Jump to content
IGNORED

S-Video For Atari 400


Recommended Posts

I like my little Atari 400 but in order to use (play games on) it, I need to get the old TV up from it's storage shelf in the basement as the 400 only has RF output.  (I also have it's RAM upgraded to 1056k.  ?  )  While I know there is the SCCC to do this, it's sold out and it's fun to make my own board.  So, I just ordered 5 of these boards and should know if it works in a little over 2 weeks.  I integrated some of the electronics from the 600XLM board for video and audio.  Also, I took my existing 400 CPU board and tested it out with a SALLY chip to make sure I had the theory right.  The idea is that a Sally will work if /halt is held high and R/W is connected to pin 36 as opposed to pin 34.  It worked.  (See second pic.  I stacked a couple sockets to jump the R/W and connect to the /halt pin as I don't want to solder to a Sally chip.)

 

23233605_400SVideoBoard.thumb.png.9f2b60b39373e84114247f8d985d82dd.png

 

1601287014_SallyTest.thumb.JPG.6d77f45511ee29f7ba79c1306da98e52.JPG

 

 

Edited by reifsnyderb
  • Like 8
Link to comment
Share on other sites

  • 3 weeks later...

I got these boards in last week and, after I assembled one, only had a black screen.  So, I put the board aside to work on the PBI memory card.  So, today, I went back to look at this and found the problem quickly.  U401 and U402 were connected wrong as I made a mistake and literally "shifted" two wires when copying the relevant part of the schematic for the CPU board from the Atari 400/800 Service manual.  Go figure.

 

Anyhow, after cutting two lines and adding two jumpers it works nicely.  I also connected POKEY and had sound.  This board uses the THS7314 video amplifier to get the job done.  The sound amplifier is a simplified single transistor design.  I borrowed the THS7314 idea from Robert de Boer's schematic for the 600XL and used the video amplifier from one of MyTek's designs.  (576 NUC, I think.)

 

I want to test this board out with a 6502B later but I don't foresee any problems as I just tested this with a SALLY chip that has the jumpers set to move the R/W and disable the /HALT line.  So, SALLY is currently working like a 6502B.  There will be some minor silkscreen changes and only 2 lines for sound as the extra ground line isn't needed.

 

The video jack is just rigged up below.

 

I'll probably order some fixed boards this week some time.  If there is some interest, I'll order extras.

 

608177010_400svideoandcompositevideocard.thumb.JPG.33ed38badf25d89a65c1019399fdaff5.JPG

 

  

Edited by reifsnyderb
  • Like 5
  • Thanks 1
Link to comment
Share on other sites

Update:  I've done some more testing and have some interference coming from somewhere.  With PacMan, the screen looks really good.  However, with the BASIC cartridge, there is a wave-like pattern that "flows" from right to left.  This does not happen if chroma is disabled.  So, I am under the belief that either my "flying" connector is the problem or that the color pin from GTIA is picking up something.  The AV circuits I am using have been tested on the 600XLM board and show no such problem.  Also, given the nature of the 2 layer boards on the Atari 400, I am leaning towards the interference coming from my "flying" connector.  A difference between this board and the 600XLM is that the color pin goes in between a couple pins on the GTIA chip.  On the next board, the color pin will be kept on the top layer and re-routed such that there is always a ground in between it and anything else.  The next board will also have the top and bottom layers filled with a ground plane in the A/V area to help shield from any other interference.  While I had the +5VDC for the A/V on a separate circuit, on the current board it ran past a lot of data and address lines.  Hypothetically, the +5VDC could have picked up interference from these lines.  The next board has that A/V +5vdc line on the top layer so as to be shielded by the second layer...the ground plane.  I also added ground lines running next to the +5VDC A/V line so as to shield it from adjacent signals.  While I really believe the problem is my "flying" connector, I want to make sure there aren't any more questions about where interference could be coming from after I get the next boards made.

 

Looking at the oscilloscope isn't helping me find the source of interference....then again, it's a pretty old scope.

 

 

Link to comment
Share on other sites

Ok.  I just took apart an 800 to borrow the 6502b and tried it with this board.  The 6502b worked perfectly.  Oddly enough, the graphics when playing PacMan look great.  No problems.  No flicker, no waviness, nothing.  But when I put in the BASIC cartridge or just go to the "memo pad", the waviness is back.  I tried Donkey Kong and the graphics are good there, too.  

 

When Donkey Kong went into attract mode, the waviness comes back.  Below are a couple pictures.  The first shows when the game is running.  Note that the colors don't look right but it's the camera.  The colors are fine.  The second shows in attract mode and you can see some of what I am referring to.  Once Donkey Kong goes back to a mostly black screen with some color on it, it's fine.    

 

One factor seems to be the amount of color on the screen.  If the screen is mostly black with some color it's fine.  But if it's all color, such as the blue screen with BASIC, or in attract mode, it goes bad.  

 

I also tried a different GTIA chip and there was a difference in that the waviness wasn't as bad.  Adjusting the color adjust clears it up just when the screen color in BASIC starts to shift from blue to purple.  On this board, the GTIA color pin (21) is running between two other pins.  So, I just tried a bypass by cutting the lands and soldering a wire so as to get the color wire away from the board.  It seemed to improve.  I am going to think on this some more for a day before ordering some new boards.  I am not sure what else this could be but interference. 

 

 

Note:  Colors are fine.  It's the camera.

979087564_donkeykonggameinsession.thumb.JPG.c0a79b17c1f831e7646b0f6fad140e92.JPG

 

 

Below shows some of the problem. 

P1100687.thumb.JPG.d8f1128b783048e5c6ad17add1209203.JPG

 

 

 

Link to comment
Share on other sites

The boards were on the DHL Courier's truck yesterday.  They have this GPS tracking online web page where you can watch as they travel around.  The truck made it up past my place on the east side of the river, made a bunch of deliveries about an hour north of where I live on both sides of the river.  After 4pm, he was on the west side (my side) of the river and my shipment was re-scheduled.  The greatest part of the story is that the guy had to drive within a mile of my place to get back to Harrisburg, PA, where the distribution center is, and travelling on my side of the river was the fastest way to return to Harrisburg.  I guess that extra mile isn't worth the trouble.  I'd use UPS but they want an extra $20 to deliver because I live in a rural area...even though the UPS truck goes by my place at least once a day.  Go figure.

 

On a more Atari-related note, the color problem may not be the board.  I started taking the board apart and found the GTIA color pin was bent but the bend wasn't visible without pulling the chip.  So, if nothing else, the changes I made to the board should add more shielding and make for a better board.  I'll know once the DHL guy finally decides to drop by and I can assemble the board.

 

 

Link to comment
Share on other sites

2 hours ago, reifsnyderb said:

The boards were on the DHL Courier's truck yesterday.  They have this GPS tracking online web page where you can watch as they travel around.  The truck made it up past my place on the east side of the river, made a bunch of deliveries about an hour north of where I live on both sides of the river.  After 4pm, he was on the west side (my side) of the river and my shipment was re-scheduled.  The greatest part of the story is that the guy had to drive within a mile of my place to get back to Harrisburg, PA, where the distribution center is, and travelling on my side of the river was the fastest way to return to Harrisburg.  I guess that extra mile isn't worth the trouble.  I'd use UPS but they want an extra $20 to deliver because I live in a rural area...even though the UPS truck goes by my place at least once a day.  Go figure.

 

On a more Atari-related note, the color problem may not be the board.  I started taking the board apart and found the GTIA color pin was bent but the bend wasn't visible without pulling the chip.  So, if nothing else, the changes I made to the board should add more shielding and make for a better board.  I'll know once the DHL guy finally decides to drop by and I can assemble the board.

 

 

THIS !!!

yup that's the nutty frustration of the data driven but not real world aspect of the digital age. You should be on the exception list as you are on the main drag, which used to be taken into consideration by delivery drivers.

Link to comment
Share on other sites

Boards are in!  I transferred the components to the new board and the S-Video color is great!  I wish all of my Atari's had color this good. There is no sign of interference, either.  The picture is very smooth.  Even the "jail bars" seen on the 800's and others are missing.  The S-Video output is much, much better than a stock Atari 800.  ? 

 

There is one change I made to the lumina resistor ladder:  I changed out the 4.7k ohm resistor with a 6.8k ohm resistor.  This is because pink was really washed out so I figured that raising the resistance of the 4.7k ohm resistor would adjust it.  It actually worked perfectly.  The pink ghost was almost white with the 4.7k ohm resistor and now it's pink.  Keep in mind the video circuit is based off of the RedHawk668 design and on the 600XLM board pink was always washed out but I didn't change out that resistor.  (https://github.com/redhawk668/Atari-600XL-S-Video-Composite)  I did modify the composite part of it because I never got composite to work on the 600XLM board...which used the same circuit.  So, composite video is there and works....it's just...well...ugly composite video.  (In other words, if it's an emergency, your TV only has composite video, and you are having M.U.L.E. withdrawal symptoms this will save your life.)

 

The PacMan colors are the best I've ever seen.  There is one weird issue, though.  With M.U.L.E., the mountains are white unless there is a planet quake.  Then the colors are correct.  I don't have an explanation for this and was more concerned until I saw the planet quake.  I really don't know why it would be like this.  Seeing the correct colors during the planetquake tells me it's not the card.  Maybe the ANTIC or GTIA chip is the problem here?

 

I have been testing the 400 with either my 48k board or my 1056k board and they both work quite well.  My next step is to get the right cable and install this board in my "good" 400 so as to have a really cool 1056k Atari 400 with S-Video.  ?

 

 

  • Like 1
Link to comment
Share on other sites

Pic of board.  I am still trying to figure out what cable would work best for it.  I'll need something with 4 conductors as I have no interest in using composite video.  Recently, I got a bunch of the right female DIN-5 connectors to install on the end of the cable.

 

 

420825696_400CPUVideoBoard.thumb.JPG.a64a4915362423d99378c4c3675a2cc8.JPG

Link to comment
Share on other sites

I bought some cables on ebay (one end DIN5F, the other end I can't recall), to cut the other end off, and use this to run out the back like the original RF cable did.  In theory these should be individually shielded wires, which should ensure good video quality w/o picking up noise.  Plus the cable is molded into the DIN5 for a nice 'professional'/'factory' look.

 

 

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

@reifsnyderb here's an approach I've been working on for better quality Atari video that may or may not be of interest to you.

UGV_V1.2_Preliminary_schema.thumb.png.5565125fd7c225e5b87b22b075cd73e5.png

I have made an attempt to correct the lack of a proper black level in this latest version. The stock Atari treats blanking and black as the same thing, which confuses the hell out of many video converters. In my circuit the 555 timer gets triggered off of the low going CSYNC pulse, thus turning ON for approximately the width of two back-to-back CSYNCs. This in turn switches on a MOSFET which applies a resistor to ground to lower the DAC output by approximately 50mv. That becomes the blanking level, whereas when the timer releases the MOSFET and the level rises by 50mv you now have the 'True Black' level. Technically a step-down to blanking is also suppose to appear just prior to the CSYNC pulse, but I wasn't able to do that without much more complexity in the circuit. Since most video converters appear to reference 'Black' following sync, this partial restoration seems to do the trick.

 

I can feed this through a very inexpensive S-Video to HDMI converter and get some pretty spectacular results.

 

image.thumb.png.31a96adebcc57053722d114de405c0b2.png

AliExpress buy LINK

 

 

Although it looks very good with my bread boarded circuit, it should be even better when my sample PCBs arrive.

 

This circuit is in the public domain with no restrictions.

 

 

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

2 hours ago, cwilbar said:

I bought some cables on ebay (one end DIN5F, the other end I can't recall), to cut the other end off, and use this to run out the back like the original RF cable did.  In theory these should be individually shielded wires, which should ensure good video quality w/o picking up noise.  Plus the cable is molded into the DIN5 for a nice 'professional'/'factory' look.

 

 

Great idea!  I'll take a look on ebay and amazon.  It would be a lot cheaper to modify a cable then buy the cables and connectors.  The best cable prices are to buy a whole spool at a time for a rather high price.

Link to comment
Share on other sites

Just now, reifsnyderb said:

@mytek:  Thanks for posting that circuit.  It looks great, has the vgate, etc. and thanks for making it public domain.  I'd happily re-do the whole board for it.  Do you have a source for the PIC and FMS6400 chips?

The FMS6400 chips are readily available from eBay as well as AliExpress (never ran into any problems sourcing them that way). I even recently got 5 of the SOIC sized PIC12F1572s from AliExpress as well. But haven't had a chance to test one yet, although they do appear genuine. When I get my PCBs that'll be the moment of truth ;-)

 

  • Thanks 1
Link to comment
Share on other sites

@mytek:

 

I've been puzzling over the composite video produced by my card and it appears that the color is way out-of-phase with the lumina.  When trying to use composite video, and if I look closely at the screen, the color is different and looks like it's attempting to put the correct color just prior to the lumina for the graphic in question.  Looking back through other posts it appears that the Atari UAV has a way of adjusting the phase.  When I look at the schematic for the 800XL, there is an amplifier circuit for the color that has 3 NPN transistors.  It just occurred to me that that 3 transistor amplifier circuit might use 3 transistors to use the propagation delay of the transistors as a phase shift.  (My theory.)  Then, looking at your schematic, there doesn't appear to be a way to compensate for a phase shift.  The circuit I am using is a modification of the redhawk688 600XL S-Video circuit.  While, my S-Video is really nice, the composite clearly has some sort of phase shift problem.  (I've been experimenting and discovered that it doesn't matter much what capacitors I use on the color line just prior to the amplifier chip.)  I am wondering how your circuit handles this phase shift and if you have any other thoughts on the subject.  Thanks!

 

My circuit:

 

111783020_400CPUandVideocard.thumb.png.37b19a96234eb2fb72b2b3ff66066a42.png

 

 

800XL (w/ chroma fix)

 

800xl.thumb.png.addd46919d72ba1fe944f73b495d613e.png

 

 

 

Link to comment
Share on other sites

5 hours ago, reifsnyderb said:

I've been puzzling over the composite video produced by my card and it appears that the color is way out-of-phase with the lumina.  When trying to use composite video, and if I look closely at the screen, the color is different and looks like it's attempting to put the correct color just prior to the lumina for the graphic in question.  Looking back through other posts it appears that the Atari UAV has a way of adjusting the phase.  When I look at the schematic for the 800XL, there is an amplifier circuit for the color that has 3 NPN transistors.  It just occurred to me that that 3 transistor amplifier circuit might use 3 transistors to use the propagation delay of the transistors as a phase shift.  (My theory.)  Then, looking at your schematic, there doesn't appear to be a way to compensate for a phase shift.  The circuit I am using is a modification of the redhawk688 600XL S-Video circuit.  While, my S-Video is really nice, the composite clearly has some sort of phase shift problem.  (I've been experimenting and discovered that it doesn't matter much what capacitors I use on the color line just prior to the amplifier chip.)  I am wondering how your circuit handles this phase shift and if you have any other thoughts on the subject.  Thanks!

Your issue is that the LUMA circuit is still using the slow CD4050 buffer with the 3904 transistor, while the delay caused by the transistors on the COLOR circuit has been eliminated. You could try replacing the CD4050 with the CD74HC4050E, it's pin compatible and may be fast enough to resolve this. You could also try other fast non-inverting hex buffers, these would probably have a different pinout/PCB footprint

 

I prefer the FMS6400 as a video buffer despite being obsolete, it creates the COMPOSITE signal internally instead of with passive components. 

  • Like 2
  • Thanks 1
Link to comment
Share on other sites

26 minutes ago, BillC said:

Your issue is that the LUMA circuit is still using the slow CD4050 buffer with the 3904 transistor, while the delay caused by the transistors on the COLOR circuit has been eliminated. You could try replacing the CD4050 with the CD74HC4050E, it's pin compatible and may be fast enough to resolve this. You could also try other fast non-inverting hex buffers, these would probably have a different pinout/PCB footprint

 

this ^^^^^^

  • Thanks 1
Link to comment
Share on other sites

Yeah I use an HCT part to buffer the luma, so no appreciable delays are inserted there, and have eliminated the transistor buffers on both the color and luma signal paths. I just let the FMS6400 handle everything , and it appears to do a good job of it.

 

As @BillC pointed out the FMS6400 does a good job creating the 3 distinct video outputs with no fuss or muss, but it is obsolete. There is a non-obsolete version named the FMS6410, but it has no way to decrease the internal gain, which I've found introduces noise problems.

 

  • Thanks 1
Link to comment
Share on other sites

  • 2 weeks later...

The 74HCT4050's arrived today.  I swapped the CD4050 with the 74HCT4050 to test it out.  There "might" be a marginal improvement with composite video.  I stress the word "might".  I guess the best way to describe composite video on this card is to say it's better than nothing.  lol

 

Well, this card is going to have to be S-Video only as the S-Video is really good.  I am not sure what else I can do about the composite video without adding a delay to the signal and there really isn't room on this board for much more.  I'll keep studying up on it to try and find a way to improve the composite video.  The two jumpers to swap between a 6502b and a 6502c Sally work great.  Audio sounds good as well.

 

On the card below, I didn't install the 2 resistors and capacitor for composite video as there is no need.

 

1772903021_400svideocard.thumb.JPG.441c28a22d4c9c6c148a341f8263b217.JPG

Edited by reifsnyderb
Link to comment
Share on other sites

Here is the 400 S-Video with Pacman.

 

1142122959_400svideowpacman.thumb.JPG.f27222adfdf1ec73a4fccde7fa73be4a.JPG

 

 

Here is the 400 S-Video with SpartaDOS X:

 

(I have the 1056k SRAM card installed.   ? )

 

Note that the purple "waves" and purple discoloration are picked up by the camera but are not visible to the naked eye.  The blue background looks great to the naked eye.  The slight, very mild, purplish "foreshadowing" of the white colors is slightly visible to the naked eye.

 

995813437_400svideowspartadosx.thumb.JPG.5797e2cc1c6d52459e4372022c7cf255.JPG

 

  • Like 3
Link to comment
Share on other sites

On 7/3/2022 at 1:37 AM, reifsnyderb said:

The shielded cable came in today.  I assembled a new card and cable with a female DIN5 connector on the end.  Lumina, Ground, Audio, and Chroma are connected.  The S-VIDEO is now even better.  I'll soon install this in my "primary" Atari 400.  ?

 

1523013216_cardandcable.thumb.JPG.7523666dc69b4e0613859da074a80e84.JPG

Sorry,

stupid question: 

 

will this module be available for purchase for other users?

 

Thx and cheers,

andY

Edited by andymanone
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...