Your technical knowledge was inspiring to see. I'm not planning to become any sort of engineer, but I would like to know how to work on my consoles and such at some point
This is an excellent video! The only suggestion I have is that you should use the pin 2 location for +5 as the position you used for the +5 will be turned off if the channel select switch is in the other position. Pin 2 has a constant +5 regardless of the channel switch selection.
You should try one more audio mod: grab the Stereo output off of the TIA. Yup, the original Atari 2600 was supposed to be stereo, with two discrete channels, L and R. Once you hear it, there's no going back to mono. It's pins 12 and 13 on the TIA. The sound is combined on the board via a jumper trace just after the TIA, so you'll have to cut that trace for the mod to work. There are several Homebrews such as Skeleton+ that pretty much require the stereo mod to complete the game, because hints come through the separate channels of audio.
I noticed there were two audio lines coming out of the TIA that were summed together, but I never would've guessed that game devs actually wrote different data to both channels! I always just assumed only one channel was being used, so I thought splitting it off to separate right/left channels after they're joined would give the same effect. Very cool - I'll have to try that one out and play Skeleton+! Thanks for the tip!
Really great information presented in a clear way. However, your video's music level are way higher than your voice levels. If you even them out more, it would be less jarring.
Thanks for letting me know - that's very helpful! This was one of my earlier videos, so I'd be curious if you think the leveling is still off in my more recent videos - particularly the one that's scheduled to release about 8 hours after I write this reply.
I thought composite video was supposed to be in 1v p2p range? Doesn't your transistor amplifier push the amplified output into multiple volts? It would seem to me that you've taken the cvideo signal, added bias and pushed it through the transistor, its probably in the 2v p2p range and is a really "hot" video signal. Many modern televisions will gladly accept that and even older ones, but it's definitely out of spec. What does it look like on an oscilloscope?
Yeah, there’s unfortunately not a lot of clear information about the input tolerance on TVs, so I’ve also have had some questions about the voltage levels myself. The signal is still 1v P2P with reference to the horizontal blanking interval, but the bias compensation is adding about 1.3v overall. I think it’s pretty normal for color data to exceed 100 IRE (I’ve seen this myself when measuring ‘standard’ composite video output signals on a scope), but the TV should interpret the saturation point at 100 IRE. So I think it’s standard for the composite input at the TV to be able to handle a higher voltage. I read somewhere a while back (sorry, don’t have the source reference handy, and a quick search didn’t pull it up) that CRTs anticipated over 30v on composite input (!?), which sounds a bit crazy to me. Compared to that, 1.3v doesn’t seem like all that much. 😁 Really, though, it would probably be best to put a resistor on the output signal to bring it back into range… though I’ve used this same amp circuit in several different game systems on many displays and haven’t cooked anything yet… doesn’t mean it won’t happen one day, though 🔥
@@whatskenmaking thanks, tbh.... I'm not the authority on it either! The only thing I can think of is color can be different when the voltage is out of spec. That 30v standard may be for pure black and white signals. My understanding of the internals is loose at best but I can imagine that as the color signal is split off the electronics decoding that can be sensitive to that voltage being so high. Where as the video signal stability is handled by more robust electronics. I think in this application I would use an opamp in a voltage follower configuration, that gets you away from the bias issue and gives you a solid current sink to draw upon.
@@prozacgodretro Well, the composite spec says that pure white should be 100 IRE. 1 IRE = 1/140v, so 100 IRE is 714mV. Black is 7.5 IRE. Anything sent above 714mV will be past the saturation point and get interpreted as max brightness white… and this is all referenced to the blanking voltage (which is 0 IRE), which is why I can bump up the voltage with the bias and the colors are still normal (and that’s also the reason why we need the bias in first place… because the blanking is referencing too high without it). I honestly don’t think the extra 1.3v is going to cause a problem with any displays, but I hear ya … the opamp is another good option… probably would be cheaper to just put a resistor on the output though… or a really, really, really long composite cable 😂
@@whatskenmaking You make me really really want to understand the standard better haha!! I'm out of my depth here! Thanks for chatting, it's good stuf!
Just what every 7800 needs!
One of the best video mod videos I have seen for the Atari 7800.
Respectfully
Glen
This is a great mod you did lad. I knew you gave your fully restored Atari 7800 a new makeover.
Your technical knowledge was inspiring to see. I'm not planning to become any sort of engineer, but I would like to know how to work on my consoles and such at some point
Definitely deserve more views. Another cool video
Great video Ken!
Very helpful video 🕹👍
This is an excellent video! The only suggestion I have is that you should use the pin 2 location for +5 as the position you used for the +5 will be turned off if the channel select switch is in the other position. Pin 2 has a constant +5 regardless of the channel switch selection.
Thank you! I actually realized myself after I filmed this video and ended up rewiring it to pin 2. Thanks for pointing that out!
You should try one more audio mod: grab the Stereo output off of the TIA. Yup, the original Atari 2600 was supposed to be stereo, with two discrete channels, L and R. Once you hear it, there's no going back to mono. It's pins 12 and 13 on the TIA. The sound is combined on the board via a jumper trace just after the TIA, so you'll have to cut that trace for the mod to work.
There are several Homebrews such as Skeleton+ that pretty much require the stereo mod to complete the game, because hints come through the separate channels of audio.
I noticed there were two audio lines coming out of the TIA that were summed together, but I never would've guessed that game devs actually wrote different data to both channels! I always just assumed only one channel was being used, so I thought splitting it off to separate right/left channels after they're joined would give the same effect. Very cool - I'll have to try that one out and play Skeleton+! Thanks for the tip!
Really great information presented in a clear way. However, your video's music level are way higher than your voice levels. If you even them out more, it would be less jarring.
Thanks for letting me know - that's very helpful! This was one of my earlier videos, so I'd be curious if you think the leveling is still off in my more recent videos - particularly the one that's scheduled to release about 8 hours after I write this reply.
hi do you have a schematic for this mod?
I thought composite video was supposed to be in 1v p2p range? Doesn't your transistor amplifier push the amplified output into multiple volts? It would seem to me that you've taken the cvideo signal, added bias and pushed it through the transistor, its probably in the 2v p2p range and is a really "hot" video signal. Many modern televisions will gladly accept that and even older ones, but it's definitely out of spec. What does it look like on an oscilloscope?
Yeah, there’s unfortunately not a lot of clear information about the input tolerance on TVs, so I’ve also have had some questions about the voltage levels myself. The signal is still 1v P2P with reference to the horizontal blanking interval, but the bias compensation is adding about 1.3v overall. I think it’s pretty normal for color data to exceed 100 IRE (I’ve seen this myself when measuring ‘standard’ composite video output signals on a scope), but the TV should interpret the saturation point at 100 IRE. So I think it’s standard for the composite input at the TV to be able to handle a higher voltage. I read somewhere a while back (sorry, don’t have the source reference handy, and a quick search didn’t pull it up) that CRTs anticipated over 30v on composite input (!?), which sounds a bit crazy to me. Compared to that, 1.3v doesn’t seem like all that much. 😁 Really, though, it would probably be best to put a resistor on the output signal to bring it back into range… though I’ve used this same amp circuit in several different game systems on many displays and haven’t cooked anything yet… doesn’t mean it won’t happen one day, though 🔥
@@whatskenmaking thanks, tbh.... I'm not the authority on it either! The only thing I can think of is color can be different when the voltage is out of spec. That 30v standard may be for pure black and white signals. My understanding of the internals is loose at best but I can imagine that as the color signal is split off the electronics decoding that can be sensitive to that voltage being so high. Where as the video signal stability is handled by more robust electronics.
I think in this application I would use an opamp in a voltage follower configuration, that gets you away from the bias issue and gives you a solid current sink to draw upon.
@@prozacgodretro Well, the composite spec says that pure white should be 100 IRE. 1 IRE = 1/140v, so 100 IRE is 714mV. Black is 7.5 IRE. Anything sent above 714mV will be past the saturation point and get interpreted as max brightness white… and this is all referenced to the blanking voltage (which is 0 IRE), which is why I can bump up the voltage with the bias and the colors are still normal (and that’s also the reason why we need the bias in first place… because the blanking is referencing too high without it). I honestly don’t think the extra 1.3v is going to cause a problem with any displays, but I hear ya … the opamp is another good option… probably would be cheaper to just put a resistor on the output though… or a really, really, really long composite cable 😂
@@whatskenmaking You make me really really want to understand the standard better haha!! I'm out of my depth here! Thanks for chatting, it's good stuf!
Could you please remove the actual content between the ads. It is quite disturbing.