What are these ROMs? Pac-Man cartridge?

[6BQ5]

My 800 system came with some cartridges from the seller. One of the cartridges, Pac-Man, doesn't work. Opening it up I see one of the ROM chips is cracked. I also noticed the chips have different part numbers and the packaging/casing looks different.

 

A picture is attached.

 

Are these really a matching set of ROM chips for Pac-Man? Where can I find a replacement?

 

Thanks!

[+Nezgar]

Wow, never seen a chip cracked like that before... Yes those part #'s match the chips in my Pac Man PCB. Interestingly, the styles/production runs of mine are basically inverse of yours.. Doesn't mean any functional difference though.

 

PCB C012963 is an 8K board, takes 2 2332 4k x 8 bit 24 pin mask ROMs. (or 1 2364 type mask ROM, but you'll probably see 2 of those in 16KB PCB's...)

 

You could inquire about direct replacements of the mask ROMs from the likes of BEST or B&C. Otherwise the easiest may be just ordering an 8K EPROM cart PCB for the brown shell cases from those sellers (about $5) and putting 2 programmed 2732, 2532, or a single 2764 in depending on pcb...

 

I think a 2532 could replace one of the two chips with no soldering, but no telling WHICH without trial and error, depending which has matching OE/CS function..

 

Here is one page showing how to manually re-wire it to directly take a 2764 EPROM:

https://www.insentricity.com/a.cl/200/putting-a-custom-eprom-on-an-atari-800-cartridge

 

Or, simpler would be using a 2364 PCB adapter from the likes of Retro Innovations, but it may not fit inside the shell... Maybe it would by forgoing sockets, or cut a hole in the back of the cart for the chip to protrude. You could have some fun with it, like put a switch for up to 2, 4, or 8 different 8K games on a single EPROM depending on the chip size...

http://store.go4retro.com/2364-adapter/

 

I killed a 16KB E.T. Cartridge of mine with an inadvertent static zap, so I've been considering what to do that PCB as well...

 

Edit: Oops Pac Man is 8K, not 16K

 

Edit2: You're lucky you have sockets unlike on my PCB. You can try extracting the chips and swapping them to the opposite sockets. Atari's 2332's ROM chip selects are handled in the chips, not by the chip location on the PCB so location is interchangable. Also re-flow the solder on the socket connections to the PCB...spray a little Isopropyl alcohol into the sockets before re-inserting the chips to help clean and ensure good contact on insertion...

[Rybags]

I believe Pacman size is dependent on the version - the one with animated interludes being rarer and 16K.

 

Possibly the crack allows air in which degrades the chip? Or are mask Roms and OTP Eproms laquer covered like some modern CPUs?

[Forrest]

Was it common to have socketed ROMs in production cartridges?

[+DrVenkman]

Was it common to have socketed ROMs in production cartridges?

In early brown metal carts, sure. Same reason why every 400, 800 and 2600 board was fully socketed during the Warner era. It wasn’t until mid/late 800XL production, and introduction of the 2600 Vader variant that mainboards started to become soldered rather than socketed.

 

A few cents per socket, especially in mass produced cartridges, starts to add up.

[leech]

In early brown metal carts, sure. Same reason why every 400, 800 and 2600 board was fully socketed during the Warner era. It wasn’t until mid/late 800XL production, and introduction of the 2600 Vader variant that mainboards started to become soldered rather than socketed.

A few cents per socket, especially in mass produced cartridges, starts to add up.

I always hear the savings of money being the reason that later models were not socketed, but I wonder if it is due to automation as well. I mean if the process went from being a by hand assembly line to just having a little robot drop the chip in a board and solder, it would be far easier and less expensive to skip the socket step. Granted I have no idea how much of the Atari stuff was done by hand. I worked for a very brief time at a company that made the USRobotics modems. Part of them were still soldered by hand.

[+DrVenkman]

I always hear the savings of money being the reason that later models were not socketed, but I wonder if it is due to automation as well. I mean if the process went from being a by hand assembly line to just having a little robot drop the chip in a board and solder, it would be far easier and less expensive to skip the socket step. Granted I have no idea how much of the Atari stuff was done by hand. I worked for a very brief time at a company that made the USRobotics modems. Part of them were still soldered by hand.

Even in the 70s, they were highly-automated. All the discrete components were machine-placed and soldered, including the sockets. To me it seems very clear that the move from fully-socketed everything in 1977 with the VCS to fully-soldered everything with the XE machines by 1985 was driven by economics and changing markets. In the 70s, things were expected to be used for years and might require repair. Something broke, you paid a repairman to fix it and kept using it. You paid retail prices that supported that expectation too. By 85, price was king. Price wars had driven several companies out of home computers entirely, videogame companies had risen and died already, and stuff was much more disposable in nature - the expectation was that when an item died, it was either cheaper just to send a new unit for a warranty return, or the consumer would just buy the next cheap (but better) thing. So no need to build for ease-of-service. Manufacturing costs had to drop in response to the ever-present crush and consumer expectation that prices would come down over time. So no more sockets was an easy way to save a chunk of change over the course of a fiscal year.

Edited May 10, 2019 by DrVenkman

[6BQ5]

Thanks, everyone! I appreciated the replies!

 

Yes, the crack is very strange. I'm not sure how it could have happened. Maybe the cartridge experienced some sort of jolt by falling on its side or was crushed while laying on its side. Perhaps the chip was faulty from the factory from the beginning and a continuous cycle of hot/cold/hot/cold in storage cause the material to crack. No matter how the crack appeared the end result is the same. The chip and cartridge are basically dead.

 

Just for fun, after I typed my initial message, I tried playing musical chips. What if the ROMs were actually two separate games? They do look very different, right? I tried every combination of location and chip order. Nothing worked. No big deal.

 

So I guess I have a blank cartridge board on my hands ... socketed, blank, cartridge board.

 

Honestly, I'm kind of intrigued about the idea of replacing the ROMs with some other game but without modifying the board or using adapters. Hmmm!

 

-------

 

Side note : I received the game Embargo from another seller. The cartridge didn't seem to fire up. Last night, after playing around with the Pac Man cartridge, I cracked open the case for Embargo. Inside was a very simple PCB with one ROM. I tried inserting the PCB and it didn't work. Then, for some reason, I got the idea to insert it backwards. Guess what? It fired right up! Someone at the factory had installed the PCB backwards!

[Rybags]

2 chips from seperate games won't work. Swapping 2 chips positions for one game won't either.

 

The last 6 bytes of the Rom have the run address and flags and if they're wrong the system won't even attempt to start the cartridge.

 

With 2 chip carts the usual order is low/high Rom with one appearing in the first 4 or 8K and the other chip occupying the last 4 or 8K.

In theory a game might partly initialize but then crash. Though many games have the startup code towards the start of the Rom so won't even get that far.

 

As for putting boards in back to front - not sure if that's much good for them, would need to check the pinout.

[leech]

If it's just Pac-man, maybe see if you can find the actual ROMS and if any EEPROM will work for them, could even put in the rarer version, right? Hopefully it's dumped somewhere.