Internal RGB/component AV switch for NESRGB - some help needed

So, I've got a Famicom AV and one of Helder/Buffalowing's Wii AV sockets. What I wanted to do was hook up an NESRGB and an RGB to component board and connect the AV so that when I plug in a Wii component cable it will output component and when I plug in a Wii RGB cable it will output RGB. This is what I (with very limited expertise) have come up with so far:



Using a MAX333 Quad SPDT CMOS switch, which will be connected as shown when it's in a low state and will output RGB, but when a Wii component cable is connected, it will bridge pins 8 and 10 on the AV output and send +5V to the chip, causing it to output YPbPr. I'm planning on using a TPS60400 voltage inverter to provide the VEE signal to the chip.

Thing is, I have only basic understanding of AC coupling and tying lines, etc so wanted to know if the placement of the resistors and capacitors is correct and if there are any other glaring issues with the idea?

 

This won't work, you're triple terminating each signal. Unless you're going to re-amplify the signal you can't terminate it. Since you have a negative supply you don't need to bias the signal either, just send it straight through. You can put a coupling capacitor on the output, but only one and only if video inputs are dc coupled (I doubt they are so you'll have to remove the coupling caps).

 

Cool, that's what I had originally, but was told on another forum that I should add coupling, so I just threw on a bunch of caps and resistors without really knowing too much about them.

 

Updated again with RGB to component circuit. Just wondering if it's OK to split the RGB signal from the NESRGB as I've done in my schematic or not?

 

Are you sure about that component encoder? It doesn't look very correct. My guess is that the BA' puts out 0.7 Vpp out when you should have 1.4 Vpp out in order to series-back-terminate in addition to parallel terminate at the TV's input). If the signal is 0.7 Vpp, you have to remove the series 75 ohm because it will attenuate the signal, or get a real video amplifier with signal gain.

The output transistor buffers are a good idea if the BA's outputs can't source enough current, but such a simple amplifier won't give you good linearity until you're working with very wide supply rails and a high current emitter resistor, and it won't have a perfect 0 ohm output impedance (you need a real op amp circuit for that), maybe up to 20 ohms. If you end up keeping the series back-termination 75 ohm R, you'll have to take the output impedance into consideration (so I'd start with a 56 ohm R instead of 75). If the signal is only 0.7 Vpp you have to remove the 75 R and know that your signal will be attenuated a little regardless because it's as if there's a 20 ohm resistor in series with the signal.

Also you don't really want to drive an ac signal *into* the switch, it's not terribly safe due to the clamp diodes. It should be dc into the switch, then ac coupled at the switch's output. You'll also save some caps.

Lastly, technically you aren't supposed to use a LM1881 like that, they output a digital TTL sync signal, TVs receive analog sync signals. When you drive a digital signal into a TV, the TV's parallel termination (75 ohms) heavily loads the TTL output, which is meant to drive 10k+ ohms, so there's no way it can supply enough current and you get a deformed signal and a stressed '1881. TVs only require a 300 mV or so signal into 75 ohms, and TTL signals are ~3 Vpp (into >10k ohms), so 10x larger in amplitude than expected which may lead to non-symmetrical clipping at the TV and further degrade the sync quality.

I don't think there's a reason for you to use a LM1881 at all with the NESRGB, there should be a real Csync output which is intended to drive a TV.

 

You are correct, there is a Csync pad you can solder to on the NESRGB labeled CS#

 

The RGB to component circuit is based on Ace's design:



I did think about the real need for the LM1881, but forgot about it as I was basically just adding Ace's circuit as is onto my current design. Pin 4 of the NESRGB is a TTL csync signal, so I've removed the LM1881.