W65C265S-Based Computer

[railsrust]

Before I get started, I do have someone who's worked with the 6502 processors professionally before helping me with some things.
He even told me that I should look into getting a logic analyzer before I do too much more.
Most of my reasoning for posting here is because he's also busy helping someone else with their project and in general just has things that are going to take priority over me.
I think we've agreed that KiCad would be a good software to work from.

I've already got some of the main chips picked out to go with the W65C265S, and I fully understood that it had quirks and limitations that I would have to work around.
The graphics chip is the Bridgetek BT816 and audio is going to be split between a Wavetable chip I found, the output of the BT816 and the ton generators of the microcontroller.

The BT816 doesn't use any external ram, it already has a built-in 1 Megabyte ram used for all video and audio data it outputs. The main issues I'm going to have with it are getting the most out of the Quad-SPI command input and how to get it to operate between multiple QSPI data inputs. It acts as a QSPI slave when getting commands from the microcontroller. For all texture and audio data it uses a different set of pins for single SPI sources, but I'd like it to go between at least two.

I went ahead and decided to go with a single parallel 1 Megabyte SRAM so I could keep chip complexity down and the room for code up.

I'm thinking of putting the whole PCB in a Mini-ITX form factor with PC power supply inputs so I can throw it into a spare computer case I have lying around.
With that in mind, I was wondering what would be some good chips I could use to operate a possible cartridge port or backplane with? I was thinking of some sort of ISA type bus or something like it. I want to be able to upgrade this machine without later on if I decide to without having to make a whole new PCB to do so. I saw from a recent post elsewhere that just leaving the cpu data and address buses open to a card edge connector wouldn't be a good idea.

Tell me what you think.

If you want to estimate my knowledge of this stuff I'll tell you what I tell other people when I get into new hobbies; I know just enough to get me into trouble and not enough to get me out.

[GARTHWILSON]

Welcome! We look forward to your success.

Do go through the 6502 primer, indexed at http://wilsonminesco.com/6502primer/ . Even though it's about the '02, much of the material applies just the same. It's organized logically into 22 sections.

A logic analyzer is nice but by no means necessary.

[railsrust]

Thanks! I figured an analyzer wasn't required, but I might as well get one anyways in case I need to troubleshoot things. I was already directed to the primer and had bookmarked it previously.

[DerTrueForce]

Welcome to our 6502 world!
There is an Introduce Yourself thread. Please, do feel free to post there!

For implementing ISA, I think you'd probably need some sort of CPLD. I haven't looked, but I suspect they don't make interface devices for those older standards anymore.
If you're willing to roll your own interface, or look around for a few, you'll probably find more success. 65SIB might be worth looking at, as might SPI-10 and I2C-6. Those last two aren't meant as an expansion bus as much as 65SIB is, though.
What chip you use would depend on what kind of interface you want.
If you want a parallel interface, you could use a 65c22. You might need supporting hardware to increase your I/O addressing beyond 256 bytes, though.
If you're OK with a serial interface, the 65SPI might be worth a look, although that's a CPLD firmware.

I don't know enough about the BT81x to say much of anything, really. I did look at using a Gameduino 3, which presents an ordinary SPI interface, but that's out of stock at the moment. It's also quite different to what you appear to be shooting for(the GD3 provides its own display). I'm not sure if these EVE chips can drive a monitor. I've only ever seen them paired fairly closely with an LCD panel.

I can second what Garth said about not needing a logic analyzer. I still don't have one, and I've gotten my SBC running without it. I've wished I had one, but I never actually needed it.

[BillO]

railsrust wrote:

Thanks! I figured an analyzer wasn't required, but I might as well get one anyways in case I need to troubleshoot things. I was already directed to the primer and had bookmarked it previously.

I'd suggest a dual trace scope would be a much better choice if you don't already have one. USB type devices can be had for well under $100 that will turn your PC into a great way to get a clue as to what is going on. Quite useful.

In fact I have one on order right now that cost less than $55 (shipped) that is supposed to have a 16mHz bandwidth. Just needed something to put on my retro computing bench to help with diagnosing issues without having to tear apart my electronics bench. I'll do a mini-review of it when I get it.

Good luck with the project.

[BigEd]

Welcome railsrust! You'll find lots of expertise and encouragement here.

Certainly it's worthwhile to read Garth's primer, all the way through, at least once. It might be that there are lots of things you don't know, and you don't even know that yet.

It's been said many times here, and often enough ignored, but it really is well worth building something simple first. It's a lot quicker to build something simple, learn while doing, and progress to something more complicated. It may feel like a step down, but it is good advice. Learn to walk before you try to run. Grant Searle's designs are probably a good place to start.

That's an interesting-looking display controller you've found. I think it may be worth a thread of its own: here's one.

[railsrust]

Looking at the 65SIB, I am definitely interested in adding something like that to the PCB. My very reason for wanting an external bus connection is so that I could farm out any additional extensions I might want off to a daughter board containing any extra devices. It is also so I can keep the main board design as simple as possible. In fact, any sound effects will to be played by the video chip, and if I want music I'm going to be putting a Wave Blaster header on there for it to play back general Midi music. I'm going to be using the DreamBlaster S2 sold by Serdaco. It just needs a single Midi TTL line to produce music output from it. The ram is just going to be a single 1 Megabyte SRAM chip. That's all this thing will ever need for what I might do with it.

[railsrust]

In case I don't go with the BT816, could someone suggest a good DAC that could be implemented 'easily'. I basically want to add the ability for voice samples to play if the EVE chip isn't what I want.

[GARTHWILSON]

railsrust wrote:

could someone suggest a good DAC that could be implemented 'easily'.

I don't know if this is what you have in mind, but I've used the simple 8-bit DAC0808 for audio and similar purposes. It has been around for decades. You just put the parallel data bits to it (from a 65c22 for example), and it settles to one-half lsb value in 150ns, about one-seventh of a microsecond. Pretty fast. See http://wilsonminesco.com/6502primer/potpourri.html#DAC, after the heading "Digital-to-analog converters." There are a few other D/A ICs and methods I've used there as well.

[railsrust]

So I got this email from "my little helper" yesterday.

Quote:

So rather than the 816 in any form, what would you think of 4 or more 6502s, closely coupled but yet able to execute independently? If we think of the theme of throwing processors at the problem I could easily adopt the convention that the “Vic Chip” is a Raspberry running a VIC emulator. Or maybe even a 6502 running fast enough to move sprites, set collisions, and yet insulate the main ’02 from the task so in effect an 02 acting as a VIC chip emulator. Heck maybe even one act as a SID by bit banging a PWM.



The fact I am even writing about this means its something that would kind of float my boat, especially if it were expandable.



Heh, run a VIC-20 AND a C64 emulation at the same time. Oh oh, just thought of a smaller array of Z80’s near by.

Then a couple emails later:

Quote:

Heh, and a bargraph led display that acts like a speedometer the more 6502s you use.


I know people have done multiprocessor ’02’s before I wonder if there is anything in open domain, namely the task dispatcher and manager. They also have done a bunch of 8051s which are mores selfcontained.


I gotta believe one FPGA can manage a crapload of ‘02s.

I mean, if he wants to engineer all of this himself... Haha!
I guarantee you've heard his name before, and you might even guess who it is based on his sense of humor.
He's produced a couple of computers before, so I can trust him to make something like that work somehow.
Anyone know of a way to manage multiple 6502s like this?
Keep in mind he was asking me if we could do this instead of the W65C265S.

[GARTHWILSON]

railsrust wrote:

Anyone know of a way to manage multiple 6502s like this?

It's getting off-topic, but that's ok. It's your own topic.

Here are some earlier topics that are very relevant, and kc5tja made valuable contributions on:

• idea: a cell-like chip based on many 6502 cores
• and André Fachat's post in this one (already cued up), about distributing the work load among the various processors in a multiprocessing system: Parallel Processing with 6502s
• See also: Dual Processor SBC.

WDC's W65C02S adds some more signals at the pins, namely ML\ (memory lock not) output (pin 5 on a DIP), BE (bus enable) input (pin 36 on a DIP), and VP\ (vector pull not) output (pin 1 on a DIP).

[railsrust]

I mean, I wasn't trying to go off topic.

He just went and dropped that on me and I was like "sure I guess, you're the expert". I'll tell you who it is in a PM.

[BigEd]

It's interesting, so I've spun it into a new thread:

• Massively parallel 6502 systems

[GaBuZoMeu]

railsrust wrote:

Anyone know of a way to manage multiple 6502s like this?
Keep in mind he was asking me if we could do this instead of the W65C265S.

Back in those days Transputers became popular to some extend. Their four independent "links" (5/10/20 Mb/s serial interfaces) allows pretty much traffic between Transputers. By arranging them in pipeline or planes fashion workload could flood Transputer farms.

The W65C265S' parallel interface bus ("PIB") cannot compete with that but in theory one could attach numerous W65C265S' to one bus where a master could issue workload and fetch results. But it is difficult to estimate what performance can be reached. This interface is poorly documented.

On the other hand - the W65C265S has four serial ports. Using them in a link-like fashion you could arrange the same network topologies as Transputers. It is of course much much slower.

[whartung]

GaBuZoMeu wrote:

The W65C265S' parallel interface bus ("PIB") cannot compete with that but in theory one could attach numerous W65C265S' to one bus where a master could issue workload and fetch results. But it is difficult to estimate what performance can be reached. This interface is poorly documented.

I wish it were better documented, it hints at some interesting features.

Quote:

On the other hand - the W65C265S has four serial ports. Using them in a link-like fashion you could arrange the same network topologies as Transputers. It is of course much much slower.

The 65C134S has a serial network built in to the chip.