One of my "wonder if it actually works" projects recently made significant progress on the Classic Amiga 1000. I got the ACA500 working on the Amiga 1000!
I saw a YouTube video from someone who managed to get it to work. I was keen to try it out so I found an ACA500 for sale on Ebay and soon enough it arrived.
I was glad (and worried) about the case that covered the ACA500 - primarily because on an Amiga 1000 the expansion needs to be fitted the wrong way around, which means the case would get in the way of the mouse and joystick ports, which are on the same side as the expansion port on the Amiga 1000...
You can see the two CF slots with some rather bent pins...I did fix them but they are very fragile, as I have seen reported elsewhere on the internet forums before.
Lucky for me the CF cards are formatted and already have Kickstart 1.3 and 3.1 installed from the previous owner (who had an Amiga 500), with a base Workbench 3.1 HD setup too.
This means that the Amiga 1000 with the ACA500 should have an automatic Kickstart 1.3 or 3.1 (no longer need a kickstart disk), and able to run Workbench 3.1 on a Hard disk! Very cool!
Previously I have been using a Gotek with the Kickstart 1.3 ADF installed on it in position 001 to boot the Amiga 1000 without the need to use a physical floppy disk:
Up front I will say that the ACA500 doesn't work very often on the Amiga 1000...
You have to power off/on the Amiga 1000 many times for the ACA500 to be recognised, but when it does get recognised on power on, it works perfectly. For this reason I can't recommend trying it yourself because it is a bit flakey. Your mileage may vary of course!
I connected the CF cards to the appropriate slots on the ACA 500:
Next I connected it to the A1000 minus any memory expansions to make sure it worked by itself
And yes, I realised later that I had the mouse plugged into the wrong port (smacks head). On the other hand it was useful to determine the issues with the ACA500 case when connected to the Amiga 1000:
As it turned out I could lift the plastic on the case just enough to run the mouse cable through it:
Not sure if that would work with an accelerator installed, but for now...
I powered on the Amiga 1000 and the ACA500 worked and presented the initial boot menu!
I then explored the options menu on the ACA500:
expert mode you can change a lot of settings, PAL/NTSC mode, mapping
the ROM (essential in this case on the Amiga 1000), and option L to load
an Amiga 1000 kickstart disk. Clearly this device has been setup with
Amiga 1000's in mind...
I should also mention that the ACA500 has 1MB of fast memory include on it, which allows for mapping kickstart images.
I then kicked off the boot process into Workbench 3.1 Hard disk from the ACA500:
The CF cards light up furiously while booting off the hard disk on the CF card:
And there it is, Workbench 3.1 on my Amiga 1000 - who would have thought I would ever be running Workbench 3.1 on it!
That done, I then added the Comspec memory expansion (covered previously in my blog here), and this extra bulk had the additional benefit of the ACA500 case no longer getting in the way of the joystick and mouse ports.
Quite busy around the Amiga 1000 now, with the Comspec memory card, Gotek, VGA scandoubler, AV out and ACA500 all connected to it:
For those curious I use the Gotek floppy connector I normally use on the CDTV floppy port as it includes the power needed to boot the Gotek. But because the Gotek needs to be on DF0 on the Amiga 1000 to boot the kickstart disk, the floppy cable runs from inside the Amiga 1000 to the gotek separately (shown well in the shot above). I covered more on the Gotek on my Amiga 1000 in my earlier blog post here.
In any case, the great news is that after connecting everything the ACA500 worked again on power on:
I then selected to boot into Workbench 3.1 from hard disk:
Workbench 3.1 on my Amiga 1000 booted quickly from CF hard disk on the ACA500. I also now get the additional memory from the Comspec memory expansion too. Fantastic:
I ran DOpus 4 on it since it was on the hard disk - just playing around really:
I then added another old Amiga 1010 external floppy drive (and disconnected the Gotek) to be able to copy stuff from floppy disk for installations, and to check the ACA500 worked to boot kickstart via mapped rom without the Gotek present, which it did!
Copying to the hard disk from floppy was nice and fast. The ACA500 means I don't need the Gotek to boot kickstart anymore!
As usual though, I didn't leave it there...
I wanted to be adventurous and see if I could add an Amiga 1200 accelerator to it! Specifically, a spare Apollo 060-66 accelerator I have - I no longer have an Amiga 1200 as I gave it to a friend, but I still have this accelerator for some reason! time to put it to use again!
So, I pulled apart the ACA500 case so I could put the accelerator in:
And here she is, the Apollo 060 ready to go in:
All connected up to the ACA500:
Unfortunately though the case wouldn't fit with the Apollo installed in the ACA500 accelerator connector because of the bulky extra board on top of it. It is clearly designed for the thin Individual Computers ACA1XXX accelerator boards...no surprise really.
So I installed the ACA500 with 060 accelerator into the Amiga 1000 expansion slot pass through minus the case, utilising a highly professional disk box to hold it up to the correct height to line up with the expansion connector:
Hmmm, gotta love professional installations - don't breathe on it:
I had already installed the Apollo 060 libraries from the setup disk before I powered down the Amiga 1000 for this part of the installation. they are needed to boot the Apollo in Workbench 3.1.
I thought how cool would it be to have an 060 Amiga 1000 with hard disk and Workbench 3.1? I even started thinking about running AmigaOS 3.9 on it if I could find some more memory! Hmmm, should stop thinking as it keeps costing me money :-)
Anyway, I powered on a few times and eventually the ACA500 came to life - you can see I have both the AV out and VGA functioning (to different screens):
Just to prove it does pick it up, here is the system information screen on the ACA500 showing the Apollo 060 accelerator detected.
I read somewhere that the ram on the Apollo 060 is not detected by the ACA500, but the board is detected fine.
So it did what they said! In any case the system refused to boot with the Apollo 060 card installed. It was a shame!
So, I took the accelerator out of the ACA500 and got to work making the stock Workbench 3.1 standard vanilla hard disk install a bit prettier:
MagicWB always does wonders for any Workbench 3.1 install, as does using DOpus 5.82 Magellan in Workbench replacement mode (as I had both on floppy disks ready to install easily):
Being forever ambitious I also tried (and failed) to get the USB working on the Amiga 1000 via the clockport on the ACA500 - I loaded in the Poseidon drivers:
I then attached the working subway USB card from my Amiga 600 to the ACA500 clockport, taking care as usual about the cable orientation.
But, it didn't work.
Not sure what the problem was exactly - Poseidon couldn't find the Subway device. Perhaps I need a different Subway USB driver - there is a separate one for the A604n card on my Amiga 600 when I set it up last year, so maybe there is another one for this? I didn't look into too much as I had run out of time to muck around with it for that day.
In any case I was very satisfied that I got the ACA500 working on the Amiga 1000, and in so doing got myself a booting Workbench 3.1 system and no more need for kickstart disks!
Maybe another day I'll give the USB and accelerator another go!
Saturday, 6 February 2016
Tuesday, 2 February 2016
Last time I tried it my Classic Amiga environment was running upside down on the A1222! (click to expand any images)
In addition to this I have been troubleshooting what I thought was an unrelated issue with some games not running right and found out the issue was the graphics card I was using on the A1222. It turned out they were related!
The card I am using on the A1222 is a Sapphire Radeon 7750 1Gb PCI-e card, which I originally used on the Sam460CR with the Radeon HD Southern Island drivers under AmigaOS 4.1 FE.
However, the Linux builds don't support that card for 3D acceleration. This is the same on the AmigaOne X1000 under Linux too, but I just didn't think about it being the cause of my upside down screen! Apparently this support is an issue for all PowerPC Linux builds. Thanks to Christian for helping me out with this.
It came to a obvious conclusion for me when I ran the non-SDL version of E-UAE JIT on the 7750 card and it worked fine, screen right way up.
With the example Slam Tilt on offer (as part of the image supplied), it was difficult to resist the urge to have a go before proceeding further!
Ok, well, I couldn't resist:
After a few games of Slam Tilt, I then turned my attention to setup my Amiga 4000D hard disk emulation environment on the A1222 (using non-SDL for now), and was very glad to see it ran well on the 7750 card:
The uaegfx drivers worked out of the box (I already had Picasso96 installed on my Amiga4000D), so I could quickly select my initial preferred 1024x768 screenmode in 24bit colour in Workbench Screenmode Prefs.
This is Amiga Writer running under AmigaOS 3.9 emulation on the A1222:
I was also pleased to see that internet access worked perfectly in the emulation as well. Here I am at Amigaworld.net while listening to some demo tunes from Delitracker II:
Next up you can see AGA demos and games ran well - Worms Directors Cut:
I also ran TBL's Silkcut demo which ran at normal speed thanks to the JIT support in E-UAE. It is a great thing that it is possible to run even the demanding Classic Amiga demos at proper speed on the A1222:
Anyway, as you see it is easy to become distracted with Amiga stuff, even when betatesting! :-)
With the likely SDL problem now identified, I grabbed a Radeon HD 5450 1GB PCI-e card from my X1000 (no longer used since upgrading to a higher card) to swap into the A1222. Here is the 5450 card I planned to use:
After installing it into the A1222 (in place of the 7750 card) and powering on, the upside down problem is now all fixed - the SDL version of E-UAE now also runs as expected, the right way up! YES!
With that sorted out, I went to check that my other issue around some games not working well under Debian was fixed, and I am glad to say they run properly now that 3D acceleration support is active:
Next, I set about setting a nicer resolution for my Amiga 4000 emulation on the A1222. This is easy enough to change in the E-UAE .uaerc file, modifying the gfx_width_windowed, gfx_height_windowed, gfx_width and gfx_height parameters.
With plenty of real estate on the Workbench, it is easy to stretch out:
Having got that working well, I turned my attention to my Amiga 2000 system emulation, which is of course a 68000 system with ECS chipset. Accordingly I needed to create a new configuration file for it for E-UAE:
This needs to be a standard non-RTG setup, 68000, no JIT, ECS, Kickstart 1.3 ROM, no Zorro3 memory as it is not supported on Amiga 2000, etc. On initial power on I get the familiar boot screen:
After I added the hard disk to the configuration, it then booted into my Amiga 2000 Workbench 1.3 system:
I then immediately went to run some demos:
Red Sector Cebit 90 demo - yes indeedy:
I then played some modules using Hippoplayer under Workbench 1.3:
I still have a long way to go with my Classic Amiga setups on the AmigaOne A1222, but as you can see I have already made a good start!