Thread
Ethernet MicroDock - any chance of repair?
I certainly don't know, but there is a lot going on in general and within the TREX ASIC on the bottom of the 1400 Card Cage's PCB, having added BootROM, driver support, etc. would be trivial.
If the 1400's Card Cage hadn't been heading into the Duo PCMCIA UltraDock, it would have made a lot more sense for Apple to have put all the goodies on the MoBo like in the previous incarnations of the 190 and 5300. That sub-assembly adds a significant manufacturing cost increase and a reliability hit the way they tacked it onto empty real estate on the MoBo of the 1400 with headers like that.
If the 1400's Card Cage hadn't been heading into the Duo PCMCIA UltraDock, it would have made a lot more sense for Apple to have put all the goodies on the MoBo like in the previous incarnations of the 190 and 5300. That sub-assembly adds a significant manufacturing cost increase and a reliability hit the way they tacked it onto empty real estate on the MoBo of the 1400 with headers like that.
I've noticed you seem to attribute a lot of magical powers to that T-REX ASIC... maybe I'm missing something, but from the functional description in the various devnotes it looks to me like it's just a PCMCIA socket controller, roughly equivalent to something like the Vadam VG-469. (Which is a register compatible clone of the PCMCIA support included in the 82365SL chipset.) While I'll grant that PCMCIA controllers *are* complex beasts (Look up the datasheet for one to see why; PCMCIA power management is complicated, and the controller is also required to contain a memory mapper that's essentially the same as that used in EMS/LIM 4.0 ISA expanded memory cards; that's why PCMCIA can have a 64 MB address range for SRAM/Linear Flash cards despite ISA being limited to 16MB. There's also some weird magic they can do with DMA, like "Zoomed Video".) it looks like the T-REX is a pretty bog-standard one with the one exception that its "outside" bus is that 68030 PDS subset that Apple used to hang peripherals on internally in all their pre-PCI Powerbooks instead of an ISA bus.
(The internal arrangement of all those "PDX" based Powerbooks is very similar to your typical 386 or 486 PC computer; the CPU's native "fast" bus is used for little more than RAM while every other peripheral sits on the other side of an ISA bridge. Considering the needless complexity of NuBus for internal peripherals it's not surprising that Apple kept cooking up various flavors of "PDS" busses, which of course were no longer "Processor direct busses" when a next generation CPU necessitated the use of a bridge chips like the PDX. Various subsets of the 68030 bus are essentially "Apple ISA", minus the "Standard" part. No wonder Macs were so ridiculously expensive...)
Anyway. Unless it does some *really* magical things (which the guy who wrote the T-REX PCMCIA driver for PPC Nubus-Linux didn't notice as he was reverse-engineering how to drive it from the practically non-existent documentation), it's going to need software to drive it, and I don't see why that software would already be present in a desktop system's (coincidentally or not, usually smaller) ROMs. So it would seem to follow that if you're going to try to cram one of those card cages into something with a possibly-compatible-with-the-PBX-defined-68030-PDS-bus' slot you're going to have to figure out which software pieces have to come along for the ride, and unless the card cage has a ROM on it (in addition to the T-REX chip) it's safe to assume those pieces are in the main ROM on the PCMCIA-equipped Powerbooks. The exception there is of course possibly the removable module that actually shipped for the 500 series, in which case it would be interesting to know the size and contents of the ROM that might be in the expansion module.
... I suppose it goes without saying that this is probably a really odd and "around the horn" way to get IDE in a non-IDE-equipped Mac if that's all you intend to use the PCMCIA slot for. 
I'd love to get my hands on a PCMCIA cage for the PB 500 series so we could see what chips are in it.
Therein lie the mystic powers of TREX! :approve:
I've been fascinated with T-REX since I did the block diagram comparison graphic to prove to DR. Bob that TREX was the only missing link for turning a Duo 2300c into a 640x480/48MB PB1400 sans expansion bay over on 'fritter in the 2300c/G3 thread, lo these many years ago.
Silent IDE boot and WiFi for the IIfx would be the holy grail of the kingdom of TREX . . . it'll never happen . . .
. . . but the siren song of the HDD replacement PCMCIA card cage in the Duo 2300c hack remains irresistible. }
Didn't somebody say it had a 68000 in it? If that's the case, the other parts will be boring glue logic, buffers and such and the real magic will be in the firmware.
Good point. I remember someone saying something along those lines, now that you mention it.
With the simple purchase of a SCSI->IDE->flash adapter combo and one of those ethernet->wifi bridges all your dreams could come true.
But what fun would that be? :lol:
Or they could have just cooked up their own bus from the beginning instead of using Nubus, a difficult and expensive mezzanine bus for both Apple to implement in their machines and for card makers to develop. It worked for Commodore. (see Zorro/AutoConfig, basically the same concept as PDS and ISA, even came in 32bit form).
We won't get into what Apple's engineers were thinking when they developed A/ROSE to "simplify" development of expansion cards.
The IIfx has a PDS slot... that would be a good place to start. It might be easier to develop a card for that as opposed to NuBus!Silent IDE boot and WiFi for the IIfx would be the holy grail of the kingdom of TREX . . . it'll never happen . . .. . . but the siren song of the HDD replacement PCMCIA card cage in the Duo 2300c hack remains irresistible. }
Great point about the IIfx PDS, I hadn't thought that through. I just figured that the IIfx might be able to run the lowest OS compatible with PCMCIA support. If so, it would clearly be the most cool to hack, given its wicked fast mythos.
I'd pretty much given up on the NuBus angle due to paucity of information in DCaDftMF3e, but I've now got the two earlier versions of Designing Cards & Drivers and it appears as though the information on the design example cards is a lot more complete than I'd expected, still far over my head, but fun to explore with a project like this in mind.
@G: the PBX bridge has been the most fascinating thing for me about the NuBus architecture PowerBooks and my personal favorite, the Duo 2300c. The "Pratt IC" in the PowerBook 190 does the same thing, but bridges the 68040 to what's hopefully the same 68030 compatible "slow I/O bus" of the PBX series.
Not that I'd consider this project, but given the Pratt IC interface, the Quadra 700 would be the obvious choice due to its sexiest Mac mythos. The incompatible 840AV PDS removes it from the competition.
The oddball PDS of the IIfx is much more alluring, mostly because not much has been done with it and the rest of the 68030 PDS slots are incompatible with it and each other anyway . . . stupid Apple.
I'd pretty much given up on the NuBus angle due to paucity of information in DCaDftMF3e, but I've now got the two earlier versions of Designing Cards & Drivers and it appears as though the information on the design example cards is a lot more complete than I'd expected, still far over my head, but fun to explore with a project like this in mind.
@G: the PBX bridge has been the most fascinating thing for me about the NuBus architecture PowerBooks and my personal favorite, the Duo 2300c. The "Pratt IC" in the PowerBook 190 does the same thing, but bridges the 68040 to what's hopefully the same 68030 compatible "slow I/O bus" of the PBX series.
Not that I'd consider this project, but given the Pratt IC interface, the Quadra 700 would be the obvious choice due to its sexiest Mac mythos. The incompatible 840AV PDS removes it from the competition.
The oddball PDS of the IIfx is much more alluring, mostly because not much has been done with it and the rest of the 68030 PDS slots are incompatible with it and each other anyway . . . stupid Apple.
A pre-emptive multitasking, memory protected nanokernel running on a 68000 in 1988.
Ah, what might have been ... *sigh*