MiSTer PCXT

[jordi]

A game I have been having trouble with is Castle Master on the PCXT core. I have tried with several different settings. Without EMS or high memory, with EMS and high memory, with PCXT, with Tandy 1000, with Tandy graphics and with CGA graphics.

But Castle Master always freezes on the first screen. I tried using Putty to remove into the MiSTer, kill the MiSTer process and start it in the Putty console so I could try to debug it, but there are no errors thrown on the MiSTer Main side when the game freezes.

But when I try it on 86Box running as a PCXT, it seems to have no problem getting into the game.

Maybe I should try to see if this is a problem for all the Freescape games.

Better open an issue in GitHub

[Caldor]

I testet all the Freescape games I could find... and all of them work. So I tried a different copy of Castle Master and it then runs fine as well, both CGA and Tandy

Awesome. I have copied games around without any read or write errors. Also tried using a 1MB RAM disk to copy games to to help test if there might be problems caused by disk speed. The core overall seems very stable now. I mainly test with a setup that uses 4DOS, EMS, high mem and a RAM drive. It has 610kb conventional memory free. I a boot menu that makes it easy to use different setups for testing different things. Still have not been able to get Area5150 to run without problems, but at least most of it works when running at 7mhz. It is more important that the games and apps work though and all the games that should be able to run on PCXT or Tandy seems to run on this.

I had trouble getting Time Bandit to run with Tandy graphic. I will try testing that some more, maybe I can find another version of that game that will run fine as well.

I still encounter that Tandy games remember the graphics from the last Tandy game you played if you play more than one, but with the latest version of the core, its usually just an image that lingers for a few seconds and then the graphics of the new game replaces them.

There is still also a problem with lines going through all Ascii. Its most noticeable when running Norton Commander or Volkov commander... I am guessing anything with Ascii graphics that changes the color of the full... char I guess it must be, so its not black, will have a black line at the bottom I am thinking it is. Because you see one line going through each char when looking at Volkov Commander or Norton Commander. Another example is Time Bandit where the boot sequence has some ascii graphics and even some animation done with ascii characters. But with Tandy mode enabled there are lines through everything. I guess it might be a problem with a part of the core that is 3rd party and might a problem to fix, but I am reporting it either way in case it might help.

[spark2k06]

I still encounter that Tandy games remember the graphics from the last Tandy game you played if you play more than one, but with the latest version of the core, its usually just an image that lingers for a few seconds and then the graphics of the new game replaces them.

This is because the different memory pages are shared with RAM, and if the game, before displaying them, does not delete the content, this will temporarily happen. What could perhaps improve the core in this sense is that during a hard reset the RAM is cleared, but it is certainly not a priority.

There is still also a problem with lines going through all Ascii. Its most noticeable when running Norton Commander or Volkov commander... I am guessing anything with Ascii graphics that changes the color of the full... char I guess it must be, so its not black, will have a black line at the bottom I am thinking it is. Because you see one line going through each char when looking at Volkov Commander or Norton Commander. Another example is Time Bandit where the boot sequence has some ascii graphics and even some animation done with ascii characters. But with Tandy mode enabled there are lines through everything. I guess it might be a problem with a part of the core that is 3rd party and might a problem to fix, but I am reporting it either way in case it might help.

This is common on original Tandy 1000 machines, you can use Tandy's MSDos 3.2 mode 200 application in the autoexec.bat to avoid it:

https://www.dosdays.co.uk/computers/Tan ... dy1000.php

The Tandy 1000 can switch between 80-column text mode and 40-column text mode by simply pressing the F12 key. The default 80-column text mode on the Tandy annoyingly uses 225 lines. IBM uses a 200 line text mode. All graphics modes in these machines use 200 lines. In order to get a proper aspect ratio with the graphics modes, the monitor's vertical size control has to be set almost to the point where the graphics are almost hidden by the monitor bezel. When you return to the text mode, this winds up cutting off either the top, bottom or both lines of the display. Use Tandy MS-DOS 3.2 mode.com command: mode 200, to fix this. This does not work in the TL machines.

Before thinking that this is a core problem, I strongly recommend running the same tests on an emulator such as 86box, which is based on PCem, by selecting the same system and CPU from the emulator options

86box_tandy.png (27.98 KiB) Viewed 32020 times

[TheCorfiot]

This is a fascinating core, especially the Tandy 1000 support
Are there any nice hard drive images available to download similar to the ao486 core

[Caldor]

Ahh, interesting. I had not really considered it might have been how the original hardware was meant to work that Tandy 1000 had those lines.

Then it seems the core must already be more than 90% complete?

I will try to run some more things in 86Box. I do not think I setup a Tandy 1000 machine on it yet. I will try to use that to debug Time Bandit.

[Swainy]

Total newbie to this core. How do I set it up?

[wark91]

you have some informations on the the readme to help to set it up : https://github.com/MiSTer-devel/PCXT_MiSTer#readme

[Swainy]

Ok, someone has been very kind and sent me scripts that made the roms up for me.

I’ve currently got it set like this in my PCXT games folder:

iO5TKsC.jpg (682.29 KiB) Viewed 29969 times

However, after the splash screen I get the following message:

640 KB OK
601
ERROR. (RESUME = “F1” KEY)

How do I stop getting this error?

Thanks!

[NML32]

Ok, someone has been very kind and sent me scripts that made the roms up for me.

I’ve currently got it set like this in my PCXT games folder:


However, after the splash screen I get the following message:

640 KB OK
601
ERROR. (RESUME = “F1” KEY)

How do I stop getting this error?

Thanks!

I sent you a PM.

[wark91]

Ok, someone has been very kind and sent me scripts that made the roms up for me.

I’ve currently got it set like this in my PCXT games folder:

However, after the splash screen I get the following message:

640 KB OK
601
ERROR. (RESUME = “F1” KEY)

How do I stop getting this error?

Thanks!

This error "601 ERROR" is normal for IBM 5160 bios because real floppy implementation is missing. You just need to push F1 key to pass it.

[Caldor]

Ahh, that is good to know. Was probably my scripts and I have been wondering if maybe I had done something wrong when creating the ROMs, but I was thinking it would be strange if the problem was something like this rather than just not working at all. Because the way the ROMs are put together is just putting one file after another. If its the wrong order they get put together, it would be strange if the core would accept it at all.

[spark2k06]

I have made two updates, but I have not been able to get them to work properly... I have created two branches with what has been implemented so far:

1. Upgrade of the UART module, from 16550 to 16750, available in OpenCores.org:

https://github.com/MiSTer-devel/PCXT_Mi ... -uart16750

I have instantiated the module in the same way as the 16550 we are using so far, but it does not recognise the HDD images... with this new module I was hoping that the problems regarding access to them would improve, especially at the maximum speed of the UART (921.6Kbps).

2. Mouse support over the COM2 port, based on the great work Antonio has done with the PS/2 to Serial wrapper he has developed:

https://github.com/MiSTer-devel/PCXT_Mi ... com2-mouse

For now, I am using a 1.844262 MHz clock via PLL for clk_uart for both ports, so that it is compatible with the 1200 baudrate required by the serial port. So, the disk image must be set to 115.2Kbps for it to be recognised.

The serial port is working correctly, the problem here seems to be the handling of the PS/2 signal coming from the MiSTer mouse emulator:

mouse_com2.jpg (58.77 KiB) Viewed 31523 times

Although mouse events are detected and generated, the results are not correct because it performs random movements and records mouse button presses when they are not actually taking place.

@kitune-san, maybe you could give us a clue on this, I would appreciate it... I am stuck on these points.

[suww37]

I have made two updates, but I have not been able to get them to work properly... I have created two branches with what has been implemented so far:

1. Upgrade of the UART module, from 16550 to 16750, available in OpenCores.org:

https://github.com/MiSTer-devel/PCXT_Mi ... -uart16750

I have instantiated the module in the same way as the 16550 we are using so far, but it does not recognise the HDD images... with this new module I was hoping that the problems regarding access to them would improve, especially at the maximum speed of the UART (921.6Kbps).

2. Mouse support over the COM2 port, based on the great work Antonio has done with the PS/2 to Serial wrapper he has developed:

https://github.com/MiSTer-devel/PCXT_Mi ... com2-mouse

For now, I am using a 1.844262 MHz clock via PLL for clk_uart for both ports, so that it is compatible with the 1200 baudrate required by the serial port. So, the disk image must be set to 115.2Kbps for it to be recognised.

The serial port is working correctly, the problem here seems to be the handling of the PS/2 signal coming from the MiSTer mouse emulator:

mouse_com2.jpg

Although mouse events are detected and generated, the results are not correct because it performs random movements and records mouse button presses when they are not actually taking place.

@kitune-san, maybe you could give us a clue on this, I would appreciate it... I am stuck on these points.

Great. I hope you succeed.

[Bas]

Completely unrelated question. How does the OS that runs inside the core determine the CHS-layout of the hard drive image that's attached to it?

[spark2k06]

Completely unrelated question. How does the OS that runs inside the core determine the CHS-layout of the hard drive image that's attached to it?

All the magic is done by the XTIDE BIOS, the inner workings of it, I can't tell you, I haven't reviewed the source code.

[Caldor]

Any news about a PCXT specific forum?

[breiztiger]

Here isn’t already pcxt specific ?

[spark2k06]

Any news about a PCXT specific forum?

This is the latest news I have:

Yes this is possible in the next couple of days I will create a new forum section for the core.

Thanks

Xbytez

Let's give him time to create a specific forum

[Xbytez]

Let's give him time to create a specific forum

All done!

[MicroCoreLabs]

I have instantiated the module in the same way as the 16550 we are using so far, but it does not recognise the HDD images... with this new module I was hoping that the problems regarding access to them would improve, especially at the maximum speed of the UART (921.6Kbps).

Was there any progress on the UART flow control signals? They may be needed to reliably run at higher bit-rates.

Can I ask which direction the errors are most seen? From the HPS to the core or the other way?

Maybe increasing the size of the UART RX FIFOs may make a significant difference?

[spark2k06]

Was there any progress on the UART flow control signals? They may be needed to reliably run at higher bit-rates.

Can I ask which direction the errors are most seen? From the HPS to the core or the other way?

Maybe increasing the size of the UART RX FIFOs may make a significant difference?

Indeed, I can confirm that serdrive makes use of the CTS/RTS signals:

https://github.com/spark2k06/Main_MiSTe ... uxSerial.h

Code: Select all

state.c_cflag |= CRTSCTS | CLOCAL;

So I would like to see how an upgrade to 16750 behaves, also, there is something observed on the current 16550 module, which happens on both the PCXT core and ao486 which shares the same UART module... and that is that the COMTEST application is not detecting the IRQ values, which should be 3 and 4, instead it shows a question mark symbol:

comtest.jpg (58.77 KiB) Viewed 31158 times

However, I will try to see if I can increase the size of the FIFO queue. The ultimate goal is to see if we can make the 921.6Kbps speed more reliable, because at 460.8Kbps it generally works very well.

I'm not sure, but I think the errors occur from the HPS to the core, because they are usually image reading errors. So yes, maybe increasing the FIFO queue will improve the issue.

[spark2k06]

However, I will try to see if I can increase the size of the FIFO queue. The ultimate goal is to see if we can make the 921.6Kbps speed more reliable, because at 460.8Kbps it generally works very well.

I'm not sure, but I think the errors occur from the HPS to the core, because they are usually image reading errors. So yes, maybe increasing the FIFO queue will improve the issue.

I am opening a new thread on this issue:

viewtopic.php?t=5366

[Caldor]

Here isn’t already pcxt specific ?

It is now

[riblad_om]

The TL/2 was my first PC as a child, how close is this core to that in terms of graphics and sound? From reading comments my guess is that later 1000s had sound chips that this core doesn't implement yet, but I'm not sure.

[spark2k06]

The TL/2 was my first PC as a child, how close is this core to that in terms of graphics and sound? From reading comments my guess is that later 1000s had sound chips that this core doesn't implement yet, but I'm not sure.

In terms of sound, it has an SN76489AN Compatible Implementation (Tandy Sound), and Adlib is also available through the JTOPL2 module.

On the graphics side, only the 640x200x4 and 160x200x16 modes are missing. In Tandy, these are less used modes than 320x200x16, which is available.

[Hodor]

The TL/2 was my first PC as a child, how close is this core to that in terms of graphics and sound? From reading comments my guess is that later 1000s had sound chips that this core doesn't implement yet, but I'm not sure.

In terms of sound, it has an SN76489AN Compatible Implementation (Tandy Sound), and Adlib is also available through the JTOPL2 module.

On the graphics side, only the 640x200x4 and 160x200x16 modes are missing. In Tandy, these are less used modes than 320x200x16, which is available.

For the sake of completion and preservation it would be nice to add the missing graphics modes. I don´t mean to be now because there are several important things unimplemented like floppy or IDE support, but later on.

[riblad_om]

The TL/2 was my first PC as a child, how close is this core to that in terms of graphics and sound? From reading comments my guess is that later 1000s had sound chips that this core doesn't implement yet, but I'm not sure.

In terms of sound, it has an SN76489AN Compatible Implementation (Tandy Sound), and Adlib is also available through the JTOPL2 module.

I apologize, I'm ignorant on sound chip names. I found a video of the Space Quest III intro playing on a TL/2, and the sound is the same except the digitized samples are missing. I'm guessing I will have to wait if/until later sound chip revisions are supported by the core?

[spark2k06]

I apologize, I'm ignorant on sound chip names. I found a video of the Space Quest III intro playing on a TL/2, and the sound is the same except the digitized samples are missing. I'm guessing I will have to wait if/until later sound chip revisions are supported by the core?

The current implementation is the Tandy 1000, if we find a sound module compatible with the TL/2 variant, there would be no problem to upgrade.

[alkenstein]

Hey there,
I found this page by googling to see if MiSTer was compatible with 8088 Mph impressive stuff! I was checking because 86Box devs are also working on this, and was wondering in case anyone might be interested to share info to help. The 86Box discord is https://discord.gg/v5fCgFw in case.
Best wishes,
alk

[somhi]

Is there any solution for bending the top row of the screen in the 15khz sony pvm monitor? I can't use any method. It's no use adjusting the mister.ini settings. Does anyone know why this screen is coming out?

Do you get also the same result when trying Tandy games ?

In my case I noticed more bending problems with CGA resolutions.