ACA500Plus - How to disable Blizzard 1260?

Caution: Non registered users only see threads and messages in the currently selected language, which is determined by their browser. Please create an account and log in to see all content by default. This is a limitation of the forum software.

Also users that are not logged in can not create new threads. This is a, unfortunately needed, counter measure against spam. Please create an account and log in to start new threads.

Don't Panic. Please wash hands.
  • Hi all,

    I have connected my Blizzard 1260 to the ACA500Plus and it seems to be working great.

    I'm struggling to disable it however, in order to use the ACA500Plus 68000 with different profiles (when wishing to use KS1.3 and Workbench 1.3 for instance, or running a 3.1 profile at 48Mhz).

    The wiki describes this:

    • CPU switch: If you have an A1200 accelerator card that can switch off it's CPU (ACA1233n, B1230IV, B1240, B1260), you can use this menu item to downgrade the computer to the 68ec000 processor of the ACA500plus. If selected, the 68000-only options (such as Action Replay Freezer and CPU speed settings) can be changed. The actual CPU switch is only executed when launching the profile. The CPU setting becomes part of a saved profile.

    If I set the CPU switch to 68000 I assume that this is what will disable the Blizzard when I activate the configuration. The CPU Switch option shows two possible configurations: ACA1233n 68030, and 68000. I find this odd as I expected to see reference to the Blizzard there, not an ACA card...?

    However, when i save "68000" to the NVRAM and launch the configuration, the ACA500Plus commences the boot, the rainbow flash of the Blizzard 1260 still appears and the accelerator still functions like normal, it is not disabled. Indeed, if I let it boot straight through into workbench the 060 is still active.

    So how to disable the Blizzard so I can occasionally use the ACA500Plus' 68000? I do not want to use the Cloaking Device as I still want to be able to use the ACA500Plus ram and IDE whilst the Blizzard is disabled.

    Is my firmware out of date here?

    My ACA500Plus has ROM v0.129, HW ID $08, Flash rom ID $1C/$16

    BigDismo V1.9 oct 10 2018

  • Blizzard cards are not officially supported. However, the CPU switch should work with the B1230-IV and the B1260. If it doesn't, then the B1260 is maybe a hacked B1240 with a different ID, but a 68060 CPU?

  • It may just be the autoconfig ID that we need to add to the list of supported cards. Can you do a "showconfig debug" in a shell and post the output here?

  • Before we go back to "switching this off in software": There's a 2.2G "chip ram" node in your memory list that overlaps with your "real" fastmem and crosses the $0000.0000 mark (32-bit wrap-around). Something is *really* wrong here, and this computer will NOT be stable.

    Does the accelerator show the same if used in an A1200?

  • Another hint: If the F3 page (cloaking mode) shows " Supported A1200 accelerator...... Press X to enable 68000 mode..." - what happens if you press X?

    A common mistake of some users has been to make settings, and then press F1, which will launch the full system.

    Another short test: Press F10, set CPU to 68000, then press return, not any other key (not even the enter key on the key pad!).

  • I will try this all this evening, thanks again for the reply.

    one thing I think pertinent now to mention - this machine has an A314 in the trapdoor expansion slot and this is providing the additional chip memory (to give this A500 1MB chip), so I wonder if this is the cause of the overlap you have mentioned?

    I did not mention this earlier because I forgot! And also because the machine has been working great prior to adding in this B1260 with just the ACA500Plus and the A314 - no problems!

    I will try the cloaking mode and the F10 68000 test this evening!

  • one thing I think pertinent now to mention - this machine has an A314 in the trapdoor expansion slot and this is providing the additional chip memory (to give this A500 1MB chip), so I wonder if this is the cause of the overlap you have mentioned?

    Chip ram expansion is usually not a problem. I don't know this specific one, so you might want to take a picture or link to it on BBoAH. I doubt that it has anything to do with the weird 2gig entry in your freemem list. That seems to be some software gone wild.

  • The board is this one:

    It is an interface for a raspberry pi (like all new Amiga stuff it seems!) for adding a bit of extra ram, to use the ARM as a coprocessor, and a few other things. It is providing the extra 512K which the ACA500Plus is turning into chip ram.

    I will remove this board and run showconfig debug again to see if it is part of the issue you have noticed.

  • The whole approach looks similar to the PowerPC board (A DOS-PC emulator for the trapdoor slot) that was available from a Dutch company about 25 years ago. Since the 512k chunk of this trapdoor expansion is only moved to $0008.0000 in memory space (and nothing more!), the worst thing that can happen is that the A314 software doesn't find it. This does not seem to be the case. However, since we are looking for a piece of software that adds a 2G chip ram chunk which really isn't there, the A314 is still a suspect.

  • Just looked at the schematics of the A314 - there must be something wrong. All drivers "from Amiga to 3.3V logic" are uni-directionsl. This means that there is no way that the computer can read from the 512k memory - it can only write. For bi-directional communication, I can only see the RTC lines, which are extremely slow.

    The published schematics are either not up to date (looking at schematics from April 2020, picture of the prototype from April 2019), or I am completely misunderstanding the concept. In any case, with the available information on GitHub, I'm convinced that this thing only "sniffs" the bus (probably creating a copy of the contents in local S-Ram), but requries "real memory" to be present in the computer that you can use as chip ram.

    *EDIT*: Got it - I was confused by the part number "74CBT16244" and applied the uni directional property to the part that I know from the 74xx244 part. When looking for a reason why the author(s) is/are using the weird 4.3V supply voltage, I found that the CBT series is a bidirectional "analogue" driver. I can't really see the specified properties of the chip when operated at this voltage, but reading the data sheet between the lines, it's not a completely bad choice. However, the Altera chip (which is 3.3V-only) may still get 4V at it's inputs if I read the CBT16244 datasheet right, which may recude it's life time. The better choice would be a bidirectional 3.3V driver that is 5V-tolerant (such as 74LCX16245). Saves the weird 4.3V regulator and only requires one more IO on the logic chip for direction control.

  • A few things to report!

    I removed the A314 from the system. This 'fixed' the 2.2GB chip ram anomoly and the machine now correctly reports 0.5MB chip ram :)

    Also, the removal seems to have fixed the original problem of not being able to disable the B1260 - the CPU Switch option now works as advertised and I can boot the machine with the aca500plus's 68000 :)

    So the lesson for me is to report problems with a stock machine, not with other things attached!

    I will put the A314 back in and do further tests, but to report on your previous test suggestion - if I pressed F10, 68000 and then return the machine 'booted' to a purple screen and got no futher. In Clock mode it reported a Supported A1200 accelerator.. but then pressing X similarly got no further than a purple screen.

    Thanks for your time on this, Jens, it's appreciated.