G3 -> PBX -> 68030 PDS = IIfxG3 . . .

[Trash80toHP_Mini]

. . . just wondering if this newest Bat$#!^CrazyNotion™ is impossible or merely as Highly Improbable & WILDLY IMPRACTICAL as most of my hunches. :?:

I came up with this impossible dream hack whilst toying around with the notion of hacking the 2300c/DuoDock's NuBus Chipset onto the PBX '030 Bridge ASIC on the 1400.

__ PPC603e works on a IIci PPC upgrade card, so there must be the equivalent of the PBX ASIC on board.

__CrescendoPB G3s work in place of the PPC603e CPU DaughterCard of the stock 1400.

so: Generic 68030 PDS -> PBX -> Connectors from my experimental DOA 1400 MoBo on a customized adapter card for my IIfx ought to be . . .

. . . an interesting proposition. }

p.s. what's good for the IIfx ought to be good for the IIsi . . .

. . . or the SE/30! MUAHAHAHAHAHAAAAAAAAAAAAAA!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! :lol:

 

[commodorejohn]

Hahaha, if you get a G3 working in an SE/30 I will have to send mine over for the same treatment :lol:

 

[ClassicHasClass]

The key is that the 603e works. If so (I've never tested that, so I will take you at your word), the 740 G3 (not the 750) is pin-compatible. I've even heard of a PowerBook 540 "G3" done in this manner, by hacking its 603e upgrade card.

 

[Trash80toHP_Mini]

I've poked around a bit and found piccies of a couple of 601 Upgrade Cards listed under NuBus Accelerators over on Applefritter.

http://www.applefritter.com/taxonomy/term/120,116

I haven't counted pins yet, but would these be PPC upgrade cards for the IIci PDS?

That 160 pin QFP ARM-ASIC from 1993 looks to be very much like the 1995 PBX ASIC from TI on the 2300c MoBo. PBX is a 240 pin QFP, which leads me to think that the data/address buses of the PPC601 and 603e may have differed (doubled?) during the intervening years.

Comments? I'm just spitballing here, no time for any serious research ATM.

How's my guesswork so far?

 

[Bunsen]

I thought PPC upgrades were only ever made for 040 machines?

 

[Trash80toHP_Mini]

IIRC, because the IIci was the de facto standard workstation of the DTP Industry, Apple was pretty much forced to make a PPC upgrade card for it. That was a HUGE user base and Apple had already had a nasty legal issue, spear-headed by them due to the Dirty ROM issue regarding their pre- IIci workstations. I wasn't aware that Apple made any PPC upgrades for '040 systems. I was under the impression that such beasties would have all been second source products.

Now I'm wondering if the cards in the 'fritter pics were both IIci cards? :-/

At any rate, one reason hit me as to why the ARM 680X0 (?) Bridge ASIC in the pics is only 160 pins. System RAM was located on the host MoBo, with only SRAM L2 Cache on the PPC Card. On the PBX 68030 Bus Bridge, System and L2 Cache are both on the Fast I/O Bus. Dunno . . .

. . . any other plausible reasons for the Bridge ASIC to grow in pin count? :?:

 

[ClassicHasClass]

There are definitely IIci PPCs. I even have an industrious user who claims to be running Classilla on a Sonnet PPC 601 card in his IIci with 8.6. He described it to me as "pokey but useable."

 

[Trash80toHP_Mini]

That's what made this WAG click.

PBX Bridged systems are pretty much maxed out IIci Architecture Cookie with PPC603e Chocolate Sprinkles on the Vanilla '030 PBX Bridge Icing.

YUM! )

 

[Gorgonops]

... I was going to say something, but I'm realizing I can't quite interpret Trash's idea. Is the question whether it would be theoretically possible to make a "G3 upgrade" for, uhm, a IIfx?, completely from scratch, or whether it would be possible to literally carve some ICs off a Powerbook motherboard and use them to tinker something together? My head hurts.

 

[Trash80toHP_Mini]

:lol: I just love it when the voice of reason chimes in on the subject of one of my WAGs. I really miss Dr. Bob, thanks for being there for me, EudiG!

Think really incredible cobbling together of off the "PCB Shelf" Hardware and an OS hack on a patch on a kluge that rivals the level of MSWindows . . .

. . . anytime after the last decent total overhaul of their spaghetti code . . . Win98. :

Ain't nuthin' theoretical popping up in my noggin' that winds up posted here, it's do it or di . . . erm . . . document a fail of spectacular proportions!

The first thing is to call Sonnet Tech Support and see if they have a good laugh with me yet again! )

Questions:

___Will a 6360/6500 Cache Slot Crescendo boot up without the driver like the CrescendoPB does if installed on the PBX Chipset with no Proc installed on the host system?

___Can the driver be hacked to work on the Mac II series when installed on a kluged PBX ChipSet Cache Card?

LOL! }

 

[trag]

The Turbo601 for the IIci only ever used a PPC601 chip. There was no PPC603(e) version. There was no PPC603 based upgrade for 680x0 desktop machines.

As far as I remember, no one but Daystar ever made a PPC upgrade for the IIci, so I suspect, but am not certain, that Cameron's user is misremembering and has a Daystar upgrade, not a Sonnet upgrade.

Although, I think there was a rumor than Sonnet (or someone else?) might pick up Daystar's tooling and do another run of Turbo601 upgrades after Daystar sunk. As I remember, by that time PPC601 chips were out of production and demand was dubious, so I'm certain that never happened.

Not entirely certain how this applies to JT's brain-cyclones as I too have trouble sensibly interpreting them, but that's what I know, for what it's worth.

I had a Turbo601/66 in my IIci back in the day. It was cool the way a dancing bear is cool. It's not that it dances so gracefully, it's that a bear is dancing at all. But there were many compatibility eccentricities with the type of high-end upgrades a person who would be inclined to buy a Turbo601 is likely to already have, and it really just wasn't that much faster. I decided that a nice Turbo040 is the way to go in the IIci.

If you must have PPC and NuBus slots, get one of the NuBus PowerMacs, or the very cool Power Computing Inc. Power 120.

 

[Trash80toHP_Mini]

In the OP, I was trying to get the post out quickly, so I used 603e as a generic term for early PPC. It also happens to be the Proc the PBX '030 Bus Bridge ASIC was designed to support for the NuBus Architecture PowerBooks. I've been studying that bit of Macontosh Architecture on and off for about 12 years now. I've pretty much given up on the notion of adapting a CrescendoPB-G3 to fit inside the Duo2300c . . . so I'm adapting the notion to doing a 128k/Plus using a 180c/270c LCD display with a DuoDock and two NuBus slots/cards inside.

This way I've got lots of room for adaptation of one G3 card or another to the 3200c MoBo.

As a side project, I'm going to be breaking the 1400c's PBX lines to the NuBus Chipset and adding, hopefully, three slots. This hack will be encased by the plastics of a 12" RGB Monitor.

I've tested both LCD/CRT size comptabilities and I'm really looking forward to doing some PCB work for the hacks.

What are the main differences between the 601 & 603?

 

[Bunsen]

The Turbo601 for the IIci only ever used a PPC601 chip.

IIci --> Turbo601 --> kludged 601 PDS --> Sonnet Crescendo Nubus?

603e / the Proc the PBX '030 Bus Bridge ASIC was designed to support

IIci (or IIsi / SE/30) PDS --> kludged adapter (PBX + 1400 CPU socket) --> Powerbook 603e / G3 upgrade?

breaking the 1400c's PBX lines to the NuBus Chipset and adding, hopefully, three slots

8-o

 

[Trash80toHP_Mini]

The Turbo601 for the IIci only ever used a PPC601 chip.

IIci --> Turbo601 --> kludged 601 PDS --> Sonnet Crescendo Nubus?

That'd be IIci PDS -> generic (?) '030 pinout adapter/PGA Socket -> PGA Male PBX Adapter Card/PGA Socket -> PGA Male Adapter for . . . whatever might work! :approve:

Ditto for the IIfx's non-standard '030 PDS.

That's an interesting notion about the Crescendo NuBus, that's new-n-novel, in the case of hacking support for the HPV Card!

IIci (or IIsi / SE/30) PDS --> kludged adapter (PBX + 1400 CPU socket) --> Powerbook 603e / G3 upgrade?

That's one of my lines of inquiry, testing will be simple, just make it work for a 1400 603e ProcCard and the CrescendoPB will work.

A 166MHz PPC 603e/SRAM Cache (?) is a lot easier to come by than a CrescendoPB and an excellent kick in the pants for any '030 PDS Mac in its own right! [] ]'>

breaking the 1400c's PBX lines to the NuBus Chipset and adding, hopefully, three slots

8-o

:?: That's the obvious one!

Lo these many years ago, I proved to Dr. Bob, that the 1400 and 2300 are, for all impractical purposes, the same computer. More accurately, he was shouted down by the rest of the 'fritter crew when I posted the two block diagrams side by side and color coded for matches, differences and missing blocks.

The 1400 and 2300 were developed and released in tandem, along with the first PCI architecture PB, the 3400. A few modules were swapped around, upgraded for the 1400's 800 x 600 LCD and the IRDA port hardware was tacked onto the T-REX Daughtercard, which was slated to be tacked onto the Duo's Docking Connector and housed in the base of the UltraDock.

The Duo's '030 PDS Docking Adapter was (assumedly) necked down to 16 Bit for the T-REX PCMCIA Card Cage, though if the IRDA circuitry is on the other 16 Bits, this NuBus Chipset graft will be a piece of cake! A minimum of half of the '030 PDS portion of the Duo's Docking Connector exists on the T-REX Daughtercard's Header interconnect, the other half probably leads to the Video/EtherNet Daughtercard Slot, if not the whole thing. I never looked into that or noodled out the entire T-REX Card Cage PCB, but I've got most of the layers documented in an Illustrator file along with template pics of the progressively ground down PCB layers.

Not entirely certain how this applies to JT's brain-cyclones as I too have trouble sensibly interpreting them . . .

Lovely imagery, very fitting, I like it! [] ]'>

Basically, every PBX based permutation has been flipping about in the back of my noggin' for a very long time. These just slipped out into .TXT here because PBX suddenly wound up overlaid on the Mac II series 68030 PDS as I was daydreaming about my IIsi/NuBus+PDS Cards Hack . . . with the 1400 PBX/NuBus Chipset sketch in front of me at work.

Two brainstorm fronts collided! )

 

[ClassicHasClass]

What are the main differences between the 601 & 603?

Well, the pinouts are different, for one thing, and the 601 was intended as a bridge processor so it implements POWER instructions that the 603 and up do not (and, in a like manner, the 603 implements portions of the PowerPC specification that the 601 does not).

However, as long as the OS knows about it, it shouldn't be a problem if you can solve the pin differences.

 

[Trash80toHP_Mini]

NoPro, PBX and 603e/G3 are the only hacks under consideration, the IIci 601 FudgeBridge is a non-starter.

 

[Trash80toHP_Mini]

I just took a more leisurely look at the 601 card pics posted on 'fritter. From the connectors' appearance and the four rows of soldertails, they seem to be 68040 PDS Connectors, not the EuroDIN connector types common to the Macintosh II series, which have three soldertail rows.

Has anyone got pictures of the IIci PPC Upgrade card . . . or the actual card that you might take some pictures of to post? :?:

 

[Trash80toHP_Mini]

I'm liking Bunsen'suggestion of hacking a DuoDock onto PBX on the 1400. A DeclROMless DuoDock board would give me two NuBus slots without making a three slot NuBus PCB from scratch. It's a great way to concept test this kluge.

I wonder if two NuBus Chipsets or one (or both) from a six slot MoBo would work? }