Still working on GTIA V-Gate?

[olix]

@mytek: Are you still working on the V-Gate Project? On the AtariBits Website is a massage, not to purchase the PCB cause of an error in the design.

 

I really like this project and would like to use the vgate in some of mine Atari XL.

[+mytek]

9 hours ago, olix said:

@mytek: Are you still working on the V-Gate Project? On the AtariBits Website is a massage, not to purchase the PCB cause of an error in the design.

 

I really like this project and would like to use the vgate in some of mine Atari XL.

One of the problems that has kept me away from this, is that it kept evolving as I incorporated into the various alternative motherboard projects that I released. Basically there was always something else and/or better I wanted to do with it. And the board you see pictured was also somewhat responsible for frying the U1MB I had in the XEGS full install pictured at the bottom of that webpage. It was entirely my fault, since I plugged a cable in offset by a pin or two, but it sure left a bad taste in my mouth. So if I was ever to revisit this again, first I would have to map out the pins in a better safer way, or eliminate the UiMB expansion header all together.

 

To tell you the truth I really have serious doubts about ever restarting this project. Because I like doing new things for one, and secondly there's no incentive for me to revisit this.

 

However there is source code and firmware available for most all the iterations used within the XEL, XLD, and NUC project downloads. And there are schematics for all all those projects showing incorporation, so I see nothing stopping someone else from creating what you seek. And this is all in the public domain.

 

[Panther]

What is the issue with this particular design?

 

[+mytek]

44 minutes ago, Panther said:

What is the issue with this particular design?

 

The no-color blanking during over scan is floating and subject to noise.

 

[olix]

14 hours ago, mytek said:

...

To tell you the truth I really have serious doubts about ever restarting this project. Because I like doing new things for one, and secondly there's no incentive for me to revisit this.

...

 

This is a little bit sad, but of course i understand you.
Perhaps there is someone who will continue your work.
But i would like to take the opportunity to say thank you for the other great projects of yours (TK-II, NUC, 1088 XEL, XEP-80II ... )

[+mytek]

Ok here's a new twist on the V-Gate project that will be piggybacking on 576NUC+ technology.

 

I decided to make it all inclusive, and thus incorporate video output of near UAV quality, while fully supporting V-Gate over scan correction.

 

However unlike the UAV, this one will be piggybacking the GTIA chip instead, leaving the original CD4050 circuit untouched and still usable (if so desired).

It's default operation will be with V-Gate always active, but if you wish to toggle it ON or OFF there is provision for doing so. It also incorporates automatic PAL detection same as the NUC version, so no need for manual configuration.

 

As the schematic says, this is PRELIMINARY, and most likely some of the resistor values will need to be tweaked.

 

Edited May 25, 2022 by mytek

[Panther]

Great!  Some day the PIC12F1572's will be in stock again...

[+mytek]

56 minutes ago, Panther said:

Great!  Some day the PIC12F1572's will be in stock again...

Well I'm taking a long shot and purchasing 10 of the SOIC-8 version from Ali-Express for about $2 a piece, which also includes shipping. So we'll see what happens

 

As you pointed out, all the usual sources have zero stock with a projected resupply in 2023. Seems like so much stuff is suppose to magically appear next year.

 

[Panther]

At least new factories are being built, in the U.S. and Western Europe, so maybe we really will start to see stock in a year.  There are DIP versions of the PIC12F1572 currently available, perhaps you could design a board that uses those while keeping everything else SMD?  I would certainly like to try out your new design (sooner than a year).

 

Edited May 25, 2022 by Panther
It's late.

[+mytek]

29 minutes ago, Panther said:

...perhaps you could design a board that uses those while keeping everything else SMD?  I would certainly like to try out your new design (sooner than a year).

I think I'll do a dual footprint so it can go either way.

 

[olix]

Will it be possible to program the PIC with the Joy2Pic Interface with the Atari?

 

[Gunstar]

I have apparently missed previous talk of the V-gate, and am unaware of what it is. Since it's only going to be available with Atari 8-bit remake machines it's pure curiousity on my part since I stick with upgrading original equipment. Is this something to do with hiding the overscan "graphic artifacts" (I forget the proper name) on the sides that show up on modern wide screen TV/monitors? If that's what it is, it's a non-issue to me anymore since I'm changing back to 4:3 CRT screens on my systems. Just curious...

Edited May 25, 2022 by Gunstar

[Rybags]

My impression was it was to be for any machine.  Primary purpose as you guessed - truncate (black-fill) the overscan region left/right which would mainly be useful for widescreen and capture devices.

Edited May 25, 2022 by Rybags

[+mytek]

7 hours ago, olix said:

Will it be possible to program the PIC with the Joy2Pic Interface with the Atari?

Most definitely. It'll just require a cable from the JOY2PIC's ICSP header and the one on the UVG board if using the SOIC-8 PIC chip.

 

2 hours ago, Gunstar said:

I have apparently missed previous talk of the V-gate, and am unaware of what it is. Since it's only going to be available with Atari 8-bit remake machines it's pure curiousity on my part since I stick with upgrading original equipment. Is this something to do with hiding the overscan "graphic artifacts" (I forget the proper name) on the sides that show up on modern wide screen TV/monitors? If that's what it is, it's a non-issue to me anymore since I'm changing back to 4:3 CRT screens on my systems. Just curious...

If you are indeed reverting to 4:3 monitors, then yes the over scan elimination feature will not be something you need.

 

Here's a demo of the over scan elimination feature on an earlier unreleased V-Gate prototype from a few years back (sorry for the poor YouTube video quality, it was a cheap camera).

 

 

1 hour ago, Rybags said:

My impression was it was to be for any machine.  Primary purpose as you guessed - truncate (black-fill) the overscan region left/right which would mainly be useful for widescreen and capture devices.

Yep this is correct - since it piggybacks GTIA, any machine is game.

 

And although this will incorporate the V-Gate aspect for over scan elimination on wider screens, it should be kept in mind that it is much more than this. It also includes an Atari independent Composite and S-Video output of similar quality to a UAV upgrade, although lacking the artifact adjustment of the UAV. So for users that don't presently have a UAV installed, it provides two solutions to common Atari related video issues (poor video quality and over scan).

 

---------------

 

BTW, due to the nature of deriving the video signals directly from the piggybacked GTIA, the original CD4050 video is unaffected and still usable. And even a UAV upgraded version will still work (so in essence you would have duplicated video outputs available - one with over scan and one without).

 

Edited May 25, 2022 by mytek

[olix]

These are really great news. I am very happy that you work on this project again after that long time.  

 

Will you release the Gerber files like you do with your other projects?

[+mytek]

20 minutes ago, olix said:

Will you release the Gerber files like you do with your other projects?

Yes I will, although that'll be a bit down the road since I haven't even begun the PCB layout yet.

 

[Panther]

10 hours ago, mytek said:

I think I'll do a dual footprint so it can go either way.

This would be fantastic!  I'll order a few of the DIP version while they're available.

 

[+mytek]

I've been testing a breadboard prototype for the last couple of days and have finalized the design, and slightly changed what 'UVG' stands for.

 

Note: in the final board design the PIC chip will have a dual footprint to accept either an SOIC-8 or DIP-8 device. All other ICs will be SOIC only.

I was able to increase the resistor DAC values which yielded better results, while staying well within the current capabilities of the SN74HCT08 outputs. This was not the case for the original 576NUC+ video circuit due to considerations that needed to be taken to minimize noise induced by the close proximity of digital switching signals (the price you pay for extreme miniaturization). In the NUC the HCT08 outputs current capabilities were exceeded by a factor of 2x, which although still safe for the HCT08 chip in use, does cause some non-linearity in the video luminance. This can be corrected on the NUC by piggybacking another HCT08 on top of the original one.

 

My plan is to also layout a satellite board that contains a Mini-DIN4 and two RCA jacks, which will break-out the video signals derived from 'J3' into the industry standard for Composite and S-Video. The audio jack will derive its signal from the audio feeding into the stock DIN5 jack using a jumper wire. This satellite board is purely optional, and the video signals could instead be sent out the stock Atari DIN5 jack in a similar manner to what people have done with the UAV.

 

The V-Gate function can be either locked ON with a shorting block on 'J2" or controlled by an external switch or U1MB.

 

Utilizing the satellite board, will make this a virtually 'solder-less' installation (only the audio jumper would need to be soldered to the Atari motherboard).

 

Programming the V-Gate PIC via the ICSP will utilize a standard 5-pin 0.1" pitch header to mate with either a PICkit programmer or the JOY2PIC. For people contemplating production runs, leaving out the header and using spring loaded POGO pins might be a better option with the right adapter.

 

 

Next on my agenda is to layout the GTIA piggyback PCB.

 

[Panther]

1 hour ago, mytek said:

Note: in the final board design the PIC chip will have a dual footprint to accept either an SOIC-8 or DIP-8 device. All other ICs will be SOIC only.

? ?

 

There goes the remaining world supply of PIC12F1572 chips!

[+mytek]

Preliminary PCB created (it looks ready, but I still need to look it over to be sure).

 

EDIT: I changed the part ID assignments, so this will no longer match the last schematic I uploaded - when this is considered done and working properly I'll upload a new schematic.

 

TOP

 

BOTTOM

 

I should probably identify the GTIA as the target on the silk screen.

 

It'll require two 20-position machine pin single row terminal strips that will have their insulator blocks sitting flush with the top of the board. Or alternatively a standard 40-position machine pin socket could be used with all of its plastic cross bars removed.

 

 

The headers will all be 0.1" pitch right angle style. Resistors are 1/8 watt THT, and the capacitors are axial THT.

 

If a THT version of the PIC12F1572 (U1) is to be used, it may require stacking another machine pin socket on top for clearance.

 

For all of the XL series units, this should plug into the GTIA socket without any additional stacking sockets underneath, assuming I calculated the available space properly. XEs will of course require de-soldering of the GTIA so that a socket can be soldered in place of it. Not sure if any additional stacking sockets will be required for these units.

 

I'll hopefully be assembling one of these for testing in a couple of weeks

 

Edited May 31, 2022 by mytek

[Ricky Spanish]

Lot's of exciting upgrades for our 8-bit's. It's painfully obvious I have a lot of catching up to do and money to spend haha.

[+Stephen]

42 minutes ago, Ricky Spanish said:

Lot's of exciting upgrades for our 8-bit's. It's painfully obvious I have a lot of catching up to do and money to spend haha.

I haven't had to catch up, but the constant spend since the mid 90s has never slowed down!  Every time I think I have enough, something new and cool comes out, and I have to have it too!

[reifsnyderb]

On 5/31/2022 at 4:57 PM, mytek said:

Preliminary PCB created (it looks ready, but I still need to look it over to be sure).

 

EDIT: I changed the part ID assignments, so this will no longer match the last schematic I uploaded - when this is considered done and working properly I'll upload a new schematic.

 

TOP

 

BOTTOM

 

I should probably identify the GTIA as the target on the silk screen.

 

It'll require two 20-position machine pin single row terminal strips that will have their insulator blocks sitting flush with the top of the board. Or alternatively a standard 40-position machine pin socket could be used with all of its plastic cross bars removed.

 

 

The headers will all be 0.1" pitch right angle style. Resistors are 1/8 watt THT, and the capacitors are axial THT.

 

If a THT version of the PIC12F1572 (U1) is to be used, it may require stacking another machine pin socket on top for clearance.

 

For all of the XL series units, this should plug into the GTIA socket without any additional stacking sockets underneath, assuming I calculated the available space properly. XEs will of course require de-soldering of the GTIA so that a socket can be soldered in place of it. Not sure if any additional stacking sockets will be required for these units.

 

I'll hopefully be assembling one of these for testing in a couple of weeks

 

I tried making a stacking a PCB on a socket by increasing the hole sizes for the single row terminal strips so that the insulation was sitting on the PCB and the pins would stick out the bottom of the PCB.  I found out the pins were not long enough and didn't stick out far enough to engage properly in the socket.  They tended to pop out.  Was there something I was missing?

Edited June 3, 2022 by reifsnyderb

[+mytek]

11 hours ago, reifsnyderb said:

I tried making a stacking a PCB on a socket by increasing the hole sizes for the single row terminal strips so that the insulation was sitting on the PCB and the pins would stick out the bottom of the PCB.  I found out the pins were not long enough and didn't stick out far enough to engage properly in the socket.  They tended to pop out.  Was there something I was missing?

You have to select the 0.8mm board thickness when having them made. The more standard 1.6mm PCB thickness doesn't allow the full shoulder of the machine pin to stick out on the bottom side.

 

Here's an example...

 

40 machine pin socket flush mounted in 0.8mm thick PCB, just behind a standard 0.1" header. PAD Size = 0.075" with Hole Diameter = 0.056"

Edited June 3, 2022 by mytek

[reifsnyderb]

19 minutes ago, mytek said:

You have to select the 0.8mm board thickness when having them made. The more standard 1.6mm PCB thickness doesn't allow the full shoulder of the machine pin to stick out on the bottom side.

 

Here's an example...

 

40 machine pin socket flush mounted in 0.8mm thick PCB, just behind a standard 0.1" header. PAD Size = 0.075" with Hole Diameter = 0.056"

That would do it.  Thanks!  I was using 1.6mm boards and wondering if special pin sockets were being used.  I completely overlooked board thickness.

Edited June 3, 2022 by reifsnyderb