MicroQuadra "Evil Memory Hack" Redux

by **Trash80toHP_Mini** » 01 May 2011, 13:01

Somebody did a very "Low Rent" version of one back in the day. It was done using wires and SIMM Sockets clustered and flopping about on the existing socket. Dr. Bob over on 'fritter called it something like: the most evil thing he'd ever seen.

I've always wondered if the Q605 & LC475 (among others) could benefit from a new approach to memory expansion. I nabbed an extra MicroQuadraFeetsMac™ on eBay and want to try to boost the memory with a sub-basement memory card hack, especially so now that we've got access to the requisite SIMM Sockets thanks to olePigeon.

My plate is over-full doing research for several other hacks ATM, so I thought I'd throw this one out there for the rest of the MLA to mull over.

Here's the concept:

It's basically a SIMM Saver Card done in reverse., funneling the existing Memory Bus and all available bank select lines to a card that adapts the memory bus by splaying it out/funneling it into a SIMM or DIMM Socket(or sockets) of a higher capacity generation of memory than was available on any given Mac. This would include stealing the bank select line from the pitiful allotment of MoBo RAM hobbling every compact mac, LC & Perfoma ever equipped with this BS.

Hardware Development:

Solder "Machine Pins" onto the bottom of the existing SIMM Socket that will mate up with the same pins installed on the sub-basement "Reverse SIMM Saver" Daughtercard.

Develop/adapt a "Virtual Memory" init/whatever to allow the Mac to address Memory FAR ABOVE the meager amount provided for by the memory mapping supported by System ROM. This was commonly done on SIMM equipped SE (and other) Accelerator Cards back in the day and shouldn't be difficult to reverse engineer/hack by one of our Low Level Programming Capable Comrades.

Research to be done:

Post links here to . . .
. . . original hack, in the threads here and over on 'fritter. < where are you Danamania? [}:)]

>
. . . any discussions on "expanding the memory envelope" of any Mac any of you folks are interested in MemHacking™

The we can interactively come up with an interesting solution.

BTW, the same thing should be possible to do on an actual SIMM Card designed for our beloved IIfx playtoys!

Personally, adapting this hack to my pet IIsi was my original inspiration, but sharing the notion up front (and the work! [}:)]

) seems like the best course of action.
I work best in collaboration and procrastinate best when left to my own devices! ::)

p.s. OOPS! wrong forum! gotta move this one to hacks.

by **Bunsen** » 02 May 2011, 00:37

Intriguing. My guess is that trag would be your natural ally in this endeavour, with his IIfx SIMM experience. On that topic: a 64-pin SIMM adapter that lets us use 30 or 72 pin RAM in those puppies would be a boon, indeed. AFAIR, once you get past 72 pins, you run into weirdness on board the SIMM/DIMMs themselves that complicates things, but no doubt trag would have (or has in the past had) more to say on that than I could usefully add here. I also STR there was more to the 64 pin DIMMs than a mere change of pin count.

by **trag** » 02 May 2011, 21:24

Bunsen wrote:Intriguing. My guess is that trag would be your natural ally in this endeavour, with his IIfx SIMM experience. On that topic: a 64-pin SIMM adapter that lets us use 30 or 72 pin RAM in those puppies would be a boon, indeed.

I'm coaching baseball this season and too busy to do much thinking, much less hacking. However, I did post a rather long explanation (on AppleFritter, I think) about why there's no point in trying to adapt other memory to the IIfx.

The short version is that the vast majority of memory uses bidirectional data pins (they're both DATA_IN and DATA_OUT) whereas the IIFX has separate pins for In and Out on the Data bus and there's a very good reason why it needs them. So there's no way to adapt SIMMs which have a bidirectional data bus to use in the IIfx short of measures which amount to desoldering the memory chips and putting them on a different board.

Or building an intervening circuit board which does some tricky conversion.

And if you're going to do that, why not just build IIFX SIMMs in the first place?

by **Trash80toHP_Mini** » 02 May 2011, 23:43

IIfx SIMMs are just a side issue, but don't VRAM ICs work similarly to the IIfx's Dual Ported (?) DRAM ICs? Dunno . . . :-/

The real reason for this thread is to expand Apple imposed Q605/LC475 and IIsi memory limitations by building an adapter card that'll use more modern (faster response times) SIMMS on a sub-mobo memory expansion card that'll address all available RAM Banks at maximum capacity.

The Q630 and its variants could also benefit from such a board, as well as almost ANY Mac with a piddling 4 MB of onboard RAM.

I'll have to check my archives for the original "EVIL MemHack" info/threads. I was hoping the maniac still had that info available, IIRC, she's the one who posted the links.

. . . or that some eager comrade would jump in, search them out, and post them! [;)]

by **trag** » 03 May 2011, 21:22

Trash80toHP_Mini wrote:The real reason for this thread is to expand Apple imposed Q605/LC475 and IIsi memory limitations by building an adapter card that'll use more modern (faster response times) SIMMS on a sub-mobo memory expansion card that'll address all available RAM Banks at maximum capacity.]

That would be interesting. The Q605 definitely does at least up to 260 MB of RAM. I know that from experience. And the 4MB on board can probably be replaced with 64MB, so I imagine that machine could go to 256 + 64 = 320MB at a minimum, perhaps more.

The IIsi should go to 128MB. My understanding is that it's essentially a IIci with a microphone jack, so the memory pins should be there for 128MB.

by **Trash80toHP_Mini** » 03 May 2011, 23:56

Thanks, comrade trag! [:D]

I had a notion that the gang might find this one interesting! [;)]

I wasn't about to do all the work on all applicable macs for this relatively simple hack anytime soon . . .
. . . at least not for anything but my pet IIsi! }:)

. . . so I tossed it out here for the rest of the MLA to get involved with development. I'd like to see us do a group project and order a bunch of boards to get the costs down after the proof of concept ProtoBoards are fabbed, debugged, up and running! [:D]

For the IIfx, I was actually thinking about its PDS, which'd make a right fine place for a RAM Disk! Even though the virtual memory addressed "Silicon Scratch Disk" would run at only half the proc's clock . . .
. . . that should still be just a tad faster than any HDD accessed VM! :approve: