TV RGB mod thread

[vol.2]

Sorry mate, the datasheet suggests that the RGB input is designed only for digital OSD input, not analog RGB.

Damn it. Does that mean I can't RGB mod it at all? Or I just have to give up the OSD?

Almost certainly. TVs of that era set all the inputs in non-volatile memory in the factory. If it wasn't designed to have an RGB input, that input will not be enabled in the settings. Those settings are non user-definable, they are set over an i2C bus with a ROM chip.

It could be possible to dump the existing ROM and reverse engineer the settings by decompiling the ROM and using the jungle IC's datasheet (assuming it's complete enough), but that would be an epic amount of work and it is both beyond the scope of this thread and outside of the wheelhouse of most of us here. You would need someone with a decent level of computer engineering skills and extensive experience with relatively outdated i2C programming

Of course, that's also assuming that there is even some unused RGB input on the jungle IC in the first place.

[lazyelm]

Is the voltage steady at the injection point around 1.1 / 1.2v, and is the voltage from your external source 1.9v steady? If so, this would indicate you have DC bias so you will want to add a coupling capacitor in between your external source and the injection point. 1uf bipolar will probably work, but if you have polar ones point the positive end towards the higher voltage which seems to be external source at the moment. It probably won't hurt to have it anyway while you are testing. If there is a bias offset then I think it could also cause your display to go black when connecting.

The voltage at the injection point is ~1.2v pretty steady yes. External source is .7v. - when i inject, the resulting voltage is ~1.9-2v

I have tried direct but also tried a .1uf cap->75ohm resistor->injection point - same results both ways. Would a 1uf vs a .1uf cap make a difference there? I could try that as well

[Delphius]

Is the voltage steady at the injection point around 1.1 / 1.2v, and is the voltage from your external source 1.9v steady? If so, this would indicate you have DC bias so you will want to add a coupling capacitor in between your external source and the injection point. 1uf bipolar will probably work, but if you have polar ones point the positive end towards the higher voltage which seems to be external source at the moment. It probably won't hurt to have it anyway while you are testing. If there is a bias offset then I think it could also cause your display to go black when connecting.

The voltage at the injection point is ~1.2v pretty steady yes. External source is .7v. - when i inject, the resulting voltage is ~1.9-2v

I have tried direct but also tried a .1uf cap->75ohm resistor->injection point - same results both ways. Would a 1uf vs a .1uf cap make a difference there? I could try that as well

0.1uf is probably fine, or at least you would be getting an image if the biasing was an issue. I would still probably keep it on there while you are testing because it shouldn't hurt anything and if you are getting a steady ~1.2v at the injection point then this indicates it has bias. I would expect at this point that you would see at least some interaction with your external RGB though. Just to clarify, is your OSD affected when you are experimenting?

[lazyelm]

Is the voltage steady at the injection point around 1.1 / 1.2v, and is the voltage from your external source 1.9v steady? If so, this would indicate you have DC bias so you will want to add a coupling capacitor in between your external source and the injection point. 1uf bipolar will probably work, but if you have polar ones point the positive end towards the higher voltage which seems to be external source at the moment. It probably won't hurt to have it anyway while you are testing. If there is a bias offset then I think it could also cause your display to go black when connecting.

The voltage at the injection point is ~1.2v pretty steady yes. External source is .7v. - when i inject, the resulting voltage is ~1.9-2v

I have tried direct but also tried a .1uf cap->75ohm resistor->injection point - same results both ways. Would a 1uf vs a .1uf cap make a difference there? I could try that as well

0.1uf is probably fine, or at least you would be getting an image if the biasing was an issue. I would still probably keep it on there while you are testing because it shouldn't hurt anything and if you are getting a steady ~1.2v at the injection point then this indicates it has bias. I would expect at this point that you would see at least some interaction with your external RGB though. Just to clarify, is your OSD affected when you are experimenting?

Yeah changing resistor values there does impact color Intensity of the green osd text. But I get no image. Black screen. If I hook up another source to the composite image I can impact intensity of rgb as well with the pots but I’m not seeing my injected image. Could it be because the voltage of the bias is 1.2v which is much higher than my injected source?


FWIW. Also tried injecting at the backboard at the base leg on the transistors but the voltage is 12v so I think I would need to ramp up the voltage for that to work on my source


In all these scenarios though I’m getting a black screen I think the blanking and or the source shutoff when it’s not detecting an input from the av inputs is blacking the screen.

[Delphius]

The voltage at the injection point is ~1.2v pretty steady yes. External source is .7v. - when i inject, the resulting voltage is ~1.9-2v

I have tried direct but also tried a .1uf cap->75ohm resistor->injection point - same results both ways. Would a 1uf vs a .1uf cap make a difference there? I could try that as well

0.1uf is probably fine, or at least you would be getting an image if the biasing was an issue. I would still probably keep it on there while you are testing because it shouldn't hurt anything and if you are getting a steady ~1.2v at the injection point then this indicates it has bias. I would expect at this point that you would see at least some interaction with your external RGB though. Just to clarify, is your OSD affected when you are experimenting?

Yeah changing resistor values there does impact color Intensity of the green osd text. But I get no image. Black screen. If I hook up another source to the composite image I can impact intensity of rgb as well with the pots but I’m not seeing my injected image. Could it be because the voltage of the bias is 1.2v which is much higher than my injected source?


FWIW. Also tried injecting at the backboard at the base leg on the transistors but the voltage is 12v so I think I would need to ramp up the voltage for that to work on my source


In all these scenarios though I’m getting a black screen I think the blanking and or the source shutoff when it’s not detecting an input from the av inputs is blacking the screen.

I am sort of out of ideas, but I can try to find some time to study the schematic a little more. I think it is a good sign that the resistor network is having an affect on the OSD when you are injecting, but surprised you aren't seeing any external RGB if you are seeing the OSD. Usually I would expect more interaction, even if it was causing some interference from active buffering or bias on the line.

For injecting into the neck board like that I think there are other mod kits that you can purchase to help. I would be a bit careful injecting there if there is a 12v bias, but especially use the coupling capacitor while experimenting to be a bit safer.

[lazyelm]

0.1uf is probably fine, or at least you would be getting an image if the biasing was an issue. I would still probably keep it on there while you are testing because it shouldn't hurt anything and if you are getting a steady ~1.2v at the injection point then this indicates it has bias. I would expect at this point that you would see at least some interaction with your external RGB though. Just to clarify, is your OSD affected when you are experimenting?

Yeah changing resistor values there does impact color Intensity of the green osd text. But I get no image. Black screen. If I hook up another source to the composite image I can impact intensity of rgb as well with the pots but I’m not seeing my injected image. Could it be because the voltage of the bias is 1.2v which is much higher than my injected source?


FWIW. Also tried injecting at the backboard at the base leg on the transistors but the voltage is 12v so I think I would need to ramp up the voltage for that to work on my source


In all these scenarios though I’m getting a black screen I think the blanking and or the source shutoff when it’s not detecting an input from the av inputs is blacking the screen.

I am sort of out of ideas, but I can try to find some time to study the schematic a little more. I think it is a good sign that the resistor network is having an affect on the OSD when you are injecting, but surprised you aren't seeing any external RGB if you are seeing the OSD. Usually I would expect more interaction, even if it was causing some interference from active buffering or bias on the line.

For injecting into the neck board like that I think there are other mod kits that you can purchase to help. I would be a bit careful injecting there if there is a 12v bias, but especially use the coupling capacitor while experimenting to be a bit safer.

Thanks man I appreciate it. I’m at a loss too

[SkiDragon]

I've had a Magnavox TS2582 C101 for a couple of years and finally decided to see if I could mod it for RGB. From factory it only has RF, composite, and S-Video. The chassis according to the label on the back of the case is 25R505-00AA
I cannot find the regular owner’s manual, and I do not own a remote for it.
However, I found the service manual for this range of chassis models here:
https://ia801308.us.archive.org/32/item ... 207538.pdf
The service manual does not list the specific type of jungle IC, so I had to open it up.



I removed the main PCB and had to desolder and pry open an EMI shield to find the jungle chip, a TDA8369. It's hard to find much documentation on this chip, but I found a short spec sheet:
https://datasheet4u.com/datasheet/ETC/TDA8369-770219
I hesitate to call it a full datasheet, and it’s a poor quality scan with black blotches making some areas hard to see. However, on the last page I unfortunately did spot the line under Features: "RGB-control circuit with digital OSD input with intensity control"
So, SOL, right? But I had a lot of time to waste today and I started looking up mods for other jungle chips of the TDA836X and TDA837X range (using crtdatabase.com) and it seemed like all the rest had analog inputs. So I decided to poke around some more.
The chip has RGB inputs from a separate OSD chip, but I did not see any coupling capacitors on the schematic or when physically inspecting the board. (There were capacitors to ground)



So, not trusting the crappy datasheet, I decided to hook up my oscilloscope to the Green line between the microprocessor and the TDA8369 Jungle chip.



Luckily, the scope did not blow up which I think verifies that it is not 'hot chassis'.
Here is what the signal looks like when I am adjusting the volume (and thus triggering the OSD)



Immediately I thought this looked digital and felt disappointed, but I realized an analog signal here would just basically show Green as either On or Off depending on the position in the scan line… so is this analog? I am still not entirely sure of myself. Can anyone chime in?

As an aside, in the service manual there was a short blurb saying "Caution: The RS Chassis incorporates a "HOT" ground system. Always use a separate isolation transformer when applying power to the ex-posed chassis."
(I wasn’t sure if that applied to my specific model, and it would be odd I thought for a TV from 1995 with S-Video to be made like this, but it's possible. However mine is not "RS" and I am not sure what chassis that statement is even referring to. Studying the schematic more closely and it looked like the AC was in fact isolated through a transformer. There was continuity between the existing RCA jack shields and various grounds on the main board, including pin 13 GND on the actual jungle chip. I also did a diode test on between the two sides of the AC plug and the RCA jack shield, both prongs and both polarities and it was all reading as no connection. So I risked it, and so far seems to be fine.)

TLDR:
I am trying to figure out if that signal is analog and if/how I should proceed to attempting a RGB mod, preferably muxed but switched if need be.

[Delphius]

I've had a Magnavox TS2582 C101 for a couple of years and finally decided to see if I could mod it for RGB. From factory it only has RF, composite, and S-Video. The chassis according to the label on the back of the case is 25R505-00AA
I cannot find the regular owner’s manual, and I do not own a remote for it.
However, I found the service manual for this range of chassis models here:
https://ia801308.us.archive.org/32/item ... 207538.pdf
The service manual does not list the specific type of jungle IC, so I had to open it up.

https://ibb.co/0psWGZfw

I removed the main PCB and had to desolder and pry open an EMI shield to find the jungle chip, a TDA8369. It's hard to find much documentation on this chip, but I found a short spec sheet:
https://datasheet4u.com/datasheet/ETC/TDA8369-770219
I hesitate to call it a full datasheet, and it’s a poor quality scan with black blotches making some areas hard to see. However, on the last page I unfortunately did spot the line under Features: "RGB-control circuit with digital OSD input with intensity control"
So, SOL, right? But I had a lot of time to waste today and I started looking up mods for other jungle chips of the TDA836X and TDA837X range (using crtdatabase.com) and it seemed like all the rest had analog inputs. So I decided to poke around some more.
The chip has RGB inputs from a separate OSD chip, but I did not see any coupling capacitors on the schematic or when physically inspecting the board. (There were capacitors to ground)

https://ibb.co/zHWLxjHd

So, not trusting the crappy datasheet, I decided to hook up my oscilloscope to the Green line between the microprocessor and the TDA8369 Jungle chip.

https://ibb.co/Kp1VFndC

Luckily, the scope did not blow up which I think verifies that it is not 'hot chassis'.
Here is what the signal looks like when I am adjusting the volume (and thus triggering the OSD)

https://ibb.co/wNhfbKgp

Immediately I thought this looked digital and felt disappointed, but I realized an analog signal here would just basically show Green as either On or Off depending on the position in the scan line… so is this analog? I am still not entirely sure of myself. Can anyone chime in?

As an aside, in the service manual there was a short blurb saying "Caution: The RS Chassis incorporates a "HOT" ground system. Always use a separate isolation transformer when applying power to the ex-posed chassis."
(I wasn’t sure if that applied to my specific model, and it would be odd I thought for a TV from 1995 with S-Video to be made like this, but it's possible. However mine is not "RS" and I am not sure what chassis that statement is even referring to. Studying the schematic more closely and it looked like the AC was in fact isolated through a transformer. There was continuity between the existing RCA jack shields and various grounds on the main board, including pin 13 GND on the actual jungle chip. I also did a diode test on between the two sides of the AC plug and the RCA jack shield, both prongs and both polarities and it was all reading as no connection. So I risked it, and so far seems to be fine.)

TLDR:
I am trying to figure out if that signal is analog and if/how I should proceed to attempting a RGB mod, preferably muxed but switched if need be.

I don't have any experience scoping out digital OSD signals, but it would seem to me that if there are separate channels for each RGB input signals then they are probably analog. I could be totally wrong, but I don't understand the logic behind separating a digital signal for each RGB channel. I would assume they would combine those into one digital signal and encode through a clock. It seems to be 5Vpp, so maybe it is worth a try as long as you keep it in that voltage range.

[SkiDragon]

Well, unfortunately it looks like my TV uses digital RGB. I decided to makeshift an RGB cable for my PS1 to test it out and these were the results:

https://imgur.com/a/ad8Y8de

It's a shame it didn't work, but nonetheless, I enjoyed the experience, and I think I learned a few useful things. Maybe someday I'll own a CRT that can be RGB modded; then I can put the knowledge I gained to use, but until then I'll be happy with my 9-inch CRT and its composite inputs.

I know this post is a year and a half old, but if this person's jungle chip expects a digital signal from the OSD chip, and they injected an analog signal into this input, why are they getting an image that looks close to correct? Normally an analog output into a digital input would not work in the slightest. Like, imagine putting composite video into your HDMI port, totally bonkers.

So, what is meant by 'digital' here. Do people just mean 'discretized', or something like that?