Slight glitches in playfield as 2600 warms up.

[32shot]

Hello all! First time posting here after a long time lurking...

 

Was given this NTSC light 6er quite a while ago, cleaned it up and refreshed it with new capacitors on the switch board (C201 on main PCB was left untouched) as well as a new voltage regulator (kit from Console5). It works beautifully with the 15 or so cartridges i have here.

 

However, a strange issue occurs in NTSC Video Olympics (and seemingly *only* Video Olympics...). As the machine warms up, small flickering glitches appear on the top part of the playfield, right under the scoreboard (please see picture). These appear for about a minute or two just after startup and stabilise afterwards.

 

Another interesting thing is that this does not happen at all on a PAL copy of Video Olympics i happen to have on hand, showing a constant, stable display from the start.

 

I have found a great deal of info on this forum regarding TIAs warmup cycle and the effect that can have on screen but i thought it was odd this would occur on such an early release as Video Olympics, and again, only in its NTSC form?

 

In advance, many thanks for all feedback!

[+Andrew Davie]

I assume you refer to the experiments I did with one machine which showed sprite glitching based on TIA temperatures?  If not, I will dig up the post for you.

[32shot]

That's right, i saw your post, along with the bus stuffing experiments as i was looking for some info... really interesting stuff!

 

Clearly we're talking about hardware and tech that is close to half a century old, yet i can't help wondering is this normal behaviour for a stock, unmodded VCS and no hairdryer involved?

 

Thanks a lot for your input!

[+Andrew Davie]

14 hours ago, 32shot said:

That's right, i saw your post, along with the bus stuffing experiments as i was looking for some info... really interesting stuff!

 

Clearly we're talking about hardware and tech that is close to half a century old, yet i can't help wondering is this normal behaviour for a stock, unmodded VCS and no hairdryer involved?

 

Thanks a lot for your input!

 

Well, it's analogue hardware being pushed to the very limits of exact timing. There are analog components involved which are going to be affected in "performance" by variations in manufacture/materials and temperature. I'm very much not surprised a hairdryer could trigger something that was "right on the edge" in the first place. Happy to have found/demonstrated it.  So yes, it's normal behaviour. But mostly software doesn't push things right to the edge so it's never seen. That's my view on what's up.

[32shot]

Thanks for your input Andrew! The point about the analogue side of things is an important one.

 

Combat, for example, doesn't exhibit this at all, and i had assumed it would have quite a bit of code in common with Video Olympics.

 

Makes one wonder, what does Video Olympics do using the playfield that Combat doesn't...

Edited April 8, 2023 by 32shot

[danny_galaga]

Maybe the problem is just that one particular cartridge, rather than the machine? Do you have another machine? Or someone close by that does?

[32shot]

The cartridge is the first thing I suspected since everything else i have here works flawlessly. Upon opening it up I discovered a rather dodgy factory solder work on the PCB although connections checked out ok using a multimeter.

 

Reflowing properly made no difference, that’s when I thought I’d write about it here…

 

Nobody around me still do VCS as you can expect so no real access to another machine I’m afraid, especially an NTSC one.

 

I have another copy of the game on the way, will see how that one will behave…

[+SpiceWare]

On 4/7/2023 at 6:20 AM, 32shot said:

However, a strange issue occurs in NTSC Video Olympics (and seemingly *only* Video Olympics...). As the machine warms up, small flickering glitches appear on the top part of the playfield, right under the scoreboard (please see picture). These appear for about a minute or two just after startup and stabilise afterwards.

 

 

Looking at the code, the line is drawn using pointers in RAM to read values from ROM that get stored to the Playfield (registers PF0, PF1, PF2).

 

 

Your issue is showing up in PF1, so I can think of 2 potential reasons for what you're seeing:

• your ROM is starting to exhibit bitrot
• there's a problem with RAM at addresses $9D or $9E

[32shot]

I see, thanks a lot for your time clarifying all this and pointing me towards Stella's debugger!

 

I just had a quick look at the PAL ROM of Video Olympics and can see this stuff is stored in the same place in RAM as with the NTSC version.

 

So the situation being (using the only NTSC VCS i have here):

- My NTSC Video Olympics cartridge exhibits this glitch

- My PAL Video Olympics cartridge does *not* exhibit this glitch

 

Does this mean we're safe to assume the problem lies in the cartridge instead of the VCS?

[+SpiceWare]

12 minutes ago, 32shot said:

Does this mean we're safe to assume the problem lies in the cartridge instead of the VCS?

 

Yes, sounds like the NTSC cartridge is experiencing bitrot since the PAL cart works.

[32shot]

Amazing, Thanks again to all of you for all the help!

 

Will report back here as soon as that other cart turns up then.

[32shot]

... Thought i'd report back quickly on this...

 

Did get my hands on another NTSC Video Olympics cartridge and this one does work without any errors.

 

As SpiceWare had suggested earlier, the glitchy one is indeed showing signs of bit rot or generally faulty rom chip.

 

Thanks again everyone for the input, first time i'm seeing this on a cartridge of any kind actually!