Atari Cat Box, can we build an emulator ?

[Level42]

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&ih=011&sspagename=STRK%3AMEWA%3AIT&viewitem=&item=320099802086&rd=1&rd=1

That's a lot of money...

Basicaly it's not a teriibly hard device to "emulate". Schematics and ROM files are available. Shouldn't it be possible to emulate the entire thing on a computer and do the very last part in hardware only ? As far as I can understand, this device selects adresses then writes a value to that adress and reads it back to verify. Or am I over-simplifying things ?

It shouldn't be that hard to build the hardware using, f.i. a LPT1 port.

More info:
http://www.gamearchive.com/General/Data_Sheets/atari_cat_box/
http://www.gamearchive.com/Video_Games/Manufacturers/Atari/signatures/signature_howto.pdf

[Crax]

That would be awesome if there was some community built testing tool that people could take with them when going to get a cab.  I would love to have an item that can test a board/monitor as it's hard to know when you are buying something from someone that says "it's just a monitor problem" or "it's just a board problem, the monitor works great" if it's really true.

[ChadTower]

Tools like that exist.  You can take a pattern generator with you to test a monitor, or even just a compatible board to swap in.

As for testing boards, bring a small RGB monitor with an adapter that would take a signal from that board.

[Level42]

As for testing boards, bring a small RGB monitor with an adapter that would take a signal from that board.

Yeah, but that will only tell you IF the board is running or not. What we want to know (and what a Cat Box can do) is what part of the PCB is defective.

[ChadTower]

There are test rigs like that as well, though they're not portable and they are very specific to a board or set of boards.  Necessarily so, too, unless you wanted to write an app that was aware of the test routines of all boards and had an external interface that you could plug in.

Hrm, now that I think about it, that's a hella painful app to write, but could be possible.  Not feasible given the market but certainly possible.

[BobA]

If you know what you are looking for you can use a PC based scope and logic analyzer to do alot of what the specialized testers do.  It is all a matter of knowing how to apply the equipment to the specific board being tested.  You can catch specific events and look for specific data strings.  I do not think that the amount of work required to emulate one specific tester and design an interface would be warranted with what is commonly available in electronic test equipment.

[ChadTower]

I agree, specialized board testers are for testing large amounts of the same item in a short time.

[Fozzy The Bear]

Yeah, but that will only tell you IF the board is running or not. What we want to know (and what a Cat Box can do) is what part of the PCB is defective.

An ATARI Cat Box would be of very little use to anybody. They are very very specific to ATARI games from a very short period. That effectively means that you'd really have to specifically be an ATARI collector to want one. That you'd have to only be interested in ATARI games from a very specific period in the early eighties AND you'd have to be looking for those particular games.

The fact is that they really are not much use any more, even to an engineer. As a collectable arcade curiosity then yes buy one .... but to want to emulate a specific cat box like that really is a total waste of time.

Best Regards,
Julian (Fozzy The Bear)

[ChadTower]

An ATARI Cat Box would be of very little use to anybody.

I am primarily interested in early Atari cabs from around 78-84.   There are those of us who would have interest in this item.  Not enough to build one, but if it were within a price range I could afford for a test item, I'd sure buy one.

[horseboy]

I was really confused for a while here. I thought at first you guys were talking about the CatBox for the Jaguar.

[ChadTower]

Yeah, I did too, when the thread first started.  IIRC that wasn't produced by Atari, it was third party.

[Level42]

What Fozzy writes is only partialy true. Basicaly, this could be used on ANY arcade PCB. The thing missing is signature files for these games.
Now, if you'd built a "general" "Cat-box" (in-circuit tester) that is running on a computer people who have acces to working PCB's can make these signature files in a snap. Remember that a cat box didn't have any mass-storage device. If they would have, they could have created signature files of the full address range of each game. It's strange they didn't hook up an Atari 800 to do something like that...

And yes, I am mostly interested in late 70's, early 80's games...I think most of us are. I am not interested in fixing a HOTD4.

Anyway, the majority of PCB's from that age run on either a 6502 or Z80. So you would just need adapters for those processors.

I am not a hardware desinger, but I think the required hardware to do this is VERY limited. It would not cost the world.
The main thing is someone who is clever enough to write the software for it...

But aside from this all, I think it would be an interesting project. I may be a bit of a dreamer sometimes. I think it's a bit sad that an idea like this gets a "useless" rating right away....especialy on a creative forum like this  

[ChadTower]

It's a "how often will you use it" against "how much time will it take to build" issue.

[Fozzy The Bear]

Now, if you'd built a "general" "Cat-box" (in-circuit tester) that is running on a computer people who have acces to working PCB's can make these signature files in a snap.

That's very true.... But that wasn't the premise proposed. What was proposed was to emulate that specific ATARI Cat Box. In that respect my answer was accurate.

If you want a general Cat Box, then you can buy one off the shelf.... It's called an oscilloscope..... It'll do 100 times more than an ATARI cat Box would ever do and you can buy programable ones that can have specific chip and component signatures loaded into them. You can even buy PC based ones that just drop into a card slot. If you want to go this route, that would be a very good place to start.

Best Regards,
Julian (Fozzy The Bear)

[ChadTower]

I know what oscilloscope is, thanks... there are two of them on my test bench.

[Level42]

OK, I may have started off not clear enough saying what my goal is....I think it's clear now though

Yeah, I've got a scope here as well. A really old one but it works. I also have an Oszifox (http://www.wittigtechnologies.com/english/05_products/01_oszifox/oszifox_01_intro.htm) which is in fact a scope that runs on a PC.

However, I don't see why this would compare to a in-circuit tester. Regular scopes simply display what's going on on a certain pin....nothing more. The "in-circuit tester" would adress each available adress, read the "answer" from the PCB and compare it to the value (signature) that it has for a known working PCB, or the values that Atari specified for their games.
At least, that is what I understand from how it works, I may be wrong.

I've never heared of scopes that can do the component signatures, but I guess they can't do it while the part is still on the PCB and since almost all PCB's have the majority of IC's soldered directly, I don't see how this is of any use.

Apart from that, I can imagine that these are very expensive.

Just like "real" in-circuit testers

[werejag]

a device called a huntron tracker does exactly that. you can compare signatures of a good board(any baord) to a bad board to trouible shoot replacemant of parts.

[Fozzy The Bear]

I've never heared of scopes that can do the component signatures, but I guess they can't do it while the part is still on the PCB and since almost all PCB's have the majority of IC's soldered directly, I don't see how this is of any use.

Apart from that, I can imagine that these are very expensive.

Just like "real" in-circuit testers

There are at least three dedicated scopes that can do in circuit component analysis as well as measure waveforms and basic voltages. As I said.. The PC Card based ones would be the ideal starting point for creating a board or component specific one as they are programable.

Now all you need, is a software engineer to spend six months and $100,000 to write it. Personally I don't think it'd have a wide enough use or appeal for anybody to write as freeware. 99.9% of the people on this forum wouldn't want one, or know what to do with one if they had it.

Best Regards,
Julian (Fozzy The Bear) 

[ChadTower]

There's another approach, too.  Boards of this era are so easy to solder in sockets and have so few commonly failing parts.  Just socket chips and swap with new ones one by one until the board comes back to life.  Anything you actually replace is a positive anyway they're so old.  You can usually narrow the failure down to a range of possibilities and then swap away until you're good.