A Better 65C265 Board

[BitWise]

Today the postman bought me some new chips (from WDC indirectly).

Chips galore

I thought I'd build something like the 65C265SXB but with more SRAM (1 or 2MB), a faster oscillator (at least 8Mhz), EEPROM and maybe a bit of memory mapping in the first 32/64K region

[Martin_H]

Neat, looking forward to seeing what you build.

[whartung]

An 8Mhz W65C265SXB with more RAM would be really nice. I've contemplated picking one of those up and making a RAM card for it, but I don't know anything about board layout and all that to make a plug in board for one. Ideally it would be really straight forward with the modern high density static RAMs. Not sure how much decoding logic it would need, not much I would think because of the CS lines on the MCU. But the MCU already has enough I/O for my purposes, I'd just like some more RAM. The 8MHz wouldn't hurt either.

The $30 board just kind of crippled itself with the CS lines. But the other MCU board is still < 4Mhz. I think the monitor only has setting for up to 6+MHz in terms of having settings for the baud rate generator using the in built monitor. So I don't know what would be involved with getting an 8MHz one up.

I tried to make some heads or tails around the PC board design tools I could find for the Mac, but they're not for the casual explorer it seems.

[handyandy]

For what it's worth, I've converted the 265romlist.pdf file to plain text. I fixed any typos I found and made it as readable as I could. I've gone over it three times so far and fixed some errors I found; there may be more. The file was for a 2600AD assembler so further editing may be necessary for it to be fed to another assembler of choice. All of the code is "AS WAS"; I didn't make any changes to it but saw lots of places where code could be optimized. It looked like several people were involved with the project judging from the style of coding. I picked this thread since I thought a better 65c265 board should have better rom or ram code...

Cheers,
Andy

edit: further editing and checking for errors: 265iromlist_a.txt
I placed (???) and initials adw at points of interest 2/24/2019 and 3/3/2019

265iromlist_a.txt

(153.04 KiB) Downloaded 127 times

If I were to assemble I would use ORCA/M assembler and edit to suit. Not there yet...

[BitWise]

This is still on my list of projects but sadly its been pushed down a long way.

Last week I was sorting out a shopping bag full of all the bits and pieces I'd bought in the last year or two that was cluttering up my study and found those chips along with some GY-521 3-axis gyros, FPM10A fingerprint readers, CH376S USB and DS3234 RTC modules, HC-05 and HM-10 Bluetooth modules, hx711 load sensors, some Intel 8085 MPUs and AMD 9511 FPUs.

Quite a bit of my spare time is spent helping out on other people's projects like this NeoPixel based art project ..
https://www.youtube.com/watch?v=T_NLTu2_PWY
and the multi-processor (2x PIC24 + 2x PIC16) performance monitor I'm building to attach to this open hardware rowing machine ..
https://openergo.webs.com/

I'd like to get back to this 65C265 project later in the year.

[whartung]

That's great to get that monitor listing out of that awful PDF.

Have you assembled it?

Have you had a chance to check it against the ROM? (I don't have the machine myself, just curious.)

[handyandy]

No I haven't assembled it or checked it against ROM yet...

Cheers,
Andy

[handyandy]

It should be possible to modify the 265iromlist to implement a "loader" e(e)prom (cs3b) in lower 32k to load ram in upper 32k bank 0 (cs4b) using the $800 vector in internal rom. It's puzzled me as to the purpose of this vector for a while. This would be one use of it. The cs3b "rom" could then be switched out with either cs5b ram or an alternate cs3b ram. I personally prefer ram over rom in bank 0. I also don't particularly like storing user vectors, etc. in page 1 stack memory. In a minimalist way I can understand why but I think it's just asking for trouble.

Cheers,
Andy

[whartung]

I think the 0800 vector is more a development thing than anything else, since RESET will reset all of the flags and what not on the chip, but no memory itself. Lets a developer gain quick access to the boot process.

The nice thing about the ROM switch out, is that if you have RAM shadowing it, you can copy the ROM out to low RAM, disable it (thus enabling the RAM) and then copy it back.

The vectors being in 0100 is simply because that's where the memory is. The chip only has 5xx bytes of RAM.

In '816 mode, it's really unimportant. As 0100 is no more special than any other spot of Bank 0. In '02 mode, then, sure. It constrains the stack. But even then, not that much. That has to be a really deep stack to start bonking in to the vectors. On your own ROM, you can move them wherever you like. It would be nice if it didn't blot out 32K of Bank 0 for the ROM, however.

[BigEd]

Related, the "8-bit guy" (YouTuber David Murray) has started a year-long project to build an '816 based retrocomputer, and seems to be using a '265 as the basis. See here:
https://stardot.org.uk/forums/viewtopic ... 45&t=16544

[whartung]

Good luck to him, but I think he's in for a rude shock with the video, as we've seen everyone over the years struggle with this aspect of the design.

He's also looking for a product, not a hobby project. Which is fine, more power to him.

That Gameduino thing looks really nice, actually.

It's interesting that apparently the small sub set of folks that he talked to only care about vintage computers (i.e. C64) as gaming machines. Getting any real work done isn't on the radar.

I like and play games like anyone else, but I've always been more interested in making the machines do Real Work(tm).

But if in a year he shows up with a working machine that can take an SD Card, and a keyboard/VGA monitor, with some real (> 64K) memory for $50-$100, I might be interested.

[drogon]

I'm following Dave's "commander 16" group on Facebook on that one. (and have seen one or 2 of our members here on it too

I do hope it succeeds, however right now it's full of people wanting this, that, the other. I hope Dave can lock it down and get the specs. sorted quickly. I have exchanged some emails with him about it all. His idea of using the Gameduino (or really a big FPGA) as the graphics chip accessed via SPI or a parallel port is the same as that memory chip that I mentioned to the other recent project here a while back - and the same as my Pi Zero graphics solution that I'm going to roll out on my '816 system - the same as long as you treat it a a "black box" you poke commands at to draw lines, sprites, etc. rather than a traditional poke pixel into RAM type of thing.

I'm also following the Foenix project - which was created in response to 8-bit Daves initial request early last year about his perfect 8-bit micro too, although that project seems to have slowed down somewhat, but it looks very intersting - an '816 with some FPGAs to handle VGA graphics, sound and DMA to stuff..

I've stared blogging about my Ruby 6502 project - mostly for my own documentation than anything else right now but I'm getting a lot of good feedback on it. Seems there are still lots of people interested in the "retrobrew" world...

-Gordon

[drogon]

whartung wrote:

It's interesting that apparently the small sub set of folks that he talked to only care about vintage computers (i.e. C64) as gaming machines. Getting any real work done isn't on the radar.

I don't quite "get" the whole C64 thing, however no-one outside the UK really gets the BBC Micro thing, so there is probably a big sway of cultural identity. Also, possibly the age of people too - the C64 being half a generation after me - I was in my 20's when it came out (and it was late in the UK too). My Ruby 6502 system is more like a BBC Micro than an Apple II which probably reflects the computers I mostly used back then.

-Gordon

[Arlet]

whartung wrote:

That Gameduino thing looks really nice, actually.

Earlier in the video he says "No FPGA". So I really started to wonder how he would get video, and then it turns out it's got an FPGA after all.

[BigEd]

It'll be interesting to see if that's still true by the end. At present he's plugging in existing things so he can do development. He needs something VLSI - either a current chip, or an FPGA, or maybe a microcontroller. But it might be the case that he meant he doesn't want the whole system to be inside the FPGA. Maybe he'd be happy with just a subsystem.