[NTLK] Armstice board (was: Unused Messagepad)
stepwhite at gmail.com
Tue Feb 28 16:07:26 EST 2017
The ARMageddon board is from 1991, so it predates the Junior project completely. It is effectively just an ARM chip and a gob of RAM. This one is definitely not a “Newton on a board”.
The ARMistice board is from 1993, and is effectively a Newton (OMP) on a board, minus LCD. On the rear you’ll see the Zarlink P610ARMKG ARM chip, Zilog SCC, the Runt VLSI, etc. Front has the serial connectors, PCMCIA slot, etc.
Both of them have a handful of NuBus interfacing chips, and 4 FPGAs. The FPGAs are clearly labelled on the ARMistice, “DEBUGGER”, “COMM MANAGER”, etc.
I presume that the “DEBUGGER” helps with breakpoints + watchpoints. I suspect the bus manager helps intercept regions of the address space to re-route into the Mac. As an example of this, the ARMageddon can apparently run the same Newton builds that the ARMistice does. But it doesn’t have all of the requisite hardware. If you attempt to use Egger with an ARMageddon board, Egger will complain that you need to launch a separate Macintosh application, aptly named Armageddon. I would suspect this application emulates the Runt VLSI chip inside Macintosh 68k code. A suspicion, because I haven’t been able to track down this application yet. There is a certain Mo in MTV that thinks he can help with this (nudge nudge).
As further evidence, the ROM image that works with Egger and the ARMistice board has references to “:ScreenDrivers:ArmisticeDriver.c”, and through the Leibniz emulator (with the appropriate gDebugger bits set) I can observe it writing to what is normally invalid address space (0x30000000, IIRC) and complaining about “shared ram failure”. Presumably this is an easier way for it to draw the Newton’s screen inside a Mac window (easier than emulating the Sharp LCD crud the Runt would emit).
I have it on very good authority there were other full length ARM boards used for the Newton development, however I haven’t seen them. But a close friend who worked with Apple on the Newton, from late 1991 onwards, vividly remembers a board named “ARMadillo”, and can even vaguely remember what it’s splash screen looked like. To illustrate why these splash screens would be memorable, this is the one from the ARMageddon board:
As Andrei mentioned, Egger (and later on Hammer) were used to drive these boards. This is a screenshot of Egger running a 629.00AS Newton ROM image on an ARMistice board (Macintosh Iici, System 7.5):
And that is what I tried to preserve with the Leibniz emulator written about earlier. Andrei, I can add the clank sound for DebugStr dialogs if that’ll make you smile.
Both boards contain 4 rows of header pins (~ 32 pins per row, so presumably one row for address, one row for data, and the other two hooking up to the FPGAs??). There are photos of the “ARMageddon Fat Man Board”, which seems to utilize them to attach a daughter card to the ARMageddon. I wish the CHM had higher quality images, but it can be seen here:
Perhaps Steve Capps remembers what the add-on was for, and will be able to talk about it later this week. If there’s a Q&A session and some freak asks about Ralph and/or NuBus cards, you’ll know I decided to brave the traffic to get there.
Additionally there is the smaller “Universal Newton Adapter”, which was a conduit to external Newton hardware. It doesn’t include an ARM or anything fun. It is absolutely dependent on an external Newton device.
A ribbon cable would run from this card, connecting to an external device. Presumably a 50-pin SCSI cable, since the floors of Apple campuses were probably littered with them in the early 90s. The external devices I’m aware of are the “Bunwarmers” and “ROMulators”.
And with the top cover removed to help show the ribbon cable connector:
A ROMulator for a Bic:
ROMulators existed for other devices, but I don’t recall having ever seen pictures of any. I’ve definitely seen references to them though, e.g. "Q Romulator Detected”
I am friends with the owner of that Bunwarmer, and I own that Universal Newton Adapter. It is our mutual goal to pair the two and see the Bunwarmer run again (for the first time in 20+ years). Just a matter of getting our free time to sync up, which is exceedingly difficult…
> On Feb 28, 2017, at 12:08 , Grant Hutchinson <grant at splorp.com> wrote:
> On 2017-02-28, at 10:54 AM, Andrei Chichak wrote:
>> I can’t find a picture of the board, but imagine a computer expansion card with a lonely ARM processor, a huge array of memory, and I seem to remember a huge LSI logic chip. Forgive me if I got it wrong, that was 1993.
> Here are Steve's photos of his ARMistice board.
> https://www.flickr.com/photos/pablo_marx/4683061782/ (front)
> https://www.flickr.com/photos/pablo_marx/4683061958/ (back)
> And the ARMageddon board:
> https://www.flickr.com/photos/pablo_marx/4683061364/ (front)
> https://www.flickr.com/photos/pablo_marx/4683061542/ (back)
> In fact, the whole album is chock full of Newton developer wonderfulness:
More information about the NewtonTalk