SGC successor brainstorming

Nagging hardware related question? Post here!
User avatar
Peter
Aurora
Posts: 917
Joined: Sat Jan 22, 2011 8:47 am

SGC successor brainstorming

Postby Peter » Fri Nov 16, 2018 1:12 pm

Hi,

not that I have time for actual QL work at the moment - just a little pondering when I have spare minutes. No announcement intended, only trying to collect some feedback:
  • Is a SGC successor for the original QL case still important to keep the QL scene alive? Or do the Tetroid (S)GC clones and the Q68 already cover most needs?
If a SGC successor is still important... would these restrictions be acceptable to reduce development work:
  • Video only as VESA compliant VGA output, fully compatible, but QL's historical outputs usused
  • Restricted support for external peripherals to reduce address lines. What is needed? QIMI, Qubide, QL-SD, Minerva I²C, MICE?
  • QL network and microdrives supported in hardware, but new software timing adjustments required (and no promise someone does the work)
  • Internal and external QL hardware ROM contents ignored. Images of ROMs on SD card instead.
  • No floppy connector onboard

I consider these improvements over the SGC:
  • Get rid of external video converters for the QL's non-standard video signal
  • Integrated graphics card with additional colours and resolutions
  • Further RAM expansion, e.g. 12 MB or 28 MB.
  • Moderately higher CPU speed, but 68000 based instead of 68020 for higher QL compatibility
  • One or two (micro)SD card interfaces, much higher speed than Floppy or QubIDE
  • Integrated sampling sound unit in addition to QL speaker
  • Fast PC-style serial port
  • PS/2 mouse interface
  • Re-chargeable realtime clock buffer (no battery replacement)
Thanks, Peter


User avatar
tofro
QL Wafer Drive
Posts: 1664
Joined: Sun Feb 13, 2011 10:53 pm
Location: SW Germany

Re: SGC successor brainstorming

Postby tofro » Fri Nov 16, 2018 2:10 pm

Peter,

the following is just my personal opinion on GC/SGC successors:

  • I don't see the point of an additional processor expansion card over the SGC - That area is covered IMHO with the Tetroid re-makes (although I understand availability is limited due to limited legacy chip supply). Replacing the original mainboard is IMHO the simpler option over a complicated 8/16-bit bus interface and limited space on an expansion card
  • I could completely live without microdrives, with the limited supply of cartridges, there's not much point in them. SD is a perfect replacement
  • No need to have much on-board ROM other than a small loader - Load the OS from card
  • (at least) VGA output is a must. High-color and Hi-Res up to 800x600 (not much more) is a nice option.
  • I'd like to see the network retained
  • I think there should be dedicated video memory and a split bus like on the original to keep contention down to the absolute minimum area
  • Allowing video double buffering in smaller resolutions would be nice (two screens, ideally movable)
  • A CPU with at least some cache like a 68030 seems appropriate. 68030 seems to be in good supply, still.
  • An internal I2C expansion slot would be nice. (E.g. to an optional RPi or other embedded CPU)
  • A/D sampled sound like on Q68 is nice, but the "real retro thing" is something along the lines of an AY, Yamaha OPL or SID. Also fits better to a relatively low-performance CPU. Could go to the I2C bus
  • I don't see the need for a floppy interface once you have SD cards.
  • I like the idea of a simplified expansion bus as the Q68 has. And I intend to use that in future ;). Use the original QL's expansion space. Probably add a 5V power supply, though and make the port 5V tolerant.
  • From the above: I don't see any need to keep the original expansion bus.
  • Before someone starts talking about USB or TCP/IP: I don't consider that very feasible without a modern co-CPU (see above for RPi) other than maybe for HID devices
  • Dual full UART would be nice
  • Once you go for a mainboard replacement, keyboard matrix support is obviously mandatory, PS/2 becomes an option

Tobias


ʎɐqǝ ɯoɹɟ ǝq oʇ ƃuᴉoƃ ʇou sᴉ pɹɐoqʎǝʞ ʇxǝu ʎɯ 'ɹɐǝp ɥO
User avatar
dilwyn
Mr QL
Posts: 1517
Joined: Wed Dec 01, 2010 10:39 pm
Location: Wales
Contact:

Re: SGC successor brainstorming

Postby dilwyn » Fri Nov 16, 2018 3:04 pm

I broadly agree with what has been written by Peter and Tobias. However, since the Tetroid SGC's became available I think that video output is the single thing that needs to be acted upon.

At the moment, we are reliant on SCART TV connections and SCART/HDMI type converters.

We need a reliable dedicated video connection to allow existing commonly available monitors to be used.


Dilwyn
All things QL: see www.dilwyn.me.uk
Derek_Stewart
QL Wafer Drive
Posts: 1493
Joined: Mon Dec 20, 2010 11:40 am
Location: Runcorn, Cheshire, UK

Re: SGC successor brainstorming

Postby Derek_Stewart » Fri Nov 16, 2018 3:22 pm

Hi,

I would like the same functionality of the Q68 on the QL.

I think Ethernet can be disregarded, connection can be achieved via a fast serial port to a PC.

The main upgrade of the QL would be a HI-Colour VESA display, with similar display modes of the Q68.

The new board to run SMSQ/E.

This sounds like an updated Aurora board.

I do not think a faster processor is required, as most existing QL software would not take advantage of the enhanced properties of the 68020, 030 etc...


Regards,

Derek
User avatar
Peter
Aurora
Posts: 917
Joined: Sat Jan 22, 2011 8:47 am

Re: SGC successor brainstorming

Postby Peter » Fri Nov 16, 2018 4:17 pm

dilwyn wrote:However, since the Tetroid SGC's became available I think that video output is the single thing that needs to be acted upon.
At the moment, we are reliant on SCART TV connections and SCART/HDMI type converters.

Yes Dilwyn, the lack of a reliable, standardized QL video output actually made me think of a SGC successor in the first place.

I have thought about a video-only solution that would work with (S)GC or bare QL. But the added amount of design work for a full SGC successor is relatively small.

Every video-only solution would suffer extremely from the slow 8 bit QL bus. This is immediately solved as soon as a CPU can be added to the design, which directly interfaces video memory. Now once a CPU is added, it is not a big step to also add some peripherals that I could re-use from my Q68 design.


User avatar
Andrew
Gold Card
Posts: 264
Joined: Tue Jul 17, 2018 9:10 pm

Re: SGC successor brainstorming

Postby Andrew » Fri Nov 16, 2018 4:41 pm

I think that first we should define the goal of designing a SGC card
A. Bring the QL to the specs of a 21st century computer ? I do not believe this is feasible. The hardware cam be designed and built, but there is a lack of programmers. The QL community is just too small for this
B. Keep the QL as the retro-computer that it should have been ? maybe, but we have to face the reality: I guess that all QL users are over 45 years old and already an endangered species. Retro computers that appeal to younger people are those who have a large library of games (QL does not) and have a large library of software and capabilities for writing new games and graphic demos. A new SGC can address the issue of capabilities, including bitblt :)
C. Retro-computer with a twist - add something new to appeal to larger programming community. There are lots of possibilities here, but what I would like to have is an Arduino and Raspberry Pi compatible pin header that would allow the use of all the shields and hats and sensors available on the market today.

So what is your view on this ?

What I would like to have is:
- replacement motherboard that would fit in the QL case, including keyboard adaptor
- VGA and HDMI output ( maybe using a Raspberry pi for that)
- QL standard graphics + color modes 640x480 and maybe 800x600 , at least 256 colors optimized for animation (with bitblt ? and this should be standardized)
- 2 SD card slots and QUBIDE
- keep the Floppy connector ! (I find it simpler to run some old programs from floppies than from SD)
- some sort of A/D sound (use the same raspberry pi for that ?)
- 68020 or 68030 and 4 Mb of RAM (more RAM is nice, but I do not feel it is particularly useful)
- modern serial and paralel ports (USB would be nice !)
- Arduino and Raspberry PI compatible pin head
- operating system can be loaded from the SD (as long as I can also use the QL mode)


User avatar
Peter
Aurora
Posts: 917
Joined: Sat Jan 22, 2011 8:47 am

Re: SGC successor brainstorming

Postby Peter » Fri Nov 16, 2018 5:27 pm

tofro wrote:Replacing the original mainboard is IMHO the simpler option over a complicated 8/16-bit bus interface and limited space on an expansion card

Problems with this approach:
  • British QLs have different connectors and case shape. So if the SubD variant is used, which makes sense for SER and VGA, british users need to modify their case. :(
  • Need to rip out the original mainboard. I'd hesitate to do that, unless it is damaged anyway.
  • Microdrives dead. Since the original case is targeted, I expect some remaining interest in them.
Space limitation for a SGC successor is not a big issue these days.

tofro wrote:An internal I2C expansion slot would be nice.

Yes, I²C should be added.

tofro wrote:I like the idea of a simplified expansion bus as the Q68 has. And I intend to use that in future ;).

A very important remark. I was under the impression that nobody except myself would really use it in practice, and might have abandoned it.

tofro wrote:Dual full UART would be nice

Noted. Depends on available pins.

tofro wrote:Once you go for a mainboard replacement, keyboard matrix support is obviously mandatory, PS/2 becomes an option

Agreed for keyboard. For mouse, PS/2 seems almost mandatory.


User avatar
Peter
Aurora
Posts: 917
Joined: Sat Jan 22, 2011 8:47 am

Re: SGC successor brainstorming

Postby Peter » Fri Nov 16, 2018 5:50 pm

Andrew wrote:I think that first we should define the goal of designing a SGC card

My goal would mainly be to keep the QL scene alive. In other words, if 90% of the users are okay with the Tetroid add-ons or Q68, it makes little sense.


User avatar
Andrew
Gold Card
Posts: 264
Joined: Tue Jul 17, 2018 9:10 pm

Re: SGC successor brainstorming

Postby Andrew » Fri Nov 16, 2018 5:55 pm

Peter wrote:My goal would mainly be to keep the QL scene alive. In other words, if 90% of the users are okay with the existing Tetroid add-ons or Q68, it makes little sense.

To keep the QL scene alive I believe you must add something to the QL that will attract new users.
The QL scene is small and my guess is that we all are past 45 or even 50 yo.


User avatar
dilwyn
Mr QL
Posts: 1517
Joined: Wed Dec 01, 2010 10:39 pm
Location: Wales
Contact:

Re: SGC successor brainstorming

Postby dilwyn » Fri Nov 16, 2018 6:02 pm

Peter wrote:
dilwyn wrote:However, since the Tetroid SGC's became available I think that video output is the single thing that needs to be acted upon.
At the moment, we are reliant on SCART TV connections and SCART/HDMI type converters.

Yes Dilwyn, the lack of a reliable, standardized QL video output actually made me think of a SGC successor in the first place.

I have thought about a video-only solution that would work with (S)GC or bare QL. But the added amount of design work for a full SGC successor is relatively small.

Every video-only solution would suffer extremely from the slow 8 bit QL bus. This is immediately solved as soon as a CPU can be added to the design, which directly interfaces video memory. Now once a CPU is added, it is not a big step to also add some peripherals that I could re-use from my Q68 design.


I think Tetroid's SGC compatible suffers from availability of a critical chip used, so a SGC successor actually makes sense if Tetroid is unable to make more than a few of these add-ons because of that chip shortage.

I think it is absolutely vital though, that whatever add-on card of this type is made, that the improved video connection is included.


Dilwyn
All things QL: see www.dilwyn.me.uk

Who is online

Users browsing this forum: No registered users and 4 guests