I'm not sure what the gain needs to be for the Y signal to drive the set. Here's the common mod done on the Y signal: http://www.jamma-nation-x.com/jammax/genesismods.html
The resistor on the emitter of the NPN transistor is 27 Ohms with 5v on the collector.

Yes, the signals are going to a television. A digital LCD television, specifically.

The available supply voltage in the system is single-supply 5v

The "amp" is an emitter follower. It doesn't add voltage gain, but it does add current gain, which is the ability to drive a low impedance load (presumably 75Ω).
I take it you are not creating an s-video signal, because your TV has the ability to take Y, Pr, and Pb directly.
I'll get back to the design, but I'm a little unsure about how I'll handle the DC offsets of the 3 inputs. I'll have to think about that. It's difficult (maybe impossible, without having hardware in front of me) to design a circuit to interface with a part whose datasheet is written in Japanese.

 

I have a design that's pretty much finished, except for choosing an op amp.
I looked at AD8002, because I know you have them, and they are in a thru-hole package, but they don't have adequate output swing with a 5V supply.
LTC6244 works well in simulation, but is not available in thru-hole. They are duals. The circuit requires 4 op amps, so 2 packages.
I'll keep looking.

 

How about a single quad package of this? http://cds.linear.com/docs/Datasheet/lt1253.pdf

EDIT: That one is also a CFB. Don't know what your design calls for. Does it includee a lot of circuitry to deal with the fact that it's a CFB that could be dropped out with a conventional amp?
Also, this chip is supposed to be a drop-in replacement for the CXA1145 and has an English datasheet. http://console5.com/techwiki/images/2/20/ES71145.pdf

 

How about a single quad package of this? http://cds.linear.com/docs/Datasheet/lt1253.pdf

EDIT: That one is also a CFB. Don't know what your design calls for. Does it includee a lot of circuitry to deal with the fact that it's a CFB that could be dropped out with a conventional amp?
Also, this chip is supposed to be a drop-in replacement for the CXA1145 and has an English datasheet. http://console5.com/techwiki/images/2/20/ES71145.pdf

The quad is LT1254. I don't think it has enough output swing with a 5V supply either. Besides, it is not stocked at Digikey, nor is LT1253 (dual) although you might find them elsewhere.
I like AD8044 (quad), and Digikey has them. They aren't cheap (≈9 bucks), but I think they do everything you need.

I see why the emitter follower is needed on the Y output of the CXA1145. It is intended to drive a 1kΩ load. You need it to drive a 150Ω load.

 

The quad is LT1254. I don't think it has enough output swing with a 5V supply either. Besides, it is not stocked at Digikey, nor is LT1253 (dual) although you might find them elsewhere.
I like AD8044 (quad), and Digikey has them. They aren't cheap (≈9 bucks), but I think they do everything you need.

I see why the emitter follower is needed on the Y output of the CXA1145. It is intended to drive a 1kΩ load. You need it to drive a 150Ω load.

I can happily order that if it fits your design. $9 is not too much; I'm not looking for a well-engineered solution, just one that I can do.

Is it still what you recommend?

 

Here is the schematic. If you want to accurately realize the equations for Pb and Pr, you need to use 1% resistors everywhere except for the 10k units.

The signals need to come from 75Ω sources on the VX1145 circuit, and they need to be terminated in 75Ω loads at the TV (internal, hopefully).
You may have to tweak the 5pF caps across the feedback resistors. These are in place to minimize overshoot and ringing on fast edges. Even if you don't see these effects on the signals from the CXA1145, it doesn't mean that the op amps aren't tempted to oscillate. I think I would test each input with a network analyzer, or, if you don't have access to one, a 1V p-p square wave with fast rise and fall times (<50nS), and tweak the caps as necessary, just to ensure future stability.

Are you going to build this on a PC board, or will it be a one-off breadboard? Let me know, because I will have recommendations on how you build it.

You can add the emitter follower in the Y output of the CX1145. A better option would be a voltage follower with a 75Ω series output resistor, if you have that option.

I may think of other caveats, which I will post. Come back with any questions.

Obviously, I can't guarantee that this circuit will do what you want. If it were mine, I would look at the CX1145 outputs before I started building this circuit.

 

Here is the schematic. If you want to accurately realize the equations for Pb and Pr, you need to use 1% resistors everywhere except for the 10k units.

The signals need to come from 75Ω sources on the VX1145 circuit, and they need to be terminated in 75Ω loads at the TV (internal, hopefully).
You may have to tweak the 5pF caps across the feedback resistors. These are in place to minimize overshoot and ringing on fast edges. Even if you don't see these effects on the signals from the CXA1145, it doesn't mean that the op amps aren't tempted to oscillate. I think I would test each input with a network analyzer, or, if you don't have access to one, a 1V p-p square wave with fast rise and fall times (<50nS), and tweak the caps as necessary, just to ensure future stability.

Are you going to build this on a PC board, or will it be a one-off breadboard? Let me know, because I will have recommendations on how you build it.

You can add the emitter follower in the Y output of the CX1145. A better option would be a voltage follower with a 75Ω series output resistor, if you have that option.

I may think of other caveats, which I will post. Come back with any questions.

Obviously, I can't guarantee that this circuit will do what you want. If it were mine, I would look at the CX1145 outputs before I started building this circuit.

I don't even have a scope capable of video frequencies. I'm prototyping it on a breadboard but I will build it into a project board. My breadboard has thus far proven not to be too terrible for video signals. Thank you.

 

I don't even have a scope capable of video frequencies. I'm prototyping it on a breadboard but I will build it into a project board. My breadboard has thus far proven not to be too terrible for video signals. Thank you.

I don't think you'll have much luck with a solderless breadboard. Try it if you want.
If that doesn't work, I'll tell you how I would do it.

EDIT: IIRC, solderless breadboards have about 20pF of capacitance from each pin bus to ground. In simulation, this causes 32 dB of peaking at 26MHz on the Y amplifier. Simulations are what they are, but it's not good news.

 

I don't think you'll have much luck with a solderless breadboard. Try it if you want.
If that doesn't work, I'll tell you how I would do it.

EDIT: IIRC, solderless breadboards have about 20pF of capacitance from each pin bus to ground. In simulation, this causes 32 dB of peaking at 26MHz on the Y amplifier. Simulations are what they are, but it's not good news.

It'll be a few days before I have the op-amps. How would you do it? Also, is a project board with solder bridges going to be okay for this circuit when it's done?

 

It'll be a few days before I have the op-amps. How would you do it? Also, is a project board with solder bridges going to be okay for this circuit when it's done?

What's a "project board"?

I would buy a board with a matrix of plated-through holes on 0.1" centers, with ground plane on one side only.
I'm retired now, but when I was designing video hardware for a living, I used single-sided copper-clad PC board material (it may be hard to find now). I laid out the board with a pencil on vellum printed with a 0.1" grid, taped this to the copper side of the board, and used that as a template for drilling component lead holes (this was mostly before surface mount). I marked the hole locations with an automatic center punch, then removed the template before drilling.
Then I used a Vector pad cutter (you could get them for use in a drill press, or with a wooden handle for manual use). The problem is, the resulting pads are so small that they frequently get ripped off the substrate by the pad cutter.
I just now ran across this technique.
However you do it, I believe you need to use RF techniques, which includes a generous ground plane. A wire connecting all the grounds together is not satisfactory for 5MHz circuitry, IMHO.

As I said, if I were building this today, I would use the Twin Industries 8100 series which I referenced in the beginning of this post.

 

Alternatively I can lay something out in Eagle and have a custom PCB made, yeah? You suggest a generous ground plane, but the circuit itself is grounded via a 20-gauge wire soldered to the ground pin of the CXA1145. Would you suggest I ground this a bit more seriously?

 

Alternatively I can lay something out in Eagle and have a custom PCB made, yeah? You suggest a generous ground plane, but the circuit itself is grounded via a 20-gauge wire soldered to the ground pin of the CXA1145. Would you suggest I ground this a bit more seriously?

I can't visualize what you mean, but good grounding requires low inductance. Wires have inductance. Planes (basically) don't.

 

If the CXA1145 is on a separate board, you should only need 3 coax cables connecting the two boards. The ground plane I was mentioning would be on the circuit I designed for you.

 

If the CXA1145 is on a separate board, you should only need 3 coax cables connecting the two boards. The ground plane I was mentioning would be on the circuit I designed for you.

The issue is how the ground plane is grounded. The CXA1145 is already in-place in a system from which I'm drawing the voltage and signals. I'm installing this new circuit parasitically; it will exist on its own board.

The circuit isn't grounded independently, it's connected parasitically to the ground (and voltage) of the Sega Genesis, in this case the CXA1145 inside the Genesis. To prevent a ground loop the +5v and ground have to be connected "close" to one another, so I put them both on the CXA1145, though I could also solder them to the pins of a voltage regulator.

My question is whether it's going to be sufficient to ground my ground plane with a 20 gauge wire the like of which I'm using for my signals, and to do so by "sharing" a ground with the existing circuitry around the CXA1145.

 

Can you post a photo of the Genesis board, showing where you plan to connect power, and where the converter board will mount?

 

http://i.imgur.com/zrMAqLA.jpg

This picture isn't mine, but it's the exact model I'm working with. Mine already has some 22-gauge wires soldered to the R G B out, Y and C out, and GND / VCC pins on the CXA1145

Here is a version with whatever ICs I could remember labelled:
http://i.imgur.com/ITBe6vt.jpg

The CXA1145 is on the left. The bottom right pin is pin 1 (it's a little hard to tell in the picture). My ground is soldered to pin 1 and my +5v is from pin 12. I'm not sure where the board will go when it's done, but it will probably end up mounted upside-down on the plastic case of the genesis hovering above the board. This picture also has a heatsink removed that normally obscures the CXA1145 (http://media.photobucket.com/image/recent/Evilweredragon/sys.jpg)

I was looking more carefully at that datasheet, and I notice the CXA1145 has two pins V REF and I REF which internally go into the "Reference" block. Is V REF a virtual ground, then?

EDIT: Additionally, those two "V Reg"s are voltage regulators outputting 5v.

 

Why are there two 5V regulators? Analog and digital?

 

Why are there two 5V regulators? Analog and digital?

The console itself actually takes in DC 10v, so I supposed it's split across these two regulators.

 

The console itself actually takes in DC 10v, so I supposed it's split across these two regulators.

That doesn't compute.

If you are going to mount the new board right above the CXA1145, you probably don't need 75Ω source and load terminations between them. That would require some redesign (because of the gain change), but it would get rid of some resistors.

 

That doesn't compute.

If you are going to mount the new board right above the CXA1145, you probably don't need 75Ω source and load terminations between them. That would require some redesign (because of the gain change), but it would get rid of some resistors.

Why does proximity to the CXA1145 save me the 75Ω load terminations?

Also, what about that V REF on the CXA1145? If it turns out to be a virtual ground halfway between the CXA1145's GND and +5V then can I use a connection to V REF as VG in the circuit?