Movie Cart

[+rbairos]

1 hour ago, swlovinist said:

I tested as much as possible with the very little extra time I have.    I tested two 7800 models, both modded.   One was the one Crossbow did with the Kioparsec Bios, the other an older mod.  

 

Serial No 72R4BR and #AT8444835.   I hear the audio and tested the carts but appears to have the screen issue with the video above.   The controller function pause works.   I tested this on a few different CRT TV sets.   I tried to capture the footage and nothing appeared on the screen using a retrotink.   I also tested the system with a standard 2600 cart and it worked fine.  

 

I have a previous S-video modded and AV modded 7800 that I have had as well.  Serial Number 72R4BR and #A1 75 5814704.  The 75 is a stamped with blue.   Same thing as above I hear the audio and tested the carts but appears to have the screen issue with the video above.   The controller function pause works.   I tested this on a few different CRT TV sets.   I tried to capture the footage and nothing appeared on the screen using a retrotink.   I also tested the system with a standard 2600 cart and it worked fine.  

So can anyone describe *why* some CPUs fail to load Atari 2600 carts at the hardware level?
I have zero experience past the 2600.
Are the clock timing seqeuences different or something?

If I'm being honest, the response since posting has been somewhat overwhelming, so instead of having a perfect general solution, I think round 1 may have to be limited to :

NTSC Only
Atari 2600 / Coleco-Adapter / modded FB2

Maybe a small shipment to 10 AA users (from the original ordered list I kept), who can also give feedback if the above case fails.

Then if that goes well, more of the above, while I get my hands on a 7800 and/or PAL system to attach a digital scope to, to figure out what's happening.

Thoughts?

PS. Still open to any quick firmware fixes that would address this.
 

[4ever2600]

Hoping I'm on the original ordered list... crosses fingers. ; )

[SuperZapperRecharge]

1 hour ago, Glorkbot said:

Although I scour the Atari boards every day, I somehow missed this thread until today. Guess I missed that preorder!

 

Did I see someone say there's a way to do this using Stella? EDIT: oh I see it doesn't work on Stella yet. And I don't think Gopher works on Mac.

 

Gopher does work on Mac. You may need to build it yourself. If you go top the Gopher topic there is a discussion about running Gopher on a Mac. Some developers are using Gopher to develop 2600 and 7800 carts on a Mac.

[SuperZapperRecharge]

36 minutes ago, rbairos said:

So can anyone describe *why* some CPUs fail to load Atari 2600 carts at the hardware level?
I have zero experience past the 2600.
Are the clock timing seqeuences different or something?

If I'm being honest, the response since posting has been somewhat overwhelming, so instead of having a perfect general solution, I think round 1 may have to be limited to :

NTSC Only
Atari 2600 / Coleco-Adapter / modded FB2
 

I think if you say 7800 unsupported at this time it would be sufficient.

 

Most people that have a 7800 have at least one 2600  I have a few 7800s I can test with, but they are not all in the same location and one is sealed in box. I could pop the sealed 7800 open too. It's a 1987 model, A1 76 5796620.

[Thomas Jentzsch]

2 hours ago, Glorkbot said:

Although I scour the Atari boards every day, I somehow missed this thread until today. Guess I missed that preorder!

 

Did I see someone say there's a way to do this using Stella? EDIT: oh I see it doesn't work on Stella yet. And I don't think Gopher works on Mac.

 

It used to work in Stella, where did you see that? Or was the format changed again?

[+rbairos]

14 minutes ago, Thomas Jentzsch said:

It used to work in Stella, where did you see that? Or was the format changed again?

Sept 16 2023 I extended the format to specify overscan, vsync etc to support PAL
https://github.com/stella-emu/stella/commit/b2004d0c496bbc8869439fbfd0fce16dcff3a328

It's been merged, but no compiled release has been put out yet.
 

[Thomas Jentzsch]

I works fine in the unreleased Stella.

[+-^CrossBow^-]

@rbairos Not sure if I'm on part of the original list either. I know I've expressed interest in getting one since they were first shown. I've also currently have 4 different 7800s that I can test it with along with just about ever variant of NTSC 2600. But my main interest and testing would be on the 7800s. I've got an '84 made unit and then like 2 A1 series and an A3 series. But reality is that the series numbers mean little since my '84 mainboard made 7800 was actually bought in late '88 and is an A3 series but is actually a much older mainboard internally.

 

[+rbairos]

4 minutes ago, -^CrossBow^- said:

@rbairos Not sure if I'm on part of the original list either. I know I've expressed interest in getting one since they were first shown. I've also currently have 4 different 7800s that I can test it with along with just about ever variant of NTSC 2600. But my main interest and testing would be on the 7800s. I've got an '84 made unit and then like 2 A1 series and an A3 series. But reality is that the series numbers mean little since my '84 mainboard made 7800 was actually bought in late '88 and is an A3 series but is actually a much older mainboard internally.

 

I'll PM you. 
Thanks.
 

[+swlovinist]

I might have another one to test this weekend, but buried with work until then.  

[+sfish]

I'm eager to try one out, and am happy to provide testing and feedback. Having said that, if there are other people ahead of me on the list, I'm also happy to patiently wait until its my turn in the queue.

[+MattelAquarius]

Please add me to the list.  I will likely purchase at least 2, as these will make the ultimate nerd gifts. I have a 7800 and two 2600's. 

'

[+ZeroPage Homebrew]

@rbairos I was able to get some testing done on my non-modded 7800 today. The MovieCart works with it but with some caveats.

 

1) It won't boot into the MovieCart program a certain percentage of the time. It failed 6 out of 16 boots during the video test below.

2) Unlike my UAV modded 7800 system, the picture and sound is always distorted on this one, but not beyond the point of recognition. When pausing the video, the audio will stick. You can hear it when I pause two times starting at 41 secs

 

Here's the video of me booting it up and playing around with various functions (Fast Forward, Rewind, B&W, Select, Reset) of the MovieCart and then doing multiple boots to show the boot failure percentage.

 

Hope this helps! Let me know if you need any more tests done.

 

- James

 

Video on my non-modded Atari 7800:

 

 

My non-modded Atari 7800:

 

 

 

My UAV modded Atari 7800:

 

[+rbairos]

Thanks so much James!

 

[Thomas Jentzsch]

So far I have only tested short videos. But I wonder if the forward/backward speed is sufficient for long movies. Maybe it could be made e.g. 5 or 10x as fast while the fire button is pressed. If feasible.

[+rbairos]

1 hour ago, Thomas Jentzsch said:

So far I have only tested short videos. But I wonder if the forward/backward speed is sufficient for long movies. Maybe it could be made e.g. 5 or 10x as fast while the fire button is pressed. If feasible.

Its not bad for long movies, as the speed accelerates.
A two hour movie gets to the end in a couple minutes I believe.
 

[JetSetIlly]

3 minutes ago, rbairos said:

Its not bad for long movies, as the speed accelerates.
A two hour movie gets to the end in a couple minutes I believe.

Using the driving controller to shuttle through the movie would be a nice touch.

[+rbairos]

@batari  I notice in an old post, you wrestled with 7800 timing issues to get Harmony cart to work:
 

Do you think something similar may be happening with the MovieCart?
I also monitor address changes then act accordingly.
I wonder if I'm reacting too quickly to them for example.
Part of me thinks all my troubles can be saved by a nop()
Thanks.

 

[Thomas Jentzsch]

47 minutes ago, rbairos said:

Its not bad for long movies, as the speed accelerates.
A two hour movie gets to the end in a couple minutes I believe.

Cool, I didn't know (or remember ).

Edited March 8 by Thomas Jentzsch

[Thomas Jentzsch]

42 minutes ago, JetSetIlly said:

Using the driving controller to shuttle through the movie would be a nice touch.

Do you think the low resolution would be sufficient?

Edited March 8 by Thomas Jentzsch

[JetSetIlly]

6 minutes ago, Thomas Jentzsch said:

Do you think the low resolution would be sufficient?

I don't know. I've never used one. But I like the idea 🙂 

[+batari]

1 hour ago, rbairos said:

@batari  I notice in an old post, you wrestled with 7800 timing issues to get Harmony cart to work:
 

Do you think something similar may be happening with the MovieCart?
I also monitor address changes then act accordingly.
I wonder if I'm reacting too quickly to them for example.
Part of me thinks all my troubles can be saved by a nop()
Thanks.

 

Yes, it is very possible that this is the issue. Essentially, the problem is that on these problem consoles, soon after an address change where A12 is high, it will go low for a short time before going high again. What I had to do is ignore A12 in cases where there should not have been enough time for a legitimate low to occur.

[+rbairos]

47 minutes ago, batari said:

Yes, it is very possible that this is the issue. Essentially, the problem is that on these problem consoles, soon after an address change where A12 is high, it will go low for a short time before going high again. What I had to do is ignore A12 in cases where there should not have been enough time for a legitimate low to occur.

Hm. I use A11 (ignoring A12) for FB2 compatibility.
Do you think A11 might act the same way?

Just a guess, but I wonder if I get the interrupt for A12 high,  get into the subroutine, treat it as low, exit,  but not have enough time to get back in and place data in time when it goes back high again.
I think I'll have to get my hands on a 7800 and attach the probe.
Do you remember off hand, what the dip timing looked like?

My hope is that whatever change I make to get 7800 working, also stabilizes other consoles like Pal 2600 JRs.


 

[+batari]

5 minutes ago, rbairos said:

Hm. I use A11 (ignoring A12) for FB2 compatibility.
Do you think A11 might act the same way?

Just a guess, but I wonder if I get the interrupt for A12 high,  get into the subroutine, treat it as low, exit,  but not have enough time to get back in and place data in time when it goes back high again.
I think I'll have to get my hands on a 7800 and attach the probe.
Do you remember off hand, what the dip timing looked like?

My hope is that whatever change I make to get 7800 working, also stabilizes other consoles like Pal 2600 JRs.

A11 is not affected by the 7800, so you should be good there. So it sounds like A12 is not going to cause you any problems.

 

That being said, for posterity, my recollection with A12 is that it would go low for about 100ns around the middle of the cycle, so yes, this is rather inconvenient for a lot of devices.

[+rbairos]

46 minutes ago, batari said:

A11 is not affected by the 7800, so you should be good there. So it sounds like A12 is not going to cause you any problems.

Drats.  Must still be a timing issue somewhere.
I also wonder if there's some extra current draw with the SD card causing it to glitch out, though I did add alot of 0.1 caps.
Zero Page's video is interesting in that audio plays, and most of the graphics, but seems some writes to registers are failing somehow.