One chip MSX improvement project

Prophet (2619)

08-05-2008, 17:33

Only a bare 5% free LEs?. I vote for a better MSX2+ compatibility (as much as possible) and bug fixing; there's no room at all to start big developments...

Por HRA!

Champion (289)

08-05-2008, 17:53

I think that current TZ80, ESE-VDP and ESE-PS2 are ugly circuits.
However, correcting it at a time repeats the refactoring little by little because
it is difficult.

I plan to divide into a more detailed unit.
And, the unit test verification is executed.
Stability, interchangeability, and free LEs are sure to be obtained.

Por SaebaMSX

Hero (533)

09-05-2008, 20:35

I think that current TZ80, ESE-VDP and ESE-PS2 are ugly circuits.

Could be the last part the reason because some keyboards have problems with OCM? I don't mind that one of my keyboards is not working, but this is a real pain seeing that some keyboards have some capabilities in PC (diagonal arrows + space bar) and it does not work in OCM. Of course this is not mandatory... but could you take a look?

And, the unit test verification is executed.
Stability, interchangeability, and free LEs are sure to be obtained.

That information is indeed very good. As you said OCM gets unstable when not many LEs are left. Having a good MSX2+ is more than enough for me, but what you want to get impress me!

Por Randam

Paragon (1431)

10-05-2008, 12:42

HRA!, is it possible to make those usb ports on the OCM usable? Would be great if you could connect a key board, mouse or other peripherals to it.

Por HRA!

Champion (289)

10-05-2008, 15:24

HRA!, is it possible to make those usb ports on the OCM usable? Would be great if you could connect a key board, mouse or other peripherals to it.

It is not possible to achieve it with free LE of OCM.
It is likely to be able to achieve it if something is deleted.

However, it is difficult to achieve complete USB-HOST.
Because it is necessary to add special CPU to the control.
It will become the biggest circuit in OCM.

If it is a circuit only for the mouse (only for the joystick ), it becomes simple a
little more.

This cannot feel happy so much

Por HRA!

Champion (289)

11-05-2008, 11:31

If USB-Floppy was able to be achieved, it is happy.
I will examine it only a little.

Por HRA!

Champion (289)

11-05-2008, 16:57

New update PLD:
http://www5d.biglobe.ne.jp/~hra/note/onechipmsx/files/emsx_top_20080511_001.zip

Updates:
(1) MSX2+ BIOS implemented (caro's IPL.ROM and compressed ROM image)
(2) The environment of Altera DE1 is added to the archive.
(3) The composition of VDP source is changed.
(4) The folder composition is changed.
(5) ToDo.txt updated.

The timing relation of VDP was able to be arranged considerably.
Some unit tests were added.
The timing bug can be analyzed by this :D

Por KdL

Paragon (1450)

11-05-2008, 17:05

(GNU LGPL) VHDL S/PDIF:
www.webalice.it/gnogni/ocm/misc/spdif_interface.rar ;)

---

From my old E-mail of 2007-06-19:

[KdL]
Over 1chipMSX the noise of sound output is very high when I connect the VGA cable!
moreover if I detach it there is no one noise.. why?
 
(for e.g. when you write a "COLOR15,15,15" the noise is to max!!!)
 
A digital coaxial cable is a good solution for me.
Is it possible to add a new circuit to have a S/PDIF output?

[HRA!]
Certainly. It is a good idea.
However, there are some problems.

1:
Because an internal voice of 1chipMSX is generated with the rate of 
3.58MHz, the circuit that does the down sampling to 192kHz is necessary.

2:
Because the empty capacity of FPGA installed in 1chipMSX few left, a big 
low-pass filter cannot be added.
If the down sampling is done without using the low-pass filter, it 
becomes a noise by
generating alias.

3:
An additional circuit such as digital communication terminals is needed, 
and light-hearted is lost.

It positively examines it.

Por SaebaMSX

Hero (533)

12-05-2008, 11:18

I've been testing these new versions.

I have something to say about using main RAM 4MB: Some (only european?) software won't detect it correctly and say that you have less memory than expected. Of course this is a problem in the software, but will prevent it from loading sometimes.

Example: Unknown Reality. This demo needs 256KB of Main RAM to work, but if internal memory is 4MB it will detect less than 256KB and only music menu loader will work.

I've tested Unknown Reality with previous firmware (2MB RAM) version and I think something is not emulated correctly. I do remember that when v9958 was present this demo had a problem during the scrolled text with blue background (where the sample waves were kept in the right side). This "problem" is not emulated... I think you can check this with a real turboR.

Maybe someone more could explain this behaviour from a coder point of view.

Por Manuel

Ascended (19310)

12-05-2008, 11:31

About UR: that 'real time scope' part does not really suffer from the V9958, but more from slightly different timing on turboR.
You're right about the 4MB RAM: 2MB is safer and still more than plenty.