[Coco] High Color Display Mod for MM/1?

Joel Ewy jcewy at swbell.net
Fri Sep 5 11:35:30 EDT 2008


Bob Devries wrote:
> Joel,
> It's certainly an interesting prospect. As well as that, there's the
> so called VGA mode that the MM/1 is supposed to have, but no-one seems
> to have knowledge of.
>
I wonder what 'VGA mode' means in this context?  Are we talking about
screen/color resolution, sync rates, or both?  Many of the MM/1's
graphics modes are comparable to those of the original VGA adapters,
though they're not exactly the same, and I attribute that to having to
sync at NTSC( and/or PAL?) rates.  I do also seem to recall that the
MM/1 was supposed to be able to drive a monitor at VGA sync rates. 
Maybe this is what you mean, and maybe that would also allow it to
generate screen resolutions that are really commensurate with VGA on a PC.

As far as I can tell, the MM/1 produces the sync signals internally from
some division of its clock.  If that's correct, it seems likely that
that can be altered under software control, though there may not be any
provision in K-Windows to do so.  Just looking quickly through the VSC
databook, I don't see any obvious way to change the horizontal retrace
from 15KHz to the 35 that a VGA monitor would need, but I may have
missed something.

>From a slightly different angle, I'm not sure what there is to prevent
one from running one of the higher resolution modes in 8 bits per
pixel.  IIUC, the only real limitation there is the amount of RAM
addressable by the VSC, which is 1M, but I could be wrong about that. 
On a 1M system, all available memory must be shared between video, OS,
data, and applications.  On a system with 3M+, I would think that the
entire 1M on the CPU board should be available for video memory, unless
there is some other need for that memory that I'm not aware of.

Running the VSC in 8-bit mode at 768x480 (interlaced, overscan) should
consume less than half the 1M (368640 bytes).  Or are we running up
against memory bus bandwidth limitations here?  Again, K-Windows doesn't
support this mode (maybe in part because it would use too much memory in
a 1M system?).
> It would be good to collect data sheets of all the major chips in the
> MM/1 and upload them to maltedmedia. I have a PDF file of the
> SCC66470B which I can upload.
>
Here's a link from where I think you can get to datasheets for both the
BT478 and the BT481.

JCE
> It could also be advantageous to upload the disk images of the
> original distribution disks.
>
> My own MM/1 is a CPU board with an IO board mounted on a non-active
> backplane (no memory provision), and I have the MIDI paddle board.
>
> -- 
> Regards, Bob Devries, Dalby, Queensland, Australia
>
> Isaiah 50:4 The sovereign Lord has given me
> the capacity to be his spokesman,
> so that I know how to help the weary.
>
> website: http://www.home.gil.com.au/~bdevasl
> my blog: http://bdevries.invigorated.org/
>
> ----- Original Message ----- From: "Joel Ewy" <jcewy at swbell.net>
> To: "CoCoList for Color Computer Enthusiasts" <coco at maltedmedia.com>
> Sent: Friday, September 05, 2008 3:58 PM
> Subject: [Coco] High Color Display Mod for MM/1?
>
>
>> Ok, you've all gotten me thinking about the MM/1 again.  I've long
>> wanted to experiment with restuffing all, or more likely only part of
>> the palette registers on the MM/1 to get more colors per line, like
>> what's already been done for high color displays on the CoCo.  So I
>> needed to know the address of the RAMDAC.  I was looking at the
>> schematics Bob uploaded instead of falling asleep, and then got curious
>> and went searching for a datasheet for the Brooktree BT478.  Before I
>> found it, I came across a datasheet for the Brooktree BT481, which
>> appears to be an updated, pin-compatible replacement for a number of
>> RAMDAC parts, including the BT478.  What intrigues me about this chip is
>> that it has high-color and true-color modes where it latches serial
>> bytes to feed the video DACs directly, bypassing the palette.  There is
>> one 15-bit (5-5-5) mode that looks particularly promising.
>>
>> Now the problem with this is that, at the very least, the horizontal
>> resolution would be halved.  So the interlaced overscan 384x480 8-bit
>> mode would become a rather odd 192x480 15-bit color mode.  Would it be
>> worth doing?  I don't know.  Probably not, but it's certainly
>> interesting to think about, and I'm definitely going to keep my eyes
>> open when I next look through my boxes of old video cards to see if
>> there's a BT481 out there that can be liberated from a socket or gently
>> heat-gunned off its board.
>>
>> Any thoughts?
>>
>> JCE
>>
>> -- 
>> Coco mailing list
>> Coco at maltedmedia.com
>> http://five.pairlist.net/mailman/listinfo/coco 
>
>
> -- 
> Coco mailing list
> Coco at maltedmedia.com
> http://five.pairlist.net/mailman/listinfo/coco
>




More information about the Coco mailing list