The 8-Bit Guy's new 65816 based computer

classic non-acorn hardware, software & languages
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

YouTuber 'The 8-bit Guy' is building a new 6502-compatible computer.

Bil Herd is apparently designing the main board!

https://www.youtube.com/watch?v=ayh0qebfD2g
Last edited by Andrew_Waite on Wed Feb 20, 2019 3:19 am, edited 1 time in total.
User avatar
BigEd
Posts: 3879
Joined: Sun Jan 24, 2010 10:24 am
Location: West Country
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by BigEd »

I'll watch this with interest! $50 target ($100 limit) sounds good.
Ramtop
Posts: 276
Joined: Tue Oct 23, 2018 1:40 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Ramtop »

Interesting. Looks like a very worthwhile goal, and maybe David is one of the few people with enough reach in the retrocomputing community to give the project enough momentum.

That said, I wish he wasn't using Commodore's Kernal and BASIC as a base. They're both quite primitive.
Gary
User avatar
davidb
Posts: 2996
Joined: Sun Nov 11, 2007 10:11 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by davidb »

It's interesting that he's using a Gameduino to help with prototyping. I could imagine sticking one of those on a cartridge for the Electron, though I wonder where he bought it from - the official Gameduino store has moved onto selling Gameduino 3 which is an LCD-based thing.

It's funny that he mentioned the BASIC Engine because some research around that led me to the VLSI Solution VS23S010 which looks like it would be interesting for generating a video signal.

Anyway, his new effort makes me want to look at cartridge development again - a little bit, anyway - so that can't be a bad thing. :)
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

Ramtop wrote:
Wed Feb 20, 2019 12:16 am
That said, I wish he wasn't using Commodore's Kernal and BASIC as a base. They're both quite primitive.
From the video (at 12:30), the plan is to write a custom kernel. Cloanto will not licence the C64 kernel, either.
User avatar
paulv
Posts: 3854
Joined: Tue Jan 25, 2011 6:37 pm
Location: Leicestershire
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by paulv »

Is interesting that he talks about the Pi as just a Linux machine. Surely sticking RISC OS on there and using BBC BASIC with ARM assembler gets him "close to the hardware" like he wants.

I realise it's not an 8-bit, machine at this point but it would achieve all his other goals with zero effort.

Paul
User avatar
BigEd
Posts: 3879
Joined: Sun Jan 24, 2010 10:24 am
Location: West Country
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by BigEd »

There's usually one or two mistakes like this in his videos - I resist the temptation to shout at the screen. Indeed, the Pi is a full computer, and an ARM computer, but it's not a Linux computer. Although it's true to say that most people think it is, so from a marketing perspective, building a project on a Pi might lead some potential customers to dismiss the project as an emulation.

I like the idea of building something roughly as capable as a C64 but using modern parts, and not using FPGA. Although FPGAs are both wonderful and great.

I think his starting point is the W65C265SXB from WDC, "an 8/16–bit Single Board Computer (SBC) built around the W65C265S System-On-a-Chip (SOC) microcomputer" - presumably not as a component but as a dev board. That board is $99 or £100 already, but the 265 chip is just £12 so not too bad. Rather a lot for a £50 target price point.
Ramtop
Posts: 276
Joined: Tue Oct 23, 2018 1:40 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Ramtop »

Andrew_Waite wrote:
Wed Feb 20, 2019 3:37 am
From the video (at 12:30), the plan is to write a custom kernel. Cloanto will not licence the C64 kernel, either.
His words are vague enough to be open to several interpretations, but it did sound to me like he's intending to write something similar to the C64 kernal. Given his extensive experience programming the VIC-20 and C64 that isn't wholly surprising, but I hope for something rather better.
Gary
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

Ramtop wrote:
Wed Feb 20, 2019 9:04 am
Andrew_Waite wrote:
Wed Feb 20, 2019 3:37 am
From the video (at 12:30), the plan is to write a custom kernel. Cloanto will not licence the C64 kernel, either.
His words are vague enough to be open to several interpretations, but it did sound to me like he's intending to write something similar to the C64 kernal. Given his extensive experience programming the VIC-20 and C64 that isn't wholly surprising, but I hope for something rather better.
Sorry, I get your point now. I guess The 8-Bit Guy's core viewership is mainly C64 enthusiasts, so this would make sense.
Last edited by Andrew_Waite on Wed Feb 20, 2019 7:58 pm, edited 2 times in total.
User avatar
sPhilMainwaring
Posts: 309
Joined: Tue Jan 15, 2013 7:57 pm
Location: Mid Wales
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by sPhilMainwaring »

I found this video by accident and thought I'd look on stardot to see if anyone else knew about it

Like BigEd I also wanted to shout at the screen when the Pi was described as a Unix machine. I also immediately thought of all the "Bare Metal" Pi stuff we've done and thought of people like Hoglet when he went on to mention the "Gameduino" board (draft spec here)

https://en.wikipedia.org/wiki/Draft:Gameduino

I've joined his group on FaceBook and messaged him with what I think is a CUNNING plan - I'm thinking educational machine with lots of GPIO pins to mimic user ports and "Adval" ports but also 1MHz bus and Tube so that the machine could be used to emulate *any* retro machine quickly via (optional) FPGA copro's - even IBM PCs

A kind of reboot of the whole Acorn Co-Pro ethos ... this would also keep the costs of a "Basic" machine very low and, as it's based on a 6502 backwardly compatible processor, could have a default boot as a Master

As mentioned this keeps down the cost of the Basic units, and would make "FPGA Beebs" and/or FPGA Domesday systems and all manner of other retro goodies easy to implement.

These would also then appeal to many, many different retro communities (Spectrum AGD, etc. with the new "Multi Platform" versions) and also bring in lots of skill sets from theses different communities which would assist in development. They would also be great for museums to go in old kit cases, etc. and school group soldering/construction projects whilst not being tied to a single (i.e. Commodore machine) emulation model

Thoughts?
Ramtop
Posts: 276
Joined: Tue Oct 23, 2018 1:40 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Ramtop »

BigEd wrote:
Wed Feb 20, 2019 8:03 am
I like the idea of building something roughly as capable as a C64 but using modern parts, and not using FPGA. Although FPGAs are both wonderful and great.
This is the aspect that I find most appealing. Using an FPGA or a highly integrated SoC would be easier, but I find a computer with separate components is much more satisfying to mess around with.
Gary
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

How realistic is it to build $50-$100 8/16-bit computers with multiple video modes and some I/O without FPGAs?

The Spectrum Harlequin does without either a ULA or FPGA, but replicating the functionality of the ULA takes ~40 logic chips!
Last edited by Andrew_Waite on Wed Feb 20, 2019 10:15 pm, edited 3 times in total.
Ramtop
Posts: 276
Joined: Tue Oct 23, 2018 1:40 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Ramtop »

I think it'd be no-go for $50, but $100 is not impossible if it is being sold near cost. The machine at minimum would need the 65816 CPU, some flash, some SRAM, a video chip like the VS23S010, a small CPLD for I/O and glue logic, and a sound chip. None of these are expensive, and I suspect $100 is viable if the system can be built in batches of 1000+ units.

My only concern on the hardware side would be the sound chip. There don't appear to be any modern FM synthesis chips easily available. Fitting a DAC would be an option, but not really in the spirit of things.

The actual system design should not be too complex or expensive to do. I could take those components and have a semi-functional board design in a week or two, and I'm still a novice. Someone like Bil Herd could do that work in his sleep. The 65816's 8-bit data bus keeps things simple. I'd personally prefer a 68SEC000, which is still in production, but that's a big ask on a $100 board and a pure 16-bit chip is probably pushing the ethos of the project a bit too far.
Gary
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

Some more datails of the machine, via Facebook.

https://www.facebook.com/groups/3964082 ... 093731324/

Ok. This is not a big official announcement and many things may change but I thought I’d share what much of the overall design will likely be and some specs.

CPU: 65c265 or 65c816. Speed is unknown yet and will depend on which chip we use.

RAM: either 512k or 1m of SRAM

ROM: a 32k onboard ROM which will hold the core OS as well as BASIC and a monitor.

Keyboard: PS/2 keyboard. We already know how to drive it and USB is complex. We may implement USB at some point but if so it will be limited to PS/2 compliant keyboards.

Mouse: PS/2 mouse. Just makes sense as well. The code to interface it is the same as the keyboard, all that changes is interpreting the data.

Storage: two storage methods are supported. There will be an onboard SD card. In addition there will be IEC support AKA legacy Commodore drives. There will be no integrated cassette support. It’s just too slow, too unreliable and it means we’d have to sacrifice precious board space to a card edge connector.

Joysticks. Two NES or SNES protocol controller ports. We might use a different connector though I don’t think anyone including myself likes that idea. But the legacy Atari style port will not be used. We want more buttons than the Atari supports officially and we want to keep as much IO as possible available to the user.

GPIO: we are strongly supporting the idea of using the Arduino type shields. It’s a good standard for projects, they are stackable with pass through.

Cartridges: this is up in the air but it has been suggested to have a cartridge slot for commercial software and hardware distribution. Shields are nice for projects but not for game titles. Details on this are yet to be worked out. The cartridge will likely be supported through SPI would allow a large range of ROMs, flash in different sizes as well as different kinds of SPI hardware. This is not an urgent aspect but it would allow a great and inexpensive way for developers to make commercial software titles and to distribute that software in a physical form.

RS-232: we do plan to support it.

WiFi: maybe. I think it’s good make it an available option. But it may not be integrated in version 1

Video: we are not decided on it yet. Gameduino is still a contender as are a couple other possibilities.

Audio: undecided. Audio is not needed for early development. We have some fallback options and may end up going with a custom solution.

Character set: a custom PETSCII variation. It will be closer to ASCII in someways but will retain the things that make PETSCII great.

BASIC. Initially we do want to be able to support BASIC 2.0. However don’t take that to mean we will include it. The final product will likely use a faster, smarter, easier to use BASIC with useful new commands, better disk support, graphics features, and audio commands. Many programs written in BASIC should run with little or no alteration.

KERNAL: David talked about that in his video. It will handle a standardized way to interface with much of the hardware. It will try to maintain the original VECTOR table minus features that just aren’t useful. This means in theory software that runs on the pet, Vic-20, or c-64 that uses KERNAL calls might actually work. This rules out most games and demos, but it will be interesting to try some software and see what might actually run. I imagine this number will be very low.

Memory map: this will likely be different from the map found on many Commodore machines. It’s not settled yet. So which memory is available for use is a little bit different. I feel the proposed map breaks away from a few archaic and limiting design decisions that no doubt were great in their day. But it does probably hurt compatibility even more and as far as BASIC 2.0 is concerned will actually leave less memory available for BASIC programs than the C64 had. But our new BASIC will not have this issue at all. Few BASIC programs used anywhere near that amount anyway, the largest program I ever wrote was only about 8k.

Documentation: there will be a quick start BASIC guide modeled after the ease and simplicity of the VIC-20 guide. There will be a more sophisticated programmers reference guide as well and a guide to introduce assembly language.

Hopefully that covers the bases for now.
User avatar
BigEd
Posts: 3879
Joined: Sun Jan 24, 2010 10:24 am
Location: West Country
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by BigEd »

Thanks!
Prime
Posts: 2874
Joined: Mon Jun 01, 2009 12:52 am
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Prime »

My first comment on the joysticks and using NES/SNES sticks, yep it may save you on I/O but at the expense of added complexity and severly limiting what is available to the end user. If you want to limit the I/O used mux the two ports and use the Sega Megadrive scheme which essentially muxes the additional buttons, and is pretty easy to drive. If you don't need to read the additional buttons just read it like an Atari, if you need them it's just a couple of extra instructions.

Cheers.

Phill.
Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Kazzie »

Prime wrote:
Thu Feb 21, 2019 3:26 pm
My first comment on the joysticks and using NES/SNES sticks, yep it may save you on I/O but at the expense of added complexity and severly limiting what is available to the end user. If you want to limit the I/O used mux the two ports and use the Sega Megadrive scheme which essentially muxes the additional buttons, and is pretty easy to drive. If you don't need to read the additional buttons just read it like an Atari, if you need them it's just a couple of extra instructions.

Cheers.

Phill.
That echoes my thoughts. Sega got as far as eight buttons and a d-pad through that connector, although things were getting a bit convoluted by then. The three-button-plus-start controllers were dead simple, using just a single 7400 series IC in the controller if I recall correctly.

I can't think of where I'd go to source NES/SNES controller connectors, other than buying aftermarket knockoff controllers and chopping the leads off. DE9 connectors, on the other hand? I've got some in the house already, and can go to a long list of electronics suppliers to get more.
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

David Murray, 'The 8-bit Guy', has posted another update to the machine's specification, again on Facebook.

"Just an update on possible video direction. I have been examining dozens of different video options and discussing them with other experts. Somebody has already modified the Gameduino 1 FPGA so that it can sit on the address/data bus instead of using SPI. Granted, I don't have a prototype of that just yet, but the code is done. So this way, the CPU will be able to directly write to the Gameduino's 32K RAM as if it were its own. This makes the Gameduino (assuming it works) much more likely as the final video chip. Also, initially I said it didn't have a bit-mapped mode with individually addressable pixels. This isn't entirely true. Apparently you can create a pseudo-bitmapped mode by laying all of the sprites on the screen, sort of how the VIC-20 worked with character graphics. However, you can't fill the whole screen, but certainly most of it. Since the Gameduino is available today, and it is open-source, it is starting to look like a viable video solution. It's also well documented, easy to program for, and it is more powerful than the C64's video, but not quite as powerful as other 16 or 32 bit computers, which is exactly what I was going for. I'm not saying the Gameduino is definitely going to be the chip we use, but it is starting to look more likely. The next step will be to attempt to install the new design onto the C64's memory map (which will be a challenge) for testing."
alex_farlie
Posts: 163
Joined: Sun Jul 07, 2013 10:46 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by alex_farlie »

Hmm...

How easy would it be (given various peoples expertise in FPGA) to build an accelerated "Positron" which is essentially a 65816 style chip, an I/O chip like the Electron ULA, and so on?
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

alex_farlie wrote:
Fri Mar 01, 2019 3:31 pm
How easy would it be (given various peoples expertise in FPGA) to build an accelerated "Positron" which is essentially a 65816 style chip, an I/O chip like the Electron ULA, and so on?
Hacking one of these new machines to make an 'Acorn Positron' would be a very nice project.

In the meantime there are some more details of the machine, including a memory map, published on Facebook.

For the moment we are focusing on the 65c816 and not the 65c265. While the microcontroller has some benefits in the form of built in IO, it also has a lot of annoyances and since part of our goal is to make our platform easy to program and develop we found its benefits really aren’t that beneficial to us.

For the time being we are going with the AY-3-8910 sound chip. Now we haven’t been able to verify the claim of clones still being manufactured, this chip definitely provides the 8bit feel. We are going to use two of these chips giving 6 voices. Additionally these chips give us 16 IO pins each for a total of 32 IO pins, which is more usable pins than the 265 would have given us.

For video we are still planning on the Gameduino as the starting point. The tests right now are using the SPI, but we are having a board made that gives us a parallel interface instead. Again this is an incremental step, not a final design.

For keyboard and mouse we are using PS/2. Reasons for this have already been stated and the routines to handle this are underway.

The board will have RS-232 and will be provided by a dedicated chip. There are several well suited and inexpensive options for this. It will use a standard RS-232 DB-9 port.

For controllers we are still leaning towards the SNES standard. The Atari standard has way to many marks against using it including that fact it’s not very standardized when you through all the variations to it out there. Also since we will already have RS-232 with a DB-9 adding controller ports that make it possible to plug the joystick into the serial port or to plug RS-232 into the joystick port seems foolish. Now if you want, you can always use an adapter to convert Atari or Sega controllers to the SNES port.

But now the big reveal on the memory map.

Starting at the bottom. For clarity each bank is 64k in size. But you do have long addressing. In bank zero you will have $0000 - $07ff which will be RAM and is reserved for use by direct page and the KERNEL and BASIC. Then from $0800-0fff you have the IO area. From $1000-7fff is RAM. The last part of it, $7000-7fff is the new stack area.
Then from $8000-ffff is the KERNEL area. This area can be either ROM or RAM and by default at startup the ROM copies itself to RAM and then disables the ROM to enable full speed. That wraps it up for the first bank.

Bank1 starts out with a 32k option area. This area can be either RAM, or one of up to 16 ROM slots which can be flashed with various utilities or programs. Additionally even when it is filled with RAM it can be write protected to act like ROM, which is a good way to test a program to see if it works correctly as a ROM.
The upper 32k of bank1 is the Gameduino.

The remaining banks up to 1meg are all RAM.

All the addressable space beyond 1meg are treated as bus expansion. This area can be used to add more memory or any number of things. It operates at a reduced speed, and will support DMA.
User avatar
gordonDrogon
Posts: 26
Joined: Fri Nov 23, 2018 12:39 pm
Location: Devon
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by gordonDrogon »

I wonder what issues they think they might have with the '265 rather than the '816...? Differences I can think are that it's "only" 8Mhz although I've seen others suggest they're running it much faster. The IO ports are somewhat.. well, just different to a 65c22 that we might be used to, and you do sacrifice some for IO/Memory decoding, however I'm sure it'll still be an interesting beast..

-Gordon
alex_farlie
Posts: 163
Joined: Sun Jul 07, 2013 10:46 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by alex_farlie »

Andrew_Waite wrote:
Fri Mar 15, 2019 11:09 pm
alex_farlie wrote:
Fri Mar 01, 2019 3:31 pm
How easy would it be (given various peoples expertise in FPGA) to build an accelerated "Positron" which is essentially a 65816 style chip, an I/O chip like the Electron ULA, and so on?
Hacking one of these new machines to make an 'Acorn Positron' would be a very nice project.

In the meantime there are some more details of the machine, including a memory map, published on Facebook.

For the moment we are focusing on the 65c816 and not the 65c265. While the microcontroller has some benefits in the form of built in IO, it also has a lot of annoyances and since part of our goal is to make our platform easy to program and develop we found its benefits really aren’t that beneficial to us.

...
But now the big reveal on the memory map.

Starting at the bottom. For clarity each bank is 64k in size. But you do have long addressing. In bank zero you will have $0000 - $07ff which will be RAM and is reserved for use by direct page and the KERNEL and BASIC. Then from $0800-0fff you have the IO area. From $1000-7fff is RAM. The last part of it, $7000-7fff is the new stack area.
Then from $8000-ffff is the KERNEL area. This area can be either ROM or RAM and by default at startup the ROM copies itself to RAM and then disables the ROM to enable full speed. That wraps it up for the first bank.


So you if swapped the Acorn MOS style entry points to be at FExx instead of FFxx (Due to additional high vectors on the 65816).. (I wonder how the Communicator handled this, or indeed the Jon Kortnik co processor using a patched HiBasic), maybe someone could write a Positron MOS for this new platform. Programming it would be more like programming on an 8086, with all those FAR calls.. Maybe the Positron would use a variant on an Extended Vector to do some Far Calls... Interesting....
:D
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

An update posted by ‎Lorin Millsap on the 'Commander X16 Prototype' Facebook page in the last hour or so. There will be an emulator and a kickstarter campaign :

Ok, so a product announcement since I think people are ready for this. A common question has been asked about an emulator. I am pleased to announce we are working on this emulator alongside our efforts on the hardware. So far it employs a custom 816 implementation that is cycle accurate. Because it is being developed alongside the hardware it will be able to emulate that hardware as accurately as possible. A special shoutout to Robin Raymond for his efforts in this venture.

As to release of this emulator, when we start the Kickstarter campaign we will release the emulator to backers. When the product officially launches the emulator will be available to everyone.
Last edited by Andrew_Waite on Mon Apr 08, 2019 2:33 pm, edited 2 times in total.
User avatar
gordonDrogon
Posts: 26
Joined: Fri Nov 23, 2018 12:39 pm
Location: Devon
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by gordonDrogon »

Andrew_Waite wrote:
Mon Apr 08, 2019 2:31 pm
An update posted by ‎Lorin Millsap on the 'Commander X16 Prototype' Facebook page in the last hour or so. There will be an emulator and a kickstarter campaign :

Ok, so a product announcement since I think people are ready for this. A common question has been asked about an emulator. I am pleased to announce we are working on this emulator alongside our efforts on the hardware. So far it employs a custom 816 implementation that is cycle accurate. Because it is being developed alongside the hardware it will be able to emulate that hardware as accurately as possible. A special shoutout to Robin Raymond for his efforts in this venture.

As to release of this emulator, when we start the Kickstarter campaign we will release the emulator to backers. When the product officially launches the emulator will be available to everyone.
Saw that this morning... In some ways disappointed - in that if it fails to reach their goals ... then what? Still, plenty more '816 projects in the works :-)

-Gordon
User avatar
gordonDrogon
Posts: 26
Joined: Fri Nov 23, 2018 12:39 pm
Location: Devon
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by gordonDrogon »

alex_farlie wrote:
Sun Mar 17, 2019 12:24 am
So you if swapped the Acorn MOS style entry points to be at FExx instead of FFxx (Due to additional high vectors on the 65816).. (I wonder how the Communicator handled this, or indeed the Jon Kortnik co processor using a patched HiBasic), maybe someone could write a Positron MOS for this new platform. Programming it would be more like programming on an 8086, with all those FAR calls.. Maybe the Positron would use a variant on an Extended Vector to do some Far Calls... Interesting....
:D
Does anyone think it's possible to patch-up some existing BBC Micro code to have the vectors changed? ie. scan a ROM and look for JSR $FFxx or JMP $FFxx and change them as required, then save a new ROM image?

Just wondering if some ROMs/software were that well behaved or used other means to get to the MOS (push/rts, etc. ?)

(I have a little vanity 65C02 system running which can run BBC Basic, etc. and plan to move to an '816 and a couple of the 816 vectors clash right at the top - I guess I'll just have to "suck it and see" :)

Cheers,

-Gordon
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

gordonDrogon wrote:
Mon Apr 08, 2019 2:55 pm
alex_farlie wrote:
Sun Mar 17, 2019 12:24 am
So you if swapped the Acorn MOS style entry points to be at FExx instead of FFxx (Due to additional high vectors on the 65816).. (I wonder how the Communicator handled this, or indeed the Jon Kortnik co processor using a patched HiBasic), maybe someone could write a Positron MOS for this new platform. Programming it would be more like programming on an 8086, with all those FAR calls.. Maybe the Positron would use a variant on an Extended Vector to do some Far Calls... Interesting....
:D
Does anyone think it's possible to patch-up some existing BBC Micro code to have the vectors changed? ie. scan a ROM and look for JSR $FFxx or JMP $FFxx and change them as required, then save a new ROM image?

Just wondering if some ROMs/software were that well behaved or used other means to get to the MOS (push/rts, etc. ?)

(I have a little vanity 65C02 system running which can run BBC Basic, etc. and plan to move to an '816 and a couple of the 816 vectors clash right at the top - I guess I'll just have to "suck it and see" :)

Cheers,

-Gordon
Please find a link to a diagram of the X16's memory map below.

Unfortunately they have memory mapped I/O between &0800-&0FFF, however I did ask if memory decode will be done in programmable logic, and it will be - still hopeful this machine can be the basis of a brand new Acorn Positron.

https://bl.ocks.org/rsbohn/raw/0d633f94 ... 8_uR66Kwz8
Last edited by Andrew_Waite on Mon Apr 08, 2019 3:18 pm, edited 2 times in total.
User avatar
gordonDrogon
Posts: 26
Joined: Fri Nov 23, 2018 12:39 pm
Location: Devon
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by gordonDrogon »

Andrew_Waite wrote:
Mon Apr 08, 2019 3:08 pm
gordonDrogon wrote:
Mon Apr 08, 2019 2:55 pm
alex_farlie wrote:
Sun Mar 17, 2019 12:24 am
So you if swapped the Acorn MOS style entry points to be at FExx instead of FFxx (Due to additional high vectors on the 65816).. (I wonder how the Communicator handled this, or indeed the Jon Kortnik co processor using a patched HiBasic), maybe someone could write a Positron MOS for this new platform. Programming it would be more like programming on an 8086, with all those FAR calls.. Maybe the Positron would use a variant on an Extended Vector to do some Far Calls... Interesting....
:D
Does anyone think it's possible to patch-up some existing BBC Micro code to have the vectors changed? ie. scan a ROM and look for JSR $FFxx or JMP $FFxx and change them as required, then save a new ROM image?

Just wondering if some ROMs/software were that well behaved or used other means to get to the MOS (push/rts, etc. ?)

(I have a little vanity 65C02 system running which can run BBC Basic, etc. and plan to move to an '816 and a couple of the 816 vectors clash right at the top - I guess I'll just have to "suck it and see" :)

Cheers,

-Gordon
Please find a link to a diagram of the X16's memory map below.
Thanks - but I'm (selfishly) thinking of my own little project, not his... My 65c02 has 64K from 0 upwards with only a 256 byte gap at $FExx for IO. and plans for the '168 are 1-4MB in the same fashion to enable me to 'bootstrap' it in full 65c02 emulation mode.

Is this "Positron" something someone is working on (and I've missed), or just wishful thinking? I could post more about my own little system but haven't as yet as I never thought it was that appropriate for stardot not being real Acorn hardware and all that...

Cheers,

-Gordon
User avatar
davidb
Posts: 2996
Joined: Sun Nov 11, 2007 10:11 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by davidb »

alex_farlie wrote:
Fri Mar 01, 2019 3:31 pm
How easy would it be (given various peoples expertise in FPGA) to build an accelerated "Positron" which is essentially a 65816 style chip, an I/O chip like the Electron ULA, and so on?
You'd have to ask Chris Curry because that sounds a lot like the Acorn Communicator. ;)
User avatar
Andrew_Waite
Posts: 248
Joined: Tue Aug 30, 2016 3:58 pm
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by Andrew_Waite »

David Murray has posted pictures of prototype hardware for his 'Commander 16' computer.

The pictures show his Commodore 64 upgraded with a Gameduino (for video) connected to the C64's user port and dual AY-3-8910 sound chips connected to the machine's expansion bus. The C64 is used for software development. The AY-3-8910s also give 32 I/O lines.

Eventually the Commodore machine will be replaced with a 65816 CPU with 1MB or memory and a PS/2 keyboard connector.

https://www.facebook.com/photo.php?fbid ... ater&ifg=1

https://www.facebook.com/notes/commande ... 397164827/
Last edited by Andrew_Waite on Wed May 22, 2019 4:50 pm, edited 1 time in total.
jregel
Posts: 238
Joined: Fri Dec 20, 2013 6:39 pm
Location: Gloucestershire
Contact:

Re: The 8-Bit Guy's new 65816 based computer

Post by jregel »

David Murray has posted a new video with some updates on the project:

https://www.youtube.com/watch?v=sg-6Cjzzg8s

I thought it was interesting to see how this is developing.
BBC Master Turbo, Retroclinic External Datacentre, VideoNuLA, PiTubeDirect with Pi Zero, Gotek USB Floppy Emulator
Post Reply

Return to “other vintage hardware, software and languages”