Mac 128K Strange Upgrade

[Mac128]

Aside from the incredibly reasonable price for which this 128K sold on eBay, this 128K is notable for the rather bizarre FrankenMac upgrade to the logic board. This is unlike any mod I have seen to the original logic board for a 512K upgrade. Perhaps there is more going on elsewhere around the board the seller just failed to photograph?

 

[~Coxy]

That looks a lot like the mod to increase the size of the RAMs on the Mac mobo by "piggybacking" them, except as you said that only one looks to have been done in this way!

Perhaps the original RAM on the bottom died prematurely and it was "replaced" with another of the same size using the piggyback method?

 

[tomlee59]

The chips in question aren't DRAM chips. My guess is that the mod is indeed directed toward enabling more RAM, but it's in the address decoder circuitry.

Or not.

 

[beachycove]

Early prototype?

 

[Mac128]

The chips in question aren't DRAM chips. My guess is that the mod is indeed directed toward enabling more RAM, but it's in the address decoder circuitry.

Tom, it definitely appears to be some kind of multiplexer. The RAM chips in the picture are indeed upgraded 256 on a stock 128K board (though why they did not socket them I have no idea). However, this is one of the simplest, if not cleanest, upgrades I've seen with a few wire jumpers going to the decoding circuitry and a matching chip piggy-backed on top of an existing chip (no higher than a socketed chip). All the other solutions I've seen involve a custom mini-mux board (some nicer than others) mounted in some free space on the LB, usually sticking up higher than anything else.

This brings up an interesting question for me. With consumers demanding more RAM within weeks of the 128K's launch, why did Apple take the time to redesign the 128K board, test and fabricate a new hybrid board? Why not just design a custom IC multiplexer chip to plug into those E3 holes on the board? Was anything else changed about the 128K/512K board over the original that would constitute a "necessary" improvement? As far as I know the 512K board is no faster or more efficient than 512K upgraded 128K boards. It's not like the new board was harder to upgrade to 512K (in fact it was easier), so OEM upgrades would have still been a profitable path for Apple. Even if Apple wanted to redesign the board at some point, a simple chip would have put the 512K into the market as soon as sales began to slump, or sooner if they wanted to.

 

[tomlee59]

I agree -- that is indeed a clean and clever upgrade, and one I'd never seen before. As to your excellent question about why Apple did what they did (and didn't do what would've made more sense), perhaps it's all an artifact of Steve Jobs' opposition to expansion of any kind in the first place. The design team had tried to sneak in an expansion port disguised as a diagnostic port, only to have Jobs order it taken out. Perhaps there was a similar battle (again lost by the team) with respect to RAM upgrades? As you note, only a trivial address-decoding change is needed to work with the higher-capacity DRAM, so upgrading is truly simple. Very curious, indeed.