Raven. A homemade Atari-like computer

[peters]

Maybe you can answer this question about the 68060.

I read somewhere many years ago that the 060 can run at 4x bus speed. I may have misunderstood what the docs were telling me.

Would this mean that instead of the Milan running at 30/60MHz, it could run at 25/100MHz ?

[Cyprian]

I read somewhere many years ago that the 060 can run at 4x bus speed.

Isn't 68060 bus asynchronous? And e.g. CT60 with 90MHz works fine with Falcon's 8MHz bus?

[agranlund]

Maybe you can answer this question about the 68060.

I read somewhere many years ago that the 060 can run at 4x bus speed. I may have misunderstood what the docs were telling me.

Would this mean that instead of the Milan running at 30/60MHz, it could run at 25/100MHz ?

Yes that's right. PCLK can be 1x, 2x or 4x of that of BCLK.
So in theory, yes, it should be able to do 25/100mhz.
How hard such a modification would be in practice on a Milan is anyones guess :)

The 060 takes PCLK and generates an internal "virtual" BCLK based on an externally feed bus-enable signal.
So whatever logic it has that makes the enable signal would have to be changed. And who knows if there are other dependencies to deal with.
The 040 is a bit different and takes both PCLK and BCLK directly.

I'm guessing the Milan is an 040 design with upgrade card for 060? If so, then it's probably the cpu card that generates bus-enable from the motherboards 040 bclk as hinted by examples in the 68060 user manual.


For all details surrounding BCLK see MC68060 User Manual; 7.2 FULL-, HALF-, AND QUARTER-SPEED BUS OPERATION AND BCLK

[peters]

Yes the Milan uses a small add on board that plugs into the cpu socket.

I think Rustynutt used an Amiga board to upgrade his Milan and he was looking at using an MC88LV926 to control timing. I've no idea what that does.

[peters]

I don't think this is the exact document I read, it was a long time ago.

https://websrv.cecs.uci.edu/~papers/mpr ... 080502.pdf

[agranlund]

Congratulations Anders! That's a remarkable achievement. Genuinely.
My hat, and any future hats I may don, are well and truly tipped to you.

Thank you sir! This has been, and continues to be, a great learning experience.

Looking at the previous little computer it's quite a big jump in usability.
That one was really fun too and sparked the interest to make what eventually ended up being Raven.

68k_4.jpg

We live in a great time where one can so easily find all kinds of technical information and educational videos on just about any subject.

[JezC]

Are you prepared to share some info on that earlier board as well? Looks interesting too!

I'd love to get & build a Raven someday but financial pressures and lack of time make that a pipe dream at the moment....

:sigh:

[agranlund]

Are you prepared to share some info on that earlier board as well? Looks interesting too!
I'd love to get & build a Raven someday but financial pressures and lack of time make that a pipe dream at the moment....

Those Rosco-m68k single board kits are probably much better in every possible way but it wasn't really about the end result.
More about the process of making and then getting things up and running. I think the same is true for the Raven board as well (the V4SA is a better and faster Atari clone already).

Anyway, that little computer is a bit too rough to Github and I really don't think it would contribute much -- but it was really fun to make and super rewarding when it came to life and could be interacted with. Highly recommend, regardless of level of complexity :)

It doesn't have an operating system as such, just a very simple boot rom that lets you interact with it over the serial port and upload code to run.
The board was designed around a handful of chips that I already had lying around in a box and its basically just cpu, some small amount of ram, rom, a single channel UART and an MFP. It takes either a full 68000 cpu, or a 68008 through an adapter board and I only ever tried it with the 68008 (I never actually got around to populate and test the MFP either).

I figured since that little machine actually worked then surely I should be able to repeat that but for something bigger and a bit more advanced too.

[agranlund]

Got myself an Adlib reproduction card for testing and updated the mxPlay plugin to better support OPL2 hardware.
I had only tested on hardware with OPL3 chip before and the guesstimated write timings where a bit too aggressive for the older chip.
Took the opportunity to have the plugin expose some song info to the player while at it.

The plugin(s) live here but who knows, it might make sense to get them moved into the main mxplay repo instead since this stuff works on other Atari clones with ISA ports too.

https://github.com/agranlund/atarisound

mxplay_opl.jpg

opl2_card.jpg

[nokturnal]

Yes, OPL2 has different timings for setting address data, because it's internal clock is much slower than on OPL3.. From my notes:

; OPL2 timings
; Master clock = 3,58 Mhz
; minimal time after address write mode = 12 cycles
; minimal time after data write mode = 84 cycles

; 12 / 3,58 = ~ 3,3 µs after the address write
; 84 / 3,58 = ~ 23,5 µs after the data write

; OPL3 timings
; Master clock = 14,32MHz
; minimal time after address write mode = 32 cycles
; minimal time after data write mode = 32 cycles

; 32 / 14,32 = ~ 2,23µs after the address write if we write data
; 32 / 14,32 = ~ 2,23µs after the data write

If cpu is much faster we need to introduce wait states. I generate nops in rept depending on cpu clock(8/16MhZ), but I will need to exchange it to something more elaborate in the future.