Daydreaming: a real VGA display card for the Atom

discussion of games, software, hardware & emulators relating to the Acorn Atom
User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Fri Jun 17, 2016 4:16 pm

Hi Guys,

On the Atom Connected Pi Zero, the 65C02 Co Pro is now running together along side the VGA resolution display, which means I can test with BBC Basic programs that do nice graphics.

I've been playing with the GPU today, and managed to get GPU accelerated "PLOT 85" triangle drawing working:
IMG_0470.JPG

It's actually not that fast (maybe 10 triangles per sec) because there is a huge amount of overhead to set up the GPU to draw one triangle!

The GPU driver code is horrendous, because it's cobbled together from bits and pieces from several sources:
https://github.com/hoglet67/PiTubeDirec ... v3d.c#L157
This is actually the first time I've directly programmed a GPU, so it's actually been quite interesting.

The full datasheet for the Pi's GPU is here:
https://www.broadcom.com/docs/support/v ... G100-R.pdf

Dave

User avatar
oss003
Posts: 2515
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby oss003 » Fri Jun 17, 2016 4:24 pm

WOW Dave :shock: :shock: ....... nice job =D> =D> =D> =D> =D>

Does a faster co-pro speedup the GPU?

Greetings
Kees

User avatar
jgharston
Posts: 2658
Joined: Thu Sep 24, 2009 11:22 am
Location: Whitby/Sheffield

Re: Daydreaming: a real VGA display card for the Atom

Postby jgharston » Fri Jun 17, 2016 4:52 pm

hoglet wrote:On the Atom Connected Pi Zero, the 65C02 Co Pro is now running together along side the VGA resolution display, which means I can test with BBC Basic programs that do nice graphics.
I've been playing with the GPU today, and managed to get GPU accelerated "PLOT 85" triangle drawing working:

Well done!
On that theme I updated the VDU 25 (PLOT) page on the Wiki yesterday (and the VDU 19 page earlier today).

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.25
(C) Copyright J.G.Harston 1989,2005-2015
>_

User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Fri Jun 17, 2016 4:59 pm

jgharston wrote:
hoglet wrote:On the Atom Connected Pi Zero, the 65C02 Co Pro is now running together along side the VGA resolution display, which means I can test with BBC Basic programs that do nice graphics.
I've been playing with the GPU today, and managed to get GPU accelerated "PLOT 85" triangle drawing working:

Well done!
On that theme I updated the VDU 25 (PLOT) page on the Wiki yesterday (and the VDU 19 page earlier today).

That's really helpful, thanks.

Is it worth clarifying which systems support which plot 25 commands?

Dave
Last edited by hoglet on Fri Jun 17, 2016 5:08 pm, edited 1 time in total.

User avatar
jgharston
Posts: 2658
Joined: Thu Sep 24, 2009 11:22 am
Location: Whitby/Sheffield

Re: Daydreaming: a real VGA display card for the Atom

Postby jgharston » Fri Jun 17, 2016 5:06 pm

hoglet wrote:
jgharston wrote:On that theme I updated the VDU 25 (PLOT) page on the Wiki yesterday (and the VDU 19 page earlier today).

That's really helpful, thanks.
Is it work clarifying which systems support which plot 25 commands?

Yes, I'm going through that next.

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.25
(C) Copyright J.G.Harston 1989,2005-2015
>_

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Fri Jun 17, 2016 5:08 pm

:shock: Magnificent =D>

The Atom can't draw 10 triangles like these in a second so it's quite fast :D
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
jgharston
Posts: 2658
Joined: Thu Sep 24, 2009 11:22 am
Location: Whitby/Sheffield

Re: Daydreaming: a real VGA display card for the Atom

Postby jgharston » Sun Jun 19, 2016 7:49 pm

I've added a Graphics test to the BBC VDU driver test programs.

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.25
(C) Copyright J.G.Harston 1989,2005-2015
>_

User avatar
oss003
Posts: 2515
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby oss003 » Tue Jul 12, 2016 6:24 am

Dave,

just a thought about speeding up scrolling, I don't know how it is working at the moment but can you do scrolling just with changing the screenaddress pointers? I guess you did the same with the hardware scrolling in the Godil?
This way you don't have to move all screen- and colourdata.

Greetings
Kees

User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Tue Jul 12, 2016 6:58 am

oss003 wrote: just a thought about speeding up scrolling, I don't know how it is working at the moment but can you do scrolling just with changing the screenaddress pointers? I guess you did the same with the hardware scrolling in the Godil?
This way you don't have to move all screen- and colourdata.

I don't think it's that easy; on the Pi the GPU seems to choose the base address for the frame buffer. You have control over the resolution, colours, etc, but not directly over the base address. You can control it's alignment, but only certain values are supported.

Besides, even if you could move it, I don't think it would wrap correctly.

It might be possible to use the GPU to do this work. I haven'y looked into that yet.

Are you planning to connect one of these up?

Dave

User avatar
oss003
Posts: 2515
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby oss003 » Tue Jul 12, 2016 7:04 am

I do have a RPi running Optima at the moment.
Maybe I'll try to connect it later but first I have to finish some other projects ..... :wink:
I have too many unfinished projects going on at the moment.

Greetings
Kees

User avatar
tricky
Posts: 1811
Joined: Tue Jun 21, 2011 8:25 am
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby tricky » Tue Jul 12, 2016 12:29 pm

I haven't looked too closely at the Pi's GPU as I am involved with designing them for PowerVR.
I would think it should be easy enough to write a shader to scroll (blit) the screen in not many more instructions that drawing your triangle.
It should also be easy enough to write a shader which, reads a beeb/atom byte format screen and decodes it to colours on a display (jsbeeb may do this).
Just in case Hoglet doesn't have enough skills, I thought adding 3D gfx would be good :lol:

User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Tue Jul 12, 2016 1:05 pm

tricky wrote:I haven't looked too closely at the Pi's GPU as I am involved with designing them for PowerVR.

I've actually don't have any experience with 3D Graphics, so any pointers you (or anyone else) can offer would be most welcome.

The VideoCore IV documentation is publicly available:
https://www.broadcom.com/docs/support/v ... G100-R.pdf

There is also at least one relevant patent:
http://www.google.co.uk/patents/US8692848

Overall the VideoCore IV GPU seem to be a tile mode renderer, which I think means the frame is broken up into a grid of largish tiles, that can then be rendered in parallel on multiple processors. Prior to rendering there is a binning phase, which decides which primitives overlap which tiles.

Everything seems to be geared up to rendering 3D shapes (not surprisingly I guess) from scratch.

What I can't figure out is how you would do something as mundane as scrolling, i.e. just copying one already-rendered part of the frame buffer to a another part.

Is there some trick to doing this that I'm missing?

I suspect I'm looking at this in completely the wrong way.

Dave

User avatar
tricky
Posts: 1811
Joined: Tue Jun 21, 2011 8:25 am
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby tricky » Tue Jul 12, 2016 6:16 pm

You can usually use a 2D operation, or a 3D one where you just set the depth to .5 (DX:0.0 GL:-1.0 to 1.0 should be OK, but sometimes isn't).
The screen in projection space is -1.0,-1.0..1.0,1.0 where -1,-1 is either top left or bottom left (DX/GL) and depth if used is near..far DX:0.0..1.0 or GL -1.0..1.0
To scroll a screen, you could set the current "screen" as a texture/source (not sure what GL calls them) and your scrolled one as the render target/dest (again, not a GL person) and draw 2 quads (2 triangles each) copying from the moving the lower larger chunk up and the line from the top down. You can either pin-pong you have a second single copy to put things back where they started.
For displaying a beeb style memory layout screen as colours, you would need to run a pixel shader per destination DX:pixel GL:fragment which did some bit twiddling to work out which displayed pixel it was addressing and then use that to look up the palette entry from a DX:constant buffer GL:? to output as the colour. This isn't the most efficient was, but would require hardly more CPU work than your triangle and probably render plenty fast enough.
In the pixel shader, you can get the screen position in pixels or projection space depending on your API (GL(es?) for Pi) or you can add texture coordinates to you vertices and let the hardware interpolate them across the screen. (maybe have them in bytes where the fractional part is used to decode the pixel within a byte - up to you really).
I only ever used DX in games apart from the odd foray into consoles, never GL/GLes. I'm more of an optimisation/behind the scenes guy.

ThomasHarte
Posts: 327
Joined: Sat Dec 23, 2000 5:56 pm

Re: Daydreaming: a real VGA display card for the Atom

Postby ThomasHarte » Wed Aug 03, 2016 7:19 pm

tricky wrote:To scroll a screen, you could set the current "screen" as a texture/source (not sure what GL calls them) and your scrolled one as the render target/dest (again, not a GL person) and draw 2 quads (2 triangles each) copying from the moving the lower larger chunk up and the line from the top down. You can either pin-pong you have a second single copy to put things back where they started.

In GL, you'd probably avoid the default frame buffer — if your implementation even has one — and instead create a frame buffer that used a texture as backing for the colour buffer. glFramebufferTexture2D being the one word to search for. That gives you something you can both draw to and draw from, though never at the same time. Then, as Tricky says, ping pong (and loop in the default frame buffer, if your implementation has one).

tricky wrote:For displaying a beeb style memory layout screen as colours, you would need to run a pixel shader per destination DX:pixel GL:fragment which did some bit twiddling to work out which displayed pixel it was addressing and then use that to look up the palette entry from a DX:constant buffer GL:? to output as the colour. This isn't the most efficient was, but would require hardly more CPU work than your triangle and probably render plenty fast enough.

The Pi is an ES 2 device, and ES 2 permits that hardware need not support integral types but may emulate them with integers, hence declining to define any bitwise operations for its shading language. Obviously that doesn't necessarily mean that there are no bitwise operations at the hardware level, but the chip the Pi uses is sufficiently old (and a little odd) that it'd be a fair guess.

If you're planning a dependent texture access anyway, owing to the palette, you might be better just indexing by the whole 8-bit word to grab 2–8 pixels of output. Updating the table will be more laborious but it's probably a rare operation and you can cache recently-uploaded textures to dodge most of the cost.

tricky wrote:In the pixel shader, you can get the screen position in pixels or projection space depending on your API (GL(es?) for Pi) or you can add texture coordinates to you vertices and let the hardware interpolate them across the screen. (maybe have them in bytes where the fractional part is used to decode the pixel within a byte - up to you really).

Check out gl_FragCoord as the built-in fragment-shader input in GL (including ES 2).

tricky wrote:I only ever used DX in games apart from the odd foray into consoles, never GL/GLes. I'm more of an optimisation/behind the scenes guy.

For working at PowerVR you're my hero.

User avatar
tricky
Posts: 1811
Joined: Tue Jun 21, 2011 8:25 am
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby tricky » Thu Aug 04, 2016 6:37 pm

ThomasHarte wrote:For working at PowerVR you're my hero.

If I'd looked into it more 15 years ago (10 years earlier), I would have applied then, but thought they were part of Matrox and so, not in the UK.

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Sat Oct 08, 2016 9:57 pm

@Hoglet,

Can you give any details or instructions how to compile the C-programs? I'm ready to give this a go with a PI zero. My aim is just to use the graphics interface. In the first place as a display adapter for my blue Atom but maybe later also for my green one with the matchbox co-pro (in this case I have to address the Pi 0 on another address because it will conflict with the matchbox).

But for starting just a cheap graphic board for my Atom. I already have a Pi Zero and a level conversion board. Just waiting for some additional connectors that I ordered....
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Sun Oct 09, 2016 8:48 am

roland wrote:@Hoglet,
Can you give any details or instructions how to compile the C-programs? I'm ready to give this a go with a PI zero. My aim is just to use the graphics interface. In the first place as a display adapter for my blue Atom but maybe later also for my green one with the matchbox co-pro (in this case I have to address the Pi 0 on another address because it will conflict with the matchbox).

You will need to do this on Ubuntu (or at least some compatible linux).

The instructions are mostly the same as written here:
- Build dependencies
- Running CMake
- Compiling kernel.img

The only addition is you need to slightly amend the "git clone" command (in the Running CMake section) to checkout the atom_vdu branch:

Code: Select all

git clone https://github.com/hoglet67/PiTubeDirect.git -b atom_vdu

I'm happy to build you a binary, but I assume you want to be able to make your own development changes.

Dave
Last edited by hoglet on Sun Oct 09, 2016 9:07 am, edited 1 time in total.

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Sun Oct 09, 2016 9:02 am

I'll give it a try to install the development tools on my Linux Mint laptop.
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Sun Oct 09, 2016 6:15 pm

I installed the compiler like your instructions told me but I couldn't compile anything. I got this:

C compiler wrote:/home/roland/PiTubeDirect/src/armc-cstubs.c:55:19: fatal error: errno.h: No such file or directory
#include <errno.h>
^
compilation terminated.
make[2]: *** [CMakeFiles/tube-client.dir/armc-cstubs.c.obj] Error 1
make[1]: *** [CMakeFiles/tube-client.dir/all] Error 2
make: *** [all] Error 2


I did a little search and I found out that I had to install another package. I think you want to add this to your wiki:

Code: Select all

sudo apt-get install libnewlib-arm-none-eabi


Now it compiles without errors and warnings:

C compiler wrote:[ 2%] Building C object CMakeFiles/tube-client.dir/armc-cstartup.c.obj
[ 4%] Building C object CMakeFiles/tube-client.dir/armc-cstubs.c.obj
[ 7%] Building ASM object CMakeFiles/tube-client.dir/armc-start.S.obj
[ 9%] Building C object CMakeFiles/tube-client.dir/tube-client.c.obj
[ 12%] Building C object CMakeFiles/tube-client.dir/tube-exception.c.obj
[ 14%] Building C object CMakeFiles/tube-client.dir/cache.c.obj
[ 17%] Building C object CMakeFiles/tube-client.dir/rpi-gpio.c.obj
[ 19%] Building C object CMakeFiles/tube-client.dir/rpi-aux.c.obj
[ 21%] Building C object CMakeFiles/tube-client.dir/rpi-interrupts.c.obj
[ 24%] Building C object CMakeFiles/tube-client.dir/rpi-mailbox.c.obj
[ 26%] Building C object CMakeFiles/tube-client.dir/rpi-mailbox-interface.c.obj
[ 29%] Building ASM object CMakeFiles/tube-client.dir/tube.S.obj
[ 31%] Building C object CMakeFiles/tube-client.dir/tube-ula.c.obj
[ 34%] Building C object CMakeFiles/tube-client.dir/programs.c.obj
[ 36%] Building C object CMakeFiles/tube-client.dir/info.c.obj
[ 39%] Building C object CMakeFiles/tube-client.dir/framebuffer.c.obj
[ 41%] Building C object CMakeFiles/tube-client.dir/v3d.c.obj
[ 43%] Building C object CMakeFiles/tube-client.dir/performance.c.obj
[ 46%] Building C object CMakeFiles/tube-client.dir/copro-65tube.c.obj
[ 48%] Building ASM object CMakeFiles/tube-client.dir/copro-65tubeasm.S.obj
[ 51%] Building C object CMakeFiles/tube-client.dir/copro-lib6502.c.obj
[ 53%] Building C object CMakeFiles/tube-client.dir/lib6502.c.obj
[ 56%] Building C object CMakeFiles/tube-client.dir/tuberom_6502.c.obj
[ 58%] Building C object CMakeFiles/tube-client.dir/copro-80186.c.obj
[ 60%] Building C object CMakeFiles/tube-client.dir/cpu80186/cpu80186.c.obj
[ 63%] Building C object CMakeFiles/tube-client.dir/cpu80186/i386dasm.c.obj
[ 65%] Building C object CMakeFiles/tube-client.dir/cpu80186/iop80186.c.obj
[ 68%] Building C object CMakeFiles/tube-client.dir/cpu80186/mem80186.c.obj
[ 70%] Building C object CMakeFiles/tube-client.dir/copro-arm2.c.obj
[ 73%] Building C object CMakeFiles/tube-client.dir/tuberom_arm.c.obj
[ 75%] Building C object CMakeFiles/tube-client.dir/mame/arm.c.obj
[ 78%] Building C object CMakeFiles/tube-client.dir/copro-32016.c.obj
[ 80%] Building C object CMakeFiles/tube-client.dir/NS32016/32016.c.obj
[ 82%] Building C object CMakeFiles/tube-client.dir/NS32016/Decode.c.obj
[ 85%] Building C object CMakeFiles/tube-client.dir/NS32016/mem32016.c.obj
[ 87%] Building C object CMakeFiles/tube-client.dir/NS32016/NSDis.c.obj
[ 90%] Building C object CMakeFiles/tube-client.dir/NS32016/Profile.c.obj
[ 92%] Building C object CMakeFiles/tube-client.dir/NS32016/Trap.c.obj
[ 95%] Building C object CMakeFiles/tube-client.dir/copro-z80.c.obj
[ 97%] Building C object CMakeFiles/tube-client.dir/yaze/simz80.c.obj
[100%] Building C object CMakeFiles/tube-client.dir/copro-null.c.obj
Linking C executable tube-client
Convert the ELF output file to a binary image
[100%] Built target tube-client
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Sun Oct 09, 2016 6:20 pm

Hi Roland,

I've made that change - thanks.

Dave

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Sun Oct 09, 2016 6:27 pm

That's quicker than I can download Raspbian Jessie Lite :lol:

Damn, I have no connectors so I'm stuck at the moment. I think I'm going to annoy Mrs Roland :mrgreen:
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Sun Oct 16, 2016 5:10 pm

I have now a dual monitor setup on my Atom but the second adapter (Pi Zero) doesn't keep up with my 1 MHz Atom :shock:
Any idea how to troubleshoot this?

atom-dual-screen.jpg
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Sun Oct 16, 2016 6:41 pm

roland wrote:Any idea how to troubleshoot this?

Couple of questions....

What speed is your Atom running at?

What software patches have you made to OSWRCH to write to #BEE4?

I'll try and reproduce what you are doing, see if I get similar flakiness.

It's also worth saying this a currently a huge hack, and there is no actual flow control back to the Atom. :oops:

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Sun Oct 16, 2016 6:45 pm

My Atom is running at 1 MHz and I used two methods to display some characters:
1: poking some values to #BEE4
2: patch printer routine at #FEFB: STA #BEE4;RTS
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Sun Oct 16, 2016 6:51 pm

I must admit that I have a somewhat messy construction for connecting the Pi Zero to the Atom. I used an Atom Tube board and soldered the converter PCB that I got from Kjell into the 40p tube connector. But I had to enlarge the pcb holes (which destroyed all the plated-through holes) and I had to patch these with wires. So it is possible that there is an error in the wiring and therefore my question: how to troubleshoot this?

IMG_1502.JPG


BTW 1: I removed the internal Atom Tube interface so there is no address conflict at the moment.
BTW 2: The setting of the jumper doesn't make any difference
BTW 3: The settings in the config.txt file are adjusted like the description in the Wiki
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Sun Oct 16, 2016 6:59 pm

Are you on the new Atom or the old Atom?

I'll try and get mine set up now and we can compare.

Dave

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Sun Oct 16, 2016 7:01 pm

It's my new Atom. My old (blue) Atom didn't start any more :cry:
Could it be that the wires between the Atom and interface board are too long? Although they are buffered at both sides....
A shorter cable doesn't solve the issue.
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Sun Oct 16, 2016 7:25 pm

OK, I have a working set up in front of me now.

The Pi is a Pi Zero with the "normal" overclock settings: arm_freq=1000 and core_freq=400 (as per the wiki) running the latest kernel.img built from the atom_vdu branch.

The processor in the Atom is a 65C02P4 (and this does run at 4MHZ) currently running at 1MHz.

In the new Atom INIT program I changed:

Code: Select all

305 ?#7EFB=#60

to

Code: Select all

305 P=#7EFB;[ STA #BEE4;RTS;]

As soon as INIT reboots, I get screen echoing.

I can OLD and LIST, and this echo without any corruption.

VGA80 (0.27) doesn't currently work, because I think the JSR to #FEFB is commented out.

Dave

User avatar
roland
Posts: 2777
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: Daydreaming: a real VGA display card for the Atom

Postby roland » Sun Oct 16, 2016 7:45 pm

I'll have to recheck my setup. There's something strange going on with it. Take a look at https://youtu.be/mKlaXNUvyg0 The Pi stops printing characters if my hand is not over it. I do not need to touch it, just holding my hand above it makes it print something. But it hasn't to be my hand ... a sheet of paper also does make it print :-k

Edit: the voltage on the Atom tube PCB is only 4.55V. Maybe that is too low... I'll try to increase this some time this week.
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
hoglet
Posts: 6353
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Daydreaming: a real VGA display card for the Atom

Postby hoglet » Sun Oct 16, 2016 7:46 pm

A couple more things.

I'm not running the Atom Tube software currently (though that should be possible, and did work in the past)

If I patch FPGAUTILS and uncomment JSR #FEFB that also works fine, and both screens are updated:
https://github.com/hoglet67/FpgaUtils/b ... 0.inc#L569

Do you get anything echoed if just do:

Code: Select all

?#BEE4=65

What about:

Code: Select all

F.A=32 TO 126;?#BEE4=A;N.

Dave


Return to “acorn atom”

Who is online

Users browsing this forum: No registered users and 1 guest