ktkm
Well-known member
I agree!
And just like @CC_333, I have longed for one for quite some time now.
-
Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.
SE/30 Xceed greyscale adapter cloning thread
- Thread starter pcamen
- Start date
And just like @CC_333, I have longed for one for quite some time now.
CC_333
Well-known member
@Trash80toHP_Mini Agreed totally!
In this day and age, having a semi proprietary external monitor hookup which can only drive an ancient and very difficult to find monitor at a relatively low resolution isn't worth bothering with at this point.
Get the internal grayscale circuits copied *first*, then go from there.
One minor feature that could be nice to have (at some point!) is to build in a fast QuickDraw accelerator, but something like that may end up happening inadvertently nevertheless, simply because it'd be a side effect of using modern components, which I'd imagine tend to be inherently faster than the older equivalents used on the original Xceed.
c
In this day and age, having a semi proprietary external monitor hookup which can only drive an ancient and very difficult to find monitor at a relatively low resolution isn't worth bothering with at this point.
Get the internal grayscale circuits copied *first*, then go from there.
One minor feature that could be nice to have (at some point!) is to build in a fast QuickDraw accelerator, but something like that may end up happening inadvertently nevertheless, simply because it'd be a side effect of using modern components, which I'd imagine tend to be inherently faster than the older equivalents used on the original Xceed.
c
Last edited by a moderator:
Trash80toHP_Mini
NIGHT STALKER
No expert here, but I don't think so. May need to disable a hardware check or tweak the driver to default to internal?
You're still cloning the "entire" card, but simplifying the implementation by removing a sizeable portion of the schematic, hence traces, gates and cable drivers for the external display. That can be picket up at later date in the original form as a second step. Dunno if an 8-bit RAMDAC is even available as a new part, so it might be necessary to substitute a far more capable part.
I think the notion could be most clearly illustrated by graying out much of that cable mess needed to support the external monitor. It might be reduced to just a triangle or even a simple V, dunno, the wiring harness is a separate schematic.
edit: if a 24 bit RAMDAC capable of supporting something on the order of 720p fallback part is needed as a substitute. Adding more VRAM to support higher resolutions or even the original's 640x480 in 24-bit becomes a possibility.
You're still cloning the "entire" card, but simplifying the implementation by removing a sizeable portion of the schematic, hence traces, gates and cable drivers for the external display. That can be picket up at later date in the original form as a second step. Dunno if an 8-bit RAMDAC is even available as a new part, so it might be necessary to substitute a far more capable part.
I think the notion could be most clearly illustrated by graying out much of that cable mess needed to support the external monitor. It might be reduced to just a triangle or even a simple V, dunno, the wiring harness is a separate schematic.
edit: if a 24 bit RAMDAC capable of supporting something on the order of 720p fallback part is needed as a substitute. Adding more VRAM to support higher resolutions or even the original's 640x480 in 24-bit becomes a possibility.
Last edited by a moderator:
CC_333
Well-known member
But only *AFTER* a working clone of the card gets created first!* Only THEN can extra features be considered.
c
*I don't mean to harp on this, but it's of utmost importance to tightly control feature creep by prioritizing getting a working clone developed first, or else this project will end up being a train wreck.
jessenator
Well-known member
Didn't Bolle already make great strides on this? Or am I misremembering?
Bolle
Well-known member
As long as we don’t know the internals of the ASIC there’s not much more to be done.
@maceffects how do those guys in china go about reverse engineering complex custom ICs like the one on that card? Expose the die and map out the chip layout under a microscope?
@maceffects how do those guys in china go about reverse engineering complex custom ICs like the one on that card? Expose the die and map out the chip layout under a microscope?
jessenator
Well-known member
https://spectrum.ieee.org/nanoclast/semiconductors/design/xray-tech-lays-chip-secrets-bare
Anyone using this new procedure we could solicit?
Anyone using this new procedure we could solicit?
maceffects
Well-known member
Trash80toHP_Mini
NIGHT STALKER
I'm making the point that it might not even be POSSIBLE to clone the EXCEED without substituting an available RAMDAC of far superior performance. We'd be tying down floating signals redundant to the internal grayscale, function, using only I/O required. The additionaal function is held in reserve for future development of the Alpha platform. I find it had to imagine a RAMDAC so limited as that on the EXCEED is an available part.
IOW, we likely won't find available parts crappy enough to slavishly recreate the card in its over 30 year old form. VRAM/Video Memory types currently availability is another source of curiosity for me. @trag what do you think about parts availability issues?
I'd say implement for the least (internal gs only) while making choices along the way intended to support the most, whatever that's determined to be during the course of pulling off step one.
Bolle
Well-known member
techknight
Well-known member
FPGAs can take over everything the main chip did, as long as you know the logic behind it. using SDRAM coupled to the FPGA would work fine for the video speeds this thing is working at.
CC_333
Well-known member
Right, OK. I should've been more specific in that I meant *functional* clone; in other words, not one that necessarily uses the exact same components (that would be unrealistic!), but one which mimics the original's functionality using modern components.
This, then, would allow all the functionality to be reproduced, and then, via a firmware update or something, more advanced features can be added (like the Floppy Emu, where BMoW added HD20 functionality simply by updating the firmware (though some very early models needed a hardware fix or two for it to work due to thier being a slightly different design, as I recall)).
c
Last edited by a moderator:
Trash80toHP_Mini
NIGHT STALKER
Ignore this post, move to quote tag disaster corrected version below. [:I]
Last edited by a moderator:
Trash80toHP_Mini
NIGHT STALKER
That's what I was thinking/hoping. As I try to get across below, the only thing about the EXCEED card that really requires cloning is its ability to sync with the A/B and substitution of its output to drive its proprietary neck board. The rest of it is A/B switching its output from internal grayscale resolution to its function as a mediocre VidCard driving an external display.
Much of the legwork has been done in dlv's project.
In a quick review of the topic I found Balrog's post: "https://archive.org/details/MicronXCeedTechDocs — you may have to click "Show All" under Download Options."
@Gorgonopsand I had a bit of a back and forth going about just about everything I barely understand and I'm grateful for his lessons. I dragged in a tangential discussion about developing for 030 PDS as a simpler approach than going straight to NuBus, which puts discussion there right smack into this thread. The most useful thing I brought up was probably using the Macintosh II Video Card (aka Toby) as a design example and pointing out the in depth documentation of it in the Designing Cards and Drivers for the Macintosh series.
I'm going to posit here:
A - the external only XCEED card is an 030 PDS implementation of the NuBus Toby card as its output is the same as the 640x480@8-bit Grayscale Card;
LEM
B - the Internal GS XCEED card implements the ability of the Macintosh II High Resolution Video Card to output the 512 x 384 resolutions for driving the later Apple Displays that are missing from the TobyCard. But the XCEED version was tweaked to match the 342 vertical line resolution of Compact Mac video.
LEM
Thankfully, that card's default setting is 512 x384***** which makes sense as coming up in 640x480 would be useless in the case of a system configured with the smaller display. Im thinking the same must be true for the XCEED card defaulting to internal, no? So tweaking the Internal GS driver shouldn't be necessary for doing an internal only clone?
Schematic, driver and DECLROM info for Toby are in Apple's Developer Docs, but I've yet to find PAL formula listings for the Toby card. I believe they're out there on the web, if not somewhere in Apple's Developer series that I've yet to run across. I can't imagine that dumbing a Toby card down to run the SE/30 512x342 internal display would be an insurmountable problem? As Toby constitutes the XCEED card's external monitor support function, adding external support of the same spec, if not identical in implementation to that of the XCEED's would be easily enough done in phase 2?
**** Macintosh II High-Resolution Video Card: New Features
Last edited by a moderator:
Trash80toHP_Mini
NIGHT STALKER
Interesting: https://vintageapple.org/macbooks/pdf/Micron_Technology_Xceed_SE-3010-78_1989.pdf
So the SE/30 3010-78 supports 1024x768? What's the timeline of Micron Technology's 030 PDS product lineup? Which models support internal grayscale and which don't?
So the SE/30 3010-78 supports 1024x768? What's the timeline of Micron Technology's 030 PDS product lineup? Which models support internal grayscale and which don't?
Trash80toHP_Mini
NIGHT STALKER
I searched "xceed" in the MacUser archive from January 1989 thru January 1991 w/o finding a single peep about internal grayscale. Until that point they appear to have been selling inexpensive, unaccelerated SE/30 PDS cards for external 640x480 and 1024x768 displays alongside a lineup of NuBus cards so uninspiring as to be entirely missing from LEM's NuBus Video Card page. Micron started advertising Xceed Memory Upgrade Kits in the MacUser August 1989 issue and began advertising VidCards in the November issue of that year. Things stayed pretty much the same until they listed a NuBus Card capable of 1152x870 in their November 1990 MacUser ad.
It'll be interesting to find the SE/30 Internal Grayscale announcement or a review of any of their cards.
BTW, it would seem Micron is still rolling right along in the memory business out in Boise. Crucial is one of their trade names. I wonder if anyone has been working there for thirty-odd years?
It'll be interesting to find the SE/30 Internal Grayscale announcement or a review of any of their cards.
BTW, it would seem Micron is still rolling right along in the memory business out in Boise. Crucial is one of their trade names. I wonder if anyone has been working there for thirty-odd years?
Trash80toHP_Mini
NIGHT STALKER
Can anyone take and post some really high resolution pics of both sides of the Micron Xceed 306, please?
It doesn't support internal grayscale. In stock form, the only resolution it supports is 640 x 480 with 8-bit on an external monitor using Sync on Green. It appears that much of the work has been done to reverse engineer this particular card. Tweaking it so as to create an internal grayscale only card appears to be a real possibility. Dunno, but it's definitely an approach to explore.
It doesn't support internal grayscale. In stock form, the only resolution it supports is 640 x 480 with 8-bit on an external monitor using Sync on Green. It appears that much of the work has been done to reverse engineer this particular card. Tweaking it so as to create an internal grayscale only card appears to be a real possibility. Dunno, but it's definitely an approach to explore.
pcamen
Well-known member
Here are a bunch of photos of the Xceed 306 card under both indoor and outdoor lighting conditions.
https://vintageapple.org/pic/Xceed_306_photos.zip
https://vintageapple.org/pic/Xceed_306_photos.zip
Last edited by a moderator:
