Atari 7800 S-Video mod schematic w/ pot adjust and other things

[ReilyReed91218]

This is my version of LHE's idea for the 7800 S-Video mod schematic. This schematic is just like LHE's but I never understood why there was a transistor buffer circuit in the middle. No testing equipment any longer so I did the math on the Luma values and dropped the resistance of the luma/sync output  marginally by 3/4th. You dont need those buffer transistors at all because it's combining with OR Gate, the buffering with Hex, then amplifying/buffering with FMS. It's just about the same as the board on OSHPark, but to quote my OSHPark release of this would better help explain thing I think. So here's the quote.

 

Quote

 

This is an Atari 7800 S-Video/Composite video mod. It also allows for stereo output, between left and right. Much like LHE’s board, this board is designed to drop-in, after you desolder the 74LS32. You then put a machine-tooled socket where the 74LS32 was, and solder pin headers to the underside of the board. You desolder this chip to input most of the luma information from the system.

 

Yes, there is stereo output. The 7800 MARIA/SALLY chips do not produce any audio. The TIA chip (used in the 2600) is all that is used to synthesize audio on the 7800. Grab TIA Pin 12 and 13 directly from the chip. You SHOULD disconnect these pins from the board. Or you can cut the traces on them so there’s no output to board. Then solder to those pins.

 

POKEY audio is not really required. But if you own a POKEY chip game (seriously… like 3 games contain a pokey chip) then you need to remove C10 and solder directly the input of C10. I overlayed POKEY ontop of 7800 audio with a 120k resistor. You might ask why I did that for each channel… and simply put, with the 1.8k pullup resistors that are required for audio, you need more impedance on the overlay of POKEY because that is getting pulled up to voltage as well.

This board had no 75 ohm impedance on I/O. If your image is too bright, then you can wire some 75 ohm resistors inline, and from jack to ground (I suggest between the solder cups, and grounding tabs). That should adjust the image a bit. I didn’t put any 75 ohm resistors after FMS6400 output to save space pretty much.

 

Another thing to note is that I swapped the two connections on the switching TS5 chip, to see if that would stop he vertical dot crawl some get when wiring TIA9 to the board. I didn’t know, though, if output on the 74LS174 Pin 5 was high or low when in 7800/2600 mode.

 

If you still get vertical scrolling bars/dot crawl on the output, I recommend re purposing the channel switch to turn TIA9 chrominance on/off. You can also do this with a psychical switch and drill another hole into the system, but I dont recommend that.

 

 

 

I'll post my board for it after I'm done designing it. If people are still getting issues with the TS5 switching chip, then 'll replace the switching chip with a comparable part and get rid of the TS% TI part. The two other shots above are how a similar shamanistic turned out and the other shot is of my currently placement of the schismatic. after it's done, I'll post both the board and the schematic.

Edited November 18, 2021 by ReilyReed91218

[the_crayon_king]

On 11/18/2021 at 12:50 AM, ReilyReed91218 said:

This schematic is just like LHE's but I never understood why there was a transistor buffer circuit in the middle.

Without looking at LHE's circuit I would expect that the transistor buffer was for impedance/summing reasons.

The datasheet for the FMS is not too specific; saying "low impedance sources" for the inputs.

It would be great if they would define "low impedance" for the reader.

 

 

I don't know what the heck R3 is supposed to be doing (on the schematic). At present R3 isn't doing anything AFAIK.  

I don't see the output resistors and capacitors on the Y/C and CV lines. Caps are kind of optional but I don't know of any reason to leave off the resistors.

I also still don't see a reason for two potentiometers on Luma (personally, I wouldn't use any potentiometers). 

 

I see the layout of the board/components looks nice, however some of the vias/traces look like they may be too close. 

The general rule I follow is I won't make them close unless the via or trace shares the same net as the pad.

I don't use KICAD so I could be wrong about what I am seeing here.

If you are curious for everything besides ground and voltage I use 6mil clearance and 6mil trace width. 

 

That all said I don't think you were really looking for critiques, so feel free to ignore me. 

[+batari]

On 11/19/2021 at 2:05 PM, the_crayon_king said:

Without looking at LHE's circuit I would expect that the transistor buffer was for impedance/summing reasons.

The datasheet for the FMS is not too specific; saying "low impedance sources" for the inputs.

It would be great if they would define "low impedance" for the reader.

It's been years since I played with this, but I recall the transistor was configured as a voltage follower, meaning you'll get roughly the same voltage on input as voltage output, but with a current gain. The reason for doing that is so changes in current demand on the output will not affect the voltage, which you want to be well-controlled. Without the transistor, with all of the resistors and pots on the input, any major changes in current will affect the voltage, but the transistor provides isolation from that.

 

As noted, with all of those resistors on the input, the impedance is not well-controlled and likely high, but with the isolation of the transistor that will keep impedance under control as well.

 

This all may not matter much, though, as the FMS chip itself may be OK with unknown impedance and may have relatively constant current demand, and if so you could possibly forgo the transistor.