DIY RGB to NTSC Encoder with Component Video Out?

Anyone got a schematic for a DIY RGB to NTSC Encoder with Component Video Out? I'm looking for one for a project i'm working on.

It's called a datasheet.

Yah, but I don't even know what chip to start with. I could totally start with the AD725 and follow an older tutorial on endgadget. I wouldn't know where to add in the component output part.

edit: looks like the datasheet on the AD725 says composite output only.

Just buy one. Seriously. If you want component, just buy one.

but then i can't integrate the circuit into a board i'm laying out to etch.

Murray said:

Just buy one. Seriously. If you want component, just buy one.

Click to expand...

AD725 I believe is only composite and s-video.

If you want to save money just ask Jeff Kurtz or James Rowan for a component only version.

FA-MAS said:

but then i can't integrate the circuit into a board i'm laying out to etch.

Click to expand...

What are you building?

This guy has a page on YUV to RGB video, and at the bottom, the last link is to a schematic for RGB to YUV. (For using old RGB games and computers on component video TVs)

I had intended to build one, but never got around to it. I don't know if it will work with arcade RGB without additional stuff to convert the scan rate.

It's actually a pretty elegant solution, since it's just using op-amps and precise resistor values to do the math involved in converting the RGB to YPbPr

That circuit takes standard 1Vpp into 75 ohm RGB video not crazy arcade levels, so you should add variable input impedance, buffer that, then use a voltage divider to get nice output. I don't think people do that on their Superguns though.

If you use the clamping circuit there you will be limited to whatever the LM1881 can detect, ie not 24/31 kHz video. I hear arcade video isn't typically AC coupled so you can skip the clamping, just add composite sync directly to Y.

Personally I think the most elegant solution would be to make a $1 transcoder from discrete differential amps using a few GP transistors but it's a lot more difficult to design.

Edit: I'm sort of wrong, actually the LM1881 will be able to detect the back porch of, and thus clamp video up to 150 kHz (so well over 1080p!), but you need to set the RC constant... So an auto-detecting chip such the LMH1980 would work for 24/31 kHz games.

Sorry for the lack of replies. So I had planned to try my hand at building a fully integrated supergun, similar to the vogatek. Not to save money, cause the vogatek's pretty cheap to start with. But to figure out how to etch circuit boards and superguns are pretty simple. Mainly alot of point to point connections. But I'd like to build in some of the features such as component output.

Just google for an RGB to YUV or component encoder and go from there.

You guys think this circuit would work? http://www.keohi.com/keohihdtv/interfaces/diytranscoder_johrhees.html

I don't think i'd need the V and H sync combination part though since it's already combined from the game board.

Edit:
This one might be better.

http://i18.photobucket.com/albums/b134/Segasonicfan/schems/Componentschem.png

FA-MAS said:

Edit:
This one might be better.

http://i18.photobucket.com/albums/b134/Segasonicfan/schems/Componentschem.png

Click to expand...

No offense to Zach but the S-video circuit he designed was horrible in quality. So I don't know how good that component one is. I guess only cannonball can say how good that component circuit is, that is if Zach put it in his mini supergun.

Forgot to ask, does a component circuit not need a subcarrier?

I read on his site that he didn't design the circuit, he just modified someone elses. So maybe that can account for the quality. I'll try it out.

Xian Xi said:

Just google for an RGB to YUV or component encoder and go from there.

Click to expand...

I realize this isin't the solution you're looking for but for other's this link could be helpful!

http://www.piersolar.com/board/viewtopic.php?f=3&t=420

Going to use it on my Genesis....hopefully the Megadrive scart I ordered will work on my un modded Home System too!

Hey guys, sorry to revive this thread but I'm having some problems reading the schematic for this thing. Well, not really the schematic itself, rather than the comments at the bottom.

I understand the drawing as well as the bypass capacitors on each IC.

But the next part about bypassing the initial +5v and VEE (-5) outputs using .1uf, .01uf and 100uf caps is what I don't understand. I understand the concept of the bypass, but I don't understand where to put what value resistor and why on the output (i'm assuming the +5v (VCC) directly from the PSU and the -5v from the ICL7660) But if you're bypassing at each power line for each IC anyway, I don't understand.

I also don't understand the Resistor Value substitution in the drawing. For example:

6030 = 6k8//51k

Does that mean to use a 6800 ohm or a 51000 ohm?

Here's the drawing.

Bypass means a small cap in this case the .1uf ceramic, you put one leg on the +5v line and one on the ground. This is kind of a filter.

As far as the other stuff I dont know but you can always PM Segasonicfan as he's a member here.

"//" means in parallel with, so to get 6030 ohms in series, you put a 6.8k in parallel with a 51k.

Kyuusaku said:

"//" means in parallel with, so to get 6030 ohms in series, you put a 6.8k in parallel with a 51k.

Click to expand...

I was kind of wondering about that. Seemed like there was some math to get specific resistance values. I just didn't know "//" meant in parallel. So I should wire them up like this?

Yup.

Kyuusaku said:

Yup.

Click to expand...

Thanks. I'm wondering if you might help clarify the power. I guess his notation of bypassing the initial +5v and VEE (-5) outputs using .1uf, .01uf and 100uf caps is what I don't understand.

I understand VCC (+5v) would come from the PSU, and would get filtered at every IC with a .1uf cap. This includes the ICL7660?

I understand that Ground would also come directly from the PSU and each portion of the schematic that would connect to the ground would go there.

Here's where I'm confuse. It looks like the VEE (-5) output from the ICL7660 is already bypassed by a 10uf cap. So I still bypass VEE at each IC with a .1uf cap? I don't understand where the other values came from ".1uf, .01uf and 100uf caps"

Xian Xi said:

No offense to Zach but the S-video circuit he designed was horrible in quality. So I don't know how good that component one is. I guess only cannonball can say how good that component circuit is, that is if Zach put it in his mini supergun.

Click to expand...

Xian Xi, why do you think so? Also, I'm not sure what you are referring to since I have never designed an S-Video circuit. I updated a schematic for pulling S-video from the CXA1145 which has bad S-video capability to begin with and would have little to do with the small discrete circuit added. The only other things I've done with S-video are using the AD724 and CXA1645 which were both just designs pulled from their respective datasheets.

and as mentioned, I did not design the RGB to component converter, I just updated the schematic from elm-chan over here:
http://elm-chan.org/works/yuv2rgb/report.html

His stuff is great and he's the real genius so I recommend people taking a look at his projects.

I made the schem for someone that requested it in simpler terms that also wanted to build it. This circuit works perfectly for arcade video conversion.

Thanks. I'm wondering if you might help clarify the power. I guess his notation of bypassing the initial +5v and VEE (-5) outputs using .1uf, .01uf and 100uf caps is what I don't understand.

I understand VCC (+5v) would come from the PSU, and would get filtered at every IC with a .1uf cap. This includes the ICL7660?

I understand that Ground would also come directly from the PSU and each portion of the schematic that would connect to the ground would go there.

Here's where I'm confuse. It looks like the VEE (-5) output from the ICL7660 is already bypassed by a 10uf cap. So I still bypass VEE at each IC with a .1uf cap? I don't understand where the other values came from ".1uf, .01uf and 100uf caps"

Click to expand...

The bypassing may not make or break this circuit but its good practice and should give you better video quality. the 10uf is not used just for bypassing the ICL7660 output (see the datasheet for this chip) it is a charge capacitor to produce the negative (and unregulated) voltage. the .1uf is for bypassing the opamps and should be as close to them as possible because it is still a power line and this helps make the voltage ripple less among other things. its important practice among all circuits generally to have well bypassed power especially close to the chips.

Note I have been working with more voltage inverters recently and since the ICL7660 output is unregulated I am now favoring the MAX889 as a better chip. Both will work fine but the later may give slightly better output.

I am very busy these days and am dealing with a lot of medical problems so please understand I cannot always help with these schems upon request. Posting on thewww.gamesx.com forums is even better for people that will gladly help and are more active on these forums and have the physical ability to do these things better than myself at the moment.

I'm glad to see this schem getting some use and I wish you luck with your circuit.
best wishes,

-Segasonicfan

segasonicfan said:

Xian Xi, why do you think so? Also, I'm not sure what you are referring to since I have never designed an S-Video circuit. I updated a schematic for pulling S-video from the CXA1145 which has bad S-video capability to begin with and would have little to do with the small discrete circuit added. The only other things I've done with S-video are using the AD724 and CXA1645 which were both just designs pulled from their respective datasheets.

-Segasonicfan

Click to expand...

Hey Zach,

I guess when I see your name stamped on a schematic I assume you designed the circuit. The CXA1145 has excellent S-video output. The schematic just had too many components and some of which weren't even needed and the combination produced a very over-saturated and bleeding picture. Vietlim(sp) posted a very simplified design for the circuit which does produce an amazing s-video signal from the CXA1145. Again like I said no offense was intended since I didn't know who designed the pictured circuit below.

Xian Xi said:

Forgot to ask, does a component circuit not need a subcarrier?

Click to expand...

No, why would it? No PAL/NTSC to decode.

Hewitson said:

No, why would it? No PAL/NTSC to decode.

Click to expand...

Figured it didn't, my projector always throws me off because the component option has (NTSC) next to it.