Still working on GTIA V-Gate?

[+mytek]

3 hours ago, pseudografx said:

I'm referring to this GTIA9-related issue, with dark lines in between certain luminances.

Left one is stock 130XE, right one is my UAV-modified 800XE. Both connected using S-Video.

Hmm... I'd have to pull out my 65XE with UGV installed to see what it does, although from memory I don't recall seeing the black line. I'll go do that now and then post the results.

 

 

EDIT: I just took a screen capture from my 65XE with UGV and I do indeed see a slightly darker line in the middle, albeit it's very subtle.

 

 

 

I do recall seeing this before, and it is due to pushing the output drive capability of the 74HCT08 chip being used to gate the luma as part of the VGATE circuit. This effect is even more pronounced in a 576NUC+ since that unit has an even heavier drive current requirement. In fact on the 576NUC+ I did try piggybacking another 74HCT08 on top of the first one and the darker quality of that line does go away. Of course doing the piggyback soldering of another surface mount chip is very tricky, and not recommended for such a small return.

 

Bottom line what you referred to as a GTIA timing aspect is in realty a slight lack of drive capability only seen in this unique situation. This is not something most people would normally notice while playing games or running an application.

 

So my question for you: is other than seeing this when running ACP or similar 256 color map applications, is this something that is noticeable to you elsewhere? If so can you give an example?

 

EDIT2: Aww wait a minute, perhaps we are talking about two different an unrelated things, because I see in your example image that the line is not only much darker, but also appears in two spots. So if this is some kind of bug with certain GTIA chips, can you go into more detail or provide a link that does? And as far as I recall the UAV doesn't have any circuit that would purposely make something like that go away.

[pseudografx]

1 hour ago, mytek said:

Hmm... I'd have to pull out my 65XE with UGV installed to see what it does, although from memory I don't recall seeing the black line. I'll go do that now and then post the results.

 

 

EDIT: I just took a screen capture from my 65XE with UGV and I do indeed see a slightly darker line in the middle, albeit it's very subtle.

 

 

 

I do recall seeing this before, and it is due to pushing the output drive capability of the 74HCT08 chip being used to gate the luma as part of the VGATE circuit. This effect is even more pronounced in a 576NUC+ since that unit has an even heavier drive current requirement. In fact on the 576NUC+ I did try piggybacking another 74HCT08 on top of the first one and the darker quality of that line does go away. Of course doing the piggyback soldering of another surface mount chip is very tricky, and not recommended for such a small return.

 

Bottom line what you referred to as a GTIA timing aspect is in realty a slight lack of drive capability only seen in this unique situation. This is not something most people would normally notice while playing games or running an application.

 

So my question for you: is other than seeing this when running ACP or similar 256 color map applications, is this something that is noticeable to you elsewhere? If so can you give an example?

 

EDIT2: Aww wait a minute, perhaps we are talking about two different an unrelated things, because I see in your example image that the line is not only much darker, but also appears in two spots. So if this is some kind of bug with certain GTIA chips, can you go into more detail or provide a link that does? And as far as I recall the UAV doesn't have any circuit that would purposely make something like that go away.

It might be the case of different GTIA variants, also perhaps PAL systems are more prone to the issue. I've actually encountered it when I was drawing a picture for a graphics compo a few years back. Because of the bug I had to redo part of the image because the dark artifacts looked very ugly. The issue was with the chessboard pattern I used. I had to change it to just horizontal lines where it was much less disturbing.

capitol.xex capitol-fixed.xex

Edited February 6 by pseudografx

[Overange]

10 hours ago, mytek said:

Wow that's a precision fit if I ever saw one

 

Looking very nice.

 

 

Thanks

 

Just got to produce the Gerber's and BOM for you  

 

[dmsc]

Hi!

On 2/5/2024 at 8:49 PM, mytek said:

EDIT2: Aww wait a minute, perhaps we are talking about two different an unrelated things, because I see in your example image that the line is not only much darker, but also appears in two spots. So if this is some kind of bug with certain GTIA chips, can you go into more detail or provide a link that does? And as far as I recall the UAV doesn't have any circuit that would purposely make something like that go away.

 

This is not a bug in GTIA, but on the way the luma DAC works. The dark areas occur in the places where you go from 0111 bit pattern to 1000, the GTIA output is faster to transition from 1 to 0 then from 0 to 1 (because NMOS chips pull down), so the output briefly goes trough 0000, giving the dark line. The same occurs, but less pronounced, when going from 0011 to 0100, or from 1011 to 1100.

 

The UAV fixes this by latching the GTIA output with a CMOS chip before feeding the DAC - this ensures that all bits transition at the same time.

 

Have Fun!

 

[+mytek]

I've been working on a C64 version of VGATE, and it's coming along pretty nicely. Even though they suffer from a different kind of over scan problem, the solution is very similar in some regards to the GTIA one. What made it tricky was that the Luma has already been processed into analog coming out of the VIC chip, and it was already combined with CSYNC. This required a unique approach. If you guys are at all curious here's a link to my blog where I talk about it.

[olix]

Overange was kind enough to sell me two of his extra-small UGV boards. As these were not assembled, I first had to get the components and then embark on the SMD soldering adventure.

 

Assembling the board is not necessarily suitable for a soldering beginner, but with a little experience and patience you can do it quite well. The most difficult part was actually that I had to switch between three different pairs of glasses during assembly: normal glasses for distance, reading glasses for close-up and magnifying glasses for the very small things. And then a microscope for checking. 

 

But everything worked out well and the two UGV modules worked straight away. Overange did a really good job and made the already good UGV from Mytek even better, as it now fits into the 600XL without any problems. I really love this little piece of hardware.

 

[+mytek]

21 hours ago, olix said:

Overange did a really good job and made the already good UGV from Mytek even better, as it now fits into the 600XL without any problems. I really love this little piece of hardware.

@Overange is your SM Edition ready for public release? I so, would you mind if I add it's gerbers and BOM to my UGV page? Of course I'd be sure to give you full credit

[Overange]

1 hour ago, mytek said:

@Overange is your SM Edition ready for public release? I so, would you mind if I add it's gerbers and BOM to my UGV page? Of course I'd be sure to give you full credit

Of course, please give me a bit of time, about to go on vacation, @olix has confirmed it can be built by hand, @Beeblebrox has also had a couple, so all in all it works really well.

 

I do recommend that you use 1% resistors, as this should in theory, improve the output further. 

 

@mytek please PM me your personal address and I can send you one  

 

 

 

 

[Mathy]

Hello @Overange

 

Olix showed a picture of the backside of the SM Edition on the ABBUC site.  Looks like there's a lot of unused space on the underside/backside.  Would it be possible to move the resistors to the underside to make the PCB even narrower?

 

Sincerely

 

Mathy

 

[Overange]

2 minutes ago, Mathy said:

Hello @Overange

 

Olix showed a picture of the backside of the SM Edition on the ABBUC site.  Looks like there's a lot of unused space on the underside/backside.  Would it be possible to move the resistors to the underside to make the PCB even narrower?

 

Sincerely

 

Mathy

 

Yes you are correct, there is a lot of free space, however, i wanted to keep the similar layout design as seen on the original, as to ensure the output results are keep out as close as possible as the original.

 

 

[+mytek]

3 hours ago, Overange said:

please PM me your personal address and I can send you one

You're in the UK so that would be a bit expensive. Better if I just add the board to one of my future JLCPCB purchases. But thanks for the offer for sure 👍

 

1 hour ago, Mathy said:

Olix showed a picture of the backside of the SM Edition on the ABBUC site.  Looks like there's a lot of unused space on the underside/backside.  Would it be possible to move the resistors to the underside to make the PCB even narrower?

From what I understand it presently fits all Atari 8-bit systems, so there's really no need to shrink it down more. Also with all the components on the topside as they are now, it would be relatively simple to use a stencil and reflow oven for assembly if so desired. As for economics, I suspect there would not be a noticeable difference in cost to make it smaller than what it already is.

[+mytek]

And now to agitate all you Commie Haters

 

I got the newest member of the VGATE family almost ready for show time.

 

VGATE Video Board for the C64 VIC-II

 

It's actually a more integrated 2 piece board set, with a separate interface board meant to take the place of the original RF modulator. And because there were essentially two versions (long board and short board) there are also two different interface boards as well. Nice thing is that no new holes or case modification is required, and we get the S-Video, Composite, and Audio all available from the one interface board. It even accommodates the LumaCode video digitizer output or stereo.

 

As can be seen below, not much is nested inside the socket area, with only the PIC chip and a couple other associated components, since there isn't a space issue in this application.

 

 

 

[+DjayBee]

I like ataribits.com on C64 hardware. 

[olix]

11 hours ago, Mathy said:

... Looks like there's a lot of unused space on the underside/backside.  Would it be possible to move the resistors to the underside to make the PCB even narrower?

 

Mathy

 

I think you should leave the layout as it is now. If there are components on both sides of the board, it only makes assembly unnecessarily difficult. 

You can't save that much space either, because there has to be room for at least the pin header, even if all the other components are moved from the top front to the bottom.

[+mytek]

9 hours ago, DjayBee said:

I like ataribits.com on C64 hardware. 

Glad you spotted that 👍

 

It shows that the innovation began on the Atari, and then spilled over to the C64.

 

However the idea for the RF Modulator replacement as a more universal interface came from a builder on the C64 which I incorporated into my design. Since the Atari 8-bits also has a modulator with two holes in the case (RCA Jack and channel switch) I'll have to investigate if such an interface board can be made for the UGV as well. And just like the C64 platform, there are differences between models, so it may take a couple of different versions. But I really do like the idea of the 3.5mm stereo jack being configurable to yield Composite Video, LumaCode, as well as left and right audio with the use of a dual RCA jack 'Y' adapter.