particular. If you are putting sync in via composite/AV input, and there is no resistor on it, then it could possibly be swamping the green.

To explain... sync signal is simply "ON" most of the time. Sync pulse is triggered by low-active state, a lack of voltage. Sync is carried on the Y (luma), which also includes picture information. By running the sync in from VGA at 5v, you're essentially blasting luma at more than max into the TV, through an AV input designed for 1v max. Of that, the actual sync part of that signal is only meant to be 0.3v max.

This is why I go on about the resistor on sync. This is really the bare minimum.

(If it was me, I'd also use a Schottky diode (IN58XX, reverse biased, cathode to signal, anode to GND) to CLAMP the sync to GND. A second schottky diode, in the opposite to orientation (forward biased), would then limit the sync to approximately 0.3v max. This is CLAMPING, it just keeps sync within normal range and makes it easier to for the TV to work out which is what. I've harped on about this more than I've wanted to. Please show me more pics!)

Anyways, the jungle blanks AV input (except the sync) when the OSD is active (that's why they call it a "blanking" signal), so I'm entertaining a hypothesis that the jungle is blanking a field of green, and that somehow means you have little green for your RGB OSD mod. Easiest way to test this theory would be to disconnect the blanking voltage and see what kind of image you get - if you see a bright screen, either white or greenish, then that could be the problem.

Another cause could include the Ultimarc amp (if it isn't boosting green as much), but this seems unlikely.

You mentioned to go checkout the RGB cuttoffs, I got the service manual and studied it a bit, and found out how to get into the service menu. Here in the service menu I have these particular options:



I highlighted the most obvious, but which one of those would be the cutoffs?

You'll have to experiment for yourself. Every TV is different, and each manufacturer seems to use different terms, so I can't be certain. All I could really do is make educated guesses.

XXX_BIAS are likely equivalent to "gain", concerning colour brightness. XXX_DRIVE might be colour intensity, but it also might concern the cutoffs (minimums) I mentioned. Sometimes I see the cutoffs listed as "RC", "BC", "GC".

You should also look out at settings like "YC", which probably concerns a balance between Y (luma, brightness & sync) and C (chroma, colour).

H_BLK_R and H_BLK_L may be for tweaking the horizontal position.

BS_START and BS_GAIN sound like they could be worth adjusting, but I'm not sure what they might be for.

I wonder what R/B_ANGLE is about, whether it concerns the colour mix.

Leave settings like TINT, CONTRAST, COLOR, BRIGHTNESS at neutral settings, at least until you've bedded down your setup a bit.

SHARPNESS should be set to zero (00) or a neutral setting. Go with 00. It is for making crap TV or VCR signals look less crap - doesn't do anything useful for RGB or component inputs.

You may have other pages of service mode adjustments - my China TV has 20 pages! Most are not accessible without entering more service mode codes.

[abstract3000]

ok so this is going  to be a headscratcher, i have spent a bit more time with the TV and this is the info I have gathered

Let's start off with the Resistor on Synch. It is there and always has been, i looked at the schematic posted earlier and it is on there, I used my multimeter verified this and the fact it is a 1K resistor.
I have reposted the Schematic below, this is the current setup except for the decoupling Caps that were removed.



Here is a Picture of the modification going to the TV


This is the Ultimarc Board connecting to a project board, the more important thing to note is that there is a trim pot on that board connected to the blanking, that will make more sense as I continue on with my observations.

So i found a Panasonic 3DO laying around that had an S-Video Output, I connected it to the TV swapped the Input and everything looks great colors wise, nothing seems to be out of place. I managed to snap this photo that depicted several shades of green all appearing correctly.


Now lets move back to the RGB Mod and discuss the Blanking.
You See how the Picture looks fine? Well I can't just switch from the RGB to S-Video and have a picture with great colors, I have to adjust that trim pot all the way to the other side to get S-Video to show up. Now lets Say S-Video is to the Right, and RGB is to the left (don't thing about that too much i might have it backwards) Anyways in the middle sits is a black screen and distorted graphics and what I believe is called the OSD (On screen display?) It's the text that appears on the screen when you change a channel, or hit the menu button. (Don't shoot me if I'm wrong) Well there is a sweet spot you can get to where you can see that text just barely visible unless over light colors it will be black.

This is what Paperboy looks like when the Blanking has been adjusted to where I will be able to see the menu


Now here is a photo of the menu pulled up and here is the funny thing, look at the letters themselves; they are the accurate colors of what's supposed to be displayed below them on the Screen?!?!?!


So that's where I'm at, I also tried to get into the service menu but at this point i feel it's a lost cause, So the only way I can get into the service menu is by adjusting the blanking as seen in the previous photos, then putting into regular channel mode (no RGB or SVIDEO) the menu refuses to come up in either of those 2 inputs. Then when I get it up it looks like this:


I mean I can really struggle through changing a value if needed, but when I change the value I have absolutely nothing but static to see what impact said change even made so I left it alone.

[Zebidee]

I'm guessing that there is a problem with the blanking voltage. When you use the OSD it overlays the image, showing correct colors, because the OSD is sending the right blanking voltage.

See that 75R resistor (termination) on the blanking line? Try lifting a leg to remove it from circuit, see how it goes then. That resistor might be pulling the blanking voltage too low, and I'm not sure it is needed.

EDIT: Attached schematic

[abstract3000]

I'm guessing that there is a problem with the blanking voltage. When you use the OSD it overlays the image, showing correct colors, because the OSD is sending the right blanking voltage.

See that 75R resistor (termination) on the blanking line? Try lifting a leg to remove it from circuit, see how it goes then. That resistor might be pulling the blanking voltage too low, and I'm not sure it is needed.

EDIT: Attached schematic

You might be correct in your assumptions that the 75 Ohm Resistor is not needed, I pulled the leg from Ground and the Signal Still comes through to the TV, though my observations are as follows,

The colors are still off in some sort of fashion, The OSD menu is still damn near impossible to see, so I had to push the resistor end back to ground to see and play with the menu. I found the "Favorite Color" option is the only Picture Setting that had any noticeable effect on the picture, Contrast brightness etc. all the way up or all the way down made no visible difference. Though again when i touched the resistor to ground the screen went Black & white as pictured above where I have to set the Blanking pot to see it. Though disconnected the pot no longer works and the screen is visible.

So I'm not an electronics Expert far from actually, but hear me out. That Trim Pot that's installed grounds out to a 75 Ohm resistor so Ground Terminal -> Resistor -> Ground. My understanding is everything has to have a current and a Ground, so when I pulled up the leg, we left the Pot there without a ground rendering it useless. So what if we just take the Trim Pot Ground terminal straight to Ground bypassing the resistor all together? Then I assume the trim pot would work again with the blanking. though it might just put us back in the same boat. Scratch that, Just tried that And all I get is a Black screen as seen in previous photos, and adjust too far it turns the TV off, and nothing useful. Though without the ability to adjust the blanking Signal it should be noted that I no longer have the ability to use the S-Video Input.

I also might just be thinking about this in a very stupid fashion, and you can slap some sense into me if that's the case but man I tell you those Whites are Bright!!!! Like "OverPowering" Bright, like it might be so damn bright its overriding the darker colors... The Ultimarc Board is what's feeding the power through, wondering if there is a way to "Adjust" the amount of voltage it's feeding to try and potentially bring it down :/ though what I'm thinking might be way off...

So it appears Grey is the Magic color (Adam Duritz would be so proud)


I'm going to study the manual a bit more and test this menu options out but i have it available to begin adjusting and can see it

As for the Color Issues, I grabbed the Paperboy rom and loaded it up again and here we are in the same place :/ though the windows menus and Icons are easier to read...


I also noticed this is a very strange color difference on Bubble bobble, so I can't really say where the issue is with what color....

[Zebidee]

You might be correct in your assumptions that the 75 Ohm Resistor is not needed, I pulled the leg from Ground and the Signal Still comes through to the TV, though my observations are as follows,

...

So I'm not an electronics Expert far from actually, but hear me out. That Trim Pot that's installed grounds out to a 75 Ohm resistor so Ground Terminal -> Resistor -> Ground. My understanding is everything has to have a current and a Ground, so when I pulled up the leg, we left the Pot there without a ground rendering it useless. So what if we just take the Trim Pot Ground terminal straight to Ground bypassing the resistor all together? Then I assume the trim pot would work again with the blanking. though it might just put us back in the same boat. Scratch that, Just tried that And all I get is a Black screen as seen in previous photos, and adjust too far it turns the TV off, and nothing useful. Though without the ability to adjust the blanking Signal it should be noted that I no longer have the ability to use the S-Video Input.

The pot has three legs, but looks like you are only using two. Connect the third leg, through that 75R resistor, to GND (may be easier to just use a new resistor). This should give you proper control over the blanking voltage level.

I've re-drawn the schematic and attached it below. The 75R should be attached to the "wiper", not to the output.

I also might just be thinking about this in a very stupid fashion, and you can slap some sense into me if that's the case but man I tell you those Whites are Bright!!!! Like "OverPowering" Bright, like it might be so damn bright its overriding the darker colors... The Ultimarc Board is what's feeding the power through, wondering if there is a way to "Adjust" the amount of voltage it's feeding to try and potentially bring it down :/ though what I'm thinking might be way off...

Once you get this mod's blanking sorted out, if that ever happens, maybe try it without the video amp.

Otherwise, it is easy enough to make a gang of three RGB pots. 200R pots should be sufficient, and only two legs connected this time. Below is a gang of 3 x 500R pots I made for a JAMMA PCB supergun setup + GreenAntz to component-modded TV.

The pots are Bourne cermet multiturn pots - much more precise and durable than the typical cheap carbon film ones. Not that much more expensive, and I can re-use the RGB gang with other projects.


 

For more on that project:
http://forum.arcadecontrols.com/index.php/topic,165142.msg1741501.html#msg1741501

[PL1]

To be clear, are 5v and the 75 Ohm resistor both connected to the wiper or is 5v not connected to the wiper?


Scott

[Zebidee]

To be clear, are 5v and the 75 Ohm resistor both connected to the wiper or is 5v not connected to the wiper?


Scott

Scott...

What are you asking exactly? You should not connecting 5v + 75R -> GND to same pin.
There are three pins, so I'm suggesting:

1) 5v input
2) GND + resistor
3) blanking output -> jungle

As usual for pots: three pins, three separate connections.

EDIT: which means like the third schematic pic you showed Scott.

After a sleep, I realise that I didn't erase the "U-turn arrow" pot symbol on the right side - I was treating that as just a pot symbol, rather than an actual connection, but now I get that it was confusing you to thinking the 5v was meant to be attached to both the input and wiper pins (though that seems crazy). So like this:





I don't think it really matters which way around the blanking and GND pins are, or what numbers you assign to which pin. So long as they are connected as described. I'm also not sure exactly how abstract3000's pot is physically wired WRT the "wiper" as the pics don't show that. The only pic he provided of the actual circuit board is from top and very low-res.

PL1: which pin do you call the "wiper"? The middle one? Does it matter (the pot works the same, just the high/low is swapped)? I don't know which pins are physically connected, but the pencil schematic only shows two pins connected.

So, with only two pins connected, the pot is just acting as a variable in-series resistor, rather than a voltage divider. I guess the pot would vary the blanking voltage a little, but not very much (and not very high). Further, the original schematic showing 75R resistor at the pot's blanking output would pull the voltage very low anyway.

So instead, attach the 75R "termination" resistor to the pot's GND pin (the one not currently connected). You don't need signal "termination" on this blanking anyway, it isn't even a video signal as such. Furthermore, I also think that re-assigning the resistor to the pot's GND connection is a good idea to protect somebody from completely shorting the PSU's 5v rail to GND through the pot. If anything, maybe a higher value resistor (say 1K) could be considered.

[PL1]

After a sleep, I realise that I didn't erase the "U-turn arrow" pot symbol on the right side - I was treating that as just a pot symbol, rather than an actual connection, but now I get that it was confusing you to thinking the 5v was meant to be attached to both the input and wiper pins (though that seems crazy).

I figured that the "crazy" connection of 5v to both the center pin and outer pin wasn't right, but you might have intended to use the pot as a variable resistor with 5v connected to the center pin and the 75 Ohm resistor providing a minimum resistance between 5v and ground.  Variable resistor + resistor = a fake pot with a non-zero minimum resistance -- the technique the KADE guys used for Atari 2600 paddles on miniConsole+.

I don't think it really matters which way around the blanking and GND pins are, or what numbers you assign to which pin. So long as they are connected as described. I'm also not sure exactly how abstract3000's pot is physically wired WRT the "wiper" as the pics don't show that. The only pic he provided of the actual circuit board is from top and very low-res.

PL1: which pin do you call the "wiper"? The middle one? Does it matter (the pot works the same, just the high/low is swapped)? I don't know which pins are physically connected, but the pencil schematic only shows two pins connected.

Wiper is always the center (#2) pin.  On a schematic, this connection point is indicated by the arrow.  As you mention, the connections to the outer pins (#1 and #3) on a pot can be swapped and it just reverses the direction you turn the knob to increase/decrease the voltage/resistance at the wiper.  Swapping connections to pins 1 and 3 does not change the resistance or voltage drop of the resistive element between them, it just changes the wiper voltage when the contact is at the blue line from 4.5v (5v on pin 1, ground on pin 3) to 0.5v. (ground on pin 1, 5v on pin 3)

The wiki has a more detailed theory of operation and this exploded view for anyone wanting to learn more about how pots work.
http://wiki.arcadecontrols.com/index.php/Analog_Encoders#How_a_potentiometer_provides_the_voltage_for_an_analog_encoder_to_measure




Scott

[abstract3000]

Thanks for the Feedback, much appreciated.

The only pic he provided of the actual circuit board is from top and very low-res.

I apologize for that, and @Scott I really couldn't tell you the correct way the Schematic is supposed to look but I got some Hi-Res photos and have described where everything is going so hopefully that will clear up the confusion

So This is the Top side, I know it's painfully obvious what I marked but just indicating what I was looking at. And the orientation of the Trim Pot


So here is the Explanation I should say for lack of better words.
(whoops I messed up, I fixed the image after closer inspection)
The Red Circles indicate the Pins on the Trim Pot

Left pin Went to the 75Ohm Resistor (Now removed) Then to Ground
Center Pin connected to Right Pin
The Right Pin Goes to a Diode, then Back to the TV Motherboard I assume to Blanking.

Here is a Closeup of the Diode


In other news I went through and tried every setting in the service menu, managed to fix the Vertical Overscan, but nothing with RGB Cuttoffs or any other setting made any sort of noticeable difference besides "tone" which did what it sounds like yet not a solution. I just saw The previous posts, but didn't get a chance to read closely through them as I wanted to get you guys the Exact information of what was happening. I will look over those now.

@Zebidee, I like your solution for adjusting the voltage, those pots you used could you provide me a link to the specific ones that would be needed on Mouser or DigiKey? and those solderless connectors I really like, what are they called for searching purposes?

[Zebidee]

I figured that the "crazy" connection of 5v to both the center pin and outer pin wasn't right, but you might have intended to use the pot as a variable resistor with 5v connected to the center pin and the 75 Ohm resistor providing a minimum resistance between 5v and ground.  Variable resistor + resistor = a fake pot with a non-zero minimum resistance -- the technique the KADE guys used for Atari 2600 paddles on miniConsole+.

Yeah, but no. There are many variants, but I'm just suggesting to make a simple voltage divider with the pot.

Putting a resistor on the GND pin, as a minimum load, is sensible. Not trying to be fancy here.

PL1: If you put a resistor between the wiper and the GND you change the behaviour of the pot (in terms of voltage response) to approximate a log curve (or is it anti-log? always confusing), rather than linear. This might well be appropriate and feel "natural" for the Atari 2600 paddles, but beyond the scope of what I'm talking about for this mod.

It just so happens that I covered this in depth (with pics, graphs & even a re-usable spreadsheet), a few months ago, on another post about decent volume controls:
http://forum.arcadecontrols.com/index.php/topic,167662.msg1765413.html#msg1765413

The wiki has a more detailed theory of operation and this exploded view for anyone wanting to learn more about how pots work.
http://wiki.arcadecontrols.com/index.php/Analog_Encoders#How_a_potentiometer_provides_the_voltage_for_an_analog_encoder_to_measure



Scott

Thanks for that link, it is a great resource. 

PS: I had just written all this then saw a new post from abstract3000, so will just submit this then read.

[PL1]

PL1: If you put a resistor between the wiper and the GND you change the behaviour of the pot (in terms of voltage response) to approximate a log curve (or is it anti-log? always confusing), rather than linear. This might well be appropriate and feel "natural" for the Atari 2600 paddles, but beyond the scope of what I'm talking about for this mod.

Different application and different connections, but it's still a combination of a pot and a fixed resistor.   

The Atari paddles have pots wired as variable resistors. (2 wires)

Arduino A/D inputs can only read voltage (not resistance) so the extra resistor was added externally so the original paddle wiring wouldn't need to be modified to make it work as a voltage divider. (pot)

 


Scott

[Zebidee]

Thanks for the Feedback, much appreciated.
...
So This is the Top side, I know it's painfully obvious what I marked but just indicating what I was looking at. And the orientation of the Trim Pot


So here is the Explanation I should say for lack of better words.

The Red Circles indicate the Pins on the Trim Pot
Left pin goes to a Jumper (No Jumping pin disconnected) The Other side of that Jumper goes to Pin 14 (H-Synch)
Center Pin Went to the 75Ohm Resistor (Now removed) Then to Ground

Oh wow this is very revealing.  A few points that were not apparent from the schematic.

1) WTF/What/why is that pot/jumper connected to VGA pin 14? That is the vertical sync pin.

It should be connected to a 5v source. On some VGA cards 5v is available from VGA pin 9. There are also 5v sources available from the TV.

2) Looks like the 75R resistor was attached to the wiper+GND originally, rather than the output+GND as the schematic says. It is still pulling the voltage very low. Try replacing it with a larger value like 1K.

The Right Pin Goes to a Diode, then Back to the TV Motherboard I assum to Blanking.

Here is a Closeup of the Diode

3) That diode is completely new news to me. It will pull the blanking voltage down even further (~0.7v). You probably don't need it there anyway, remove it [edit: easier to just bypass it with a short wire], it may be causing issues.

In other news I went through and tried every setting in the service menu, managed to fix the Vertical Overscan, but nothing with RGB Cuttoffs or any other setting made any sort of noticeable difference besides "tone" which did what it sounds like yet not a solution. I just saw The previous posts, but didn't get a chance to read closely through them as I wanted to get you guys the Exact information of what was happening. I will look over those now.

I'm not worrying about this much for now. Get the blanking right first.

@Zebidee, I like your solution for adjusting the voltage, those pots you used could you provide me a link to the specific ones that would be needed on Mouser or DigiKey? and those solderless connectors I really like, what are they called for searching purposes?

The specific pots I used are Bourne 500R, but as I mentioned above 200R would be sufficient.
https://www.bourns.com/products/trimpot-trimming-potentiometers/trimpot-trimming-potentiometers-multiturn/product/3296

To buy, search for:
"bourns cermet multi turn potentiometer 3296W-1-201LF" for the 200R variety, or;
"bourns cermet multi turn potentiometer 3296W-1-501LF" for the 500R.

I see that they are available on both Mouser and Digikey, as well as other online marketplaces.

[Zebidee]

PL1: If you put a resistor between the wiper and the GND you change the behaviour of the pot (in terms of voltage response) to approximate a log curve (or is it anti-log? always confusing), rather than linear. This might well be appropriate and feel "natural" for the Atari 2600 paddles, but beyond the scope of what I'm talking about for this mod.

Different application and different connections, but it's still a combination of a pot and a fixed resistor.   

The Atari paddles have pots wired as variable resistors. (2 wires)

Arduino A/D inputs can only read voltage (not resistance) so the extra resistor was added externally so the original paddle wiring wouldn't need to be modified to make it work as a voltage divider. (pot)

 


Scott

Right. We/You are measuring voltage not resistance. Your example has only 2 pins connected, thus it is just a single variable resistor. Even so, it works as a voltage divider because there is another resistor between output and GND.[edited, corrected, condensed]

As you say, "Different application and different connections, but it's still a combination of a pot and a fixed resistor.   ". Very true.

[abstract3000]

Oh wow this is very revealing.  A few points that were not apparent from the schematic.

1) WTF/What/why is that pot/jumper connected to VGA pin 14? That is the vertical sync pin.

It should be connected to a 5v source. On some VGA cards 5v is available from VGA pin 9. There are also 5v sources available from the TV.

Sorry please take another look at the photo, I updated my mistake. That Jumper was left on the board and I was told it was for testing, and not to bother with it. My guess is the other side is related to Pin 13 and it was when he was playing around with the Synch Signals. 

2) Looks like the 75R resistor was attached to the wiper+GND originally, rather than the output+GND as the schematic says. It is still pulling the voltage very low. Try replacing it with a larger value like 1K.

So it would have been on the Left pin not the Wiper, but I can try putting a 1K resistor in it's place instead to see what that does. The Wiper (center from my understanding is connected to the right pin)

3) That diode is completely new news to me. It will pull the blanking voltage down even further (~0.7v). You probably don't need it there anyway, remove it [edit: easier to just bypass it with a short wire], it may be causing issues.

Ok so I Can look at a way to pypass the diode completely, I was thinking the 5V may be coming from the TV to the right pin through the diode to ensure it doesn't somehow find its way back as precautionary, but I could have that completely wrong.

[Zebidee]

Oh wow this is very revealing.  A few points that were not apparent from the schematic.

1) WTF/What/why is that pot/jumper connected to VGA pin 14? That is the vertical sync pin.

It should be connected to a 5v source. On some VGA cards 5v is available from VGA pin 9. There are also 5v sources available from the TV.

Sorry please take another look at the photo, I updated my mistake. That Jumper was left on the board and I was told it was for testing, and not to bother with it. My guess is the other side is related to Pin 13 and it was when he was playing around with the Synch Signals. 

I think I understand. So VGA 13/14 is not connected or related to the blanking? Good, leave it for now.

2) Looks like the 75R resistor was attached to the wiper+GND originally, rather than the output+GND as the schematic says. It is still pulling the voltage very low. Try replacing it with a larger value like 1K.

So it would have been on the Left pin not the Wiper, but I can try putting a 1K resistor in it's place instead to see what that does. The Wiper (center from my understanding is connected to the right pin)

I'm confess to being completely confused now. As we just discussed with PL1, there are many ways to wire a pot for different outcomes. Your schematic shows the 75R+GND connected to output. This will pull your blanking voltage quite low.

One terminal is input, one is output, one is resistor->GND. Your pot is 1K? So I suggest a 1K resistor, 75R is too low.

If you had the output wired to the wiper then it would be like this:

1) 5v input
2) blanking output
3) resistor -> GND

3) That diode is completely new news to me. It will pull the blanking voltage down even further (~0.7v). You probably don't need it there anyway, remove it [edit: easier to just bypass it with a short wire], it may be causing issues.

Ok so I Can look at a way to pypass the diode completely, I was thinking the 5V may be coming from the TV to the right pin through the diode to ensure it doesn't somehow find its way back as precautionary, but I could have that completely wrong.

So.... the blanking 5v voltage is sourced from the TV somehow? Guess that's all I need to know for now.

The diode will only pass current one-way, but in doing so it drops voltage ~0.7v. I suspect that your blanking voltage doesn't get very high anyway, so this will just make it worse.

For now, don't remove the diode. Just solder a short wire (such as an offcut component leg) over the diode's terminals to bypass the diode completely. Much easier than removing the diode, and easier to undo later if you want to.

What we want to see is a large range of voltage (~0 to ~5v) at the pot's output. You can do this measurement yourself, at the pot's high/low limits, with a multimeter.

[abstract3000]

Alright then, So i get a bunch of resistors in on Tuesday, I will get the 1K resistor put back into place where the 75 Ohm one was, and solder in a wire skipping the Diode. This will also give me a couple days to try and trace the wire from the pot back to the board and see If I Can identify what exactly is happening there, and I will use the multimeter to try and get a voltage reading off of the pot.

[abstract3000]

well not the best of news, and feeling like an idiot at the moment. So I tested the 1k Resistor it did absolutely nothing, I was concerned I was not seeing anything so i used the multimeter to check the resistor and even tried a 2nd one and no dice. I then tested the 75k and it went back to how it was working previously...

So I leave the garage, and come back several hours later, Monitor is black I assumed I turned it off. What i didn't realize was that the PC shut down cutting the 12v to the Ultimarc adapter causing the screen to go black, and I went to desolder the diode and try without it....yeah so things got ugly, I see a flash on the screen and here the TV power off. At this moment I here a slight clicking sound when plugged in but no screen. My guess is with the mod on the diode I shorted out where ever it came from the board either destroying a capacitor or messing something else up in the power supply this is beyond my capabilities I'm afraid. Thank god I got the Panasonic...

[Zebidee]

RIP Sharp. Sorry for your loss. Seems like the RGB-OSD mod was causing too much pain anyway, and might never have worked well.

That clicking is likely the protection circuit cutting off power. Something is out-of-whack in the power circuit. This is probably good news as easier to repair than a shorted jungle IC. You might be able to find a failed part by backtracking the 5v supply line. If lucky maybe just a fusable resistor or power diode, easy to check these with a DMM. Sometimes the damaged part/s is obvious, sometimes not. Sometimes you can smell the "magic smoke". Maybe your mate can help.

[abstract3000]

You know my friend could certainly Troubleshoot and identify and even resolve the issue with parts being available, but at this point he appears to want nothing more to do with the project. I completely understand as he has dispensed enough time and effort to get me this far, wanting to move on to other paying jobs and leave this one behind, and that is completely understandable. To be honest I'm surprised you have stuck this out as long as you did Zebidee, and for that I am extremely grateful.

Many posts ago, I stated I needed to really consider the costs and initiative to this project, and where to draw a line. So far its just putting more money and effort on trying to find that "fix", I feel like such a complete idiot and should have physically checked the TV was unplugged, but its just another reminder to move on, with several red flags along the way.

I can't help but feel deflated and a bit of a hit to my ego, as I truly am fascinated with Electronics, never thought i knew to terribly much, but never thought i would make this blunderus of a move out of carelessness. Though wakabavideo's adapter is in Los Angeles Right now on it's way and I have a direct swap over. So as hard as it is to just let it go, I feel it's time to put this to bed, use the other Monitor with the plug and play option (verified on a cheap Chinese adapter) and stop dumping money into it.

If the CRT will fit into the Panasonic Housing by chance (I have little faith that it will (but if so) I will hold onto it a bit longer as it will be easier to store and maybe down the road i can have another look. Though if not I just can't have that CRT in the massive temporary wooden frame sitting around exposed, it's hard to move and not well protected, and letting it take all that space up for an inevitable event the neck gets broken by accident seems wasteful, but I will see how it goes this weakened.

[Zebidee]

Argh. Nevermind. To screwup is human. Ego takes a beating, but it will recover. Learn from mistakes. Time to move on. At least you didn't fry.

Offer the Sharp for free online, here and elsewhere.

If you measure the CRT's yoke impedance, you can attract interest from people wanting to do a tube swap.

Just set DMM to ohms (set to low ohms if not autoranging) and get the readings across both horizontal and vertical yoke coils. You can access the large pins near neck. Alternatively, you can unplug the 4-pin chassis connector and put your probe leads into the female pins. This ensures a good connection.

Good idea to zero the meter's probes first (hold probe leads together for a few seconds, let readings approach zero/settle), as the horz coil readings in particular can be quite low in some yokes, and a few ohms difference can be important.

When done, stick down some paper/masking tape somewhere convenient, and write down the H & V coil readings on it.

You'll also need to count the number of pins around back of neck (you may need to remove the neckboard to see clearly).