Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: Motormouth on March 29, 2017, 05:30:40 AM
-
How well does the Vampire 500 V2+ work with expansions?
on the A500\+ The 86 pin expansion
I guess the one I am thinking of is the Impact A500 HD8+?
Does it work with the GVP 286?
How about the A2000, does it work with zorro II SCSI cards?
How about bridgeboards?
-
I've been curious about this, myself. Asked a few times, most people say it's pretty compatible now, but that was before the cards started getting into the hands of many people. Be interesting to compile a list of tested cards! :hammer:
-
on the A500\+ The 86 pin expansion
I guess the one I am thinking of is the Impact A500 HD8+?
Does it work with the GVP 286?
How about the A2000, does it work with zorro II SCSI cards?
How about bridgeboards?
"could" cause problems..... because of the missing Autoconfig of the vampire..... I have a GVP 500 HD8 and get my vampire the next days.... I will report, if it works or not......
-
well, from what i can figure the vampire ram maps above the 24bit z2/ocs/ecs limit, so z2 boards should be fine? i've seen one guy bolt an ethernet card into his vamp'd up A2000 with no problems, and heard someone else say they had no issues with their A590+2Mb.
but once you have fast IDE, and 128meg, doesn't that make most reasons for having a hard card redundant?
just so long as it's device controller is not called "scsi.device" you should be golden.
i also have an HD8+ with 4meg, and a vamp 500 on order. so yeah, would be interesting.
also the 286 card could possibly be outgunned by PCtask purely in software? just thinking...
-
Have you seen this for the A600?
http://www.apollo-core.com/knowledge.php?b=5¬e=749&order=&x=0
-
"could" cause problems..... because of the missing Autoconfig of the vampire..... I have a GVP 500 HD8 and get my vampire the next days.... I will report, if it works or not......
One should be very carefully distinguish two things: First, the lack of proper announcement of the hardware available on the vampire itself ("no autoconfig"), and second, the abilitiy of the board to identify other autoconfig boards in the system.
The former is lack of maturity of the vampire system, though it does not prevent per se other autoconfig hardware in the chain from working.
The second would be a more serious defect of which I hope that it does not occur. This is because autoconfig itself is not controlled by the vampire, but by the buster chip on the mainboard.
-
This is because autoconfig itself is not controlled by the vampire, but by the buster chip on the mainboard.
A500, so there is no buster. CFGIN/CFGOUT don't seem to be handled by buster on the A2000 either, the slots are looped with an 74LS chip to allow empty slots to be skipped. If the vampire prevented access to 0xE80000 then you wouldn't be able to configure the card, it could also prevent access to zorro 2 ram or io space which would allow the card to be detected but not used.
-
One should be very carefully distinguish two things: First, the lack of proper announcement of the hardware available on the vampire itself ("no autoconfig"), and second, the abilitiy of the board to identify other autoconfig boards in the system.
The former is lack of maturity of the vampire system, though it does not prevent per se other autoconfig hardware in the chain from working.
The second would be a more serious defect of which I hope that it does not occur. This is because autoconfig itself is not controlled by the vampire, but by the buster chip on the mainboard.
The A500 doesn't have a Buster chip but was designed to support one Auto-config device on it's 68000 expansion bus. Only the Vampire designers know exactly how their custom FPGA logic works, but there is really no (good) reason why it should not support Auto-config.
One possible issue is the A2000/A500 expansion.library support for anything beyond ZorroII 24 bit addressing (e.g. ZorroIII 32 bit addressing).
-
One possible issue is the A2000/A500 expansion.library support for anything beyond ZorroII 24 bit addressing (e.g. ZorroIII 32 bit addressing).
That part of expansion is at least identical over all versions, i.e. decoding of the Z-III extended flags in the config defs. It would not make much of a difference anyhow as the A2000 does not have the ability to address anything beyond the 24-bit space on its Zorro.
-
All of you have been bringing up great points.
Some of the questions are can the vampire handle cards on the 24 bit address and DMA on a 86 pin or zorro II and the Buster in the A2000 for Zorro II.
This is at least we have data points from darksun9210 with the A590 working with the Vampire, and a “Ethernet” card working in an A2000 with the Vampire.
Further,
I really like the page cunnpole posted for the A600 vampire, the problem however for the vampire A500 is that most of these are clockport devices or port adapters not really directly relevant to the A500.
One of reasons I was asking about expansion was deciding whether to put this in one of my A500s or A2000s
An A2000 with a vampire and a 486 bridgeboard would be kinda awesome.
But a scsi card or 16 bit memory or even an Ethernet card is not quite like the issues a bridgeboard can have.
But I also like the idea of a Super A500. The vampire would probably make the GVP 286 graphics fast enough to be useful.
Eventhough a more modern IDE controller is very nice, having a SCSI around in an Amiga particularly the older school ECS and earlier is useful.
-
Oh one other thing, has anyone seen, used, or own the vampire a500 adapter for the A2000 cpu slot??
It is in this image
http://wiki.apollo-accelerators.com/lib/exe/detail.php/v500v2-a2000-01.jpg?id=v500_v2
-
Oh one other thing, has anyone seen, used, or own the vampire a500 adapter for the A2000 cpu slot??
It is in this image
http://wiki.apollo-accelerators.com/lib/exe/detail.php/v500v2-a2000-01.jpg?id=v500_v2
I have. Looks like this: https://vimeo.com/181966474
-
Hello, I read this and went checking my status:
and get my vampire the next days....
My status changed to "Allocated (handled by majsta)". What does this mean?
Should I do anything from my side?
-
Hello, I read this and went checking my status:
My status changed to "Allocated (handled by majsta)". What does this mean?
Should I do anything from my side?
It means a Vampire has been reserved for you. You should have had an email asking for payment so I would check your spam folder in case its in there.
-
It means a Vampire has been reserved for you. You should have had an email asking for payment so I would check your spam folder in case its in there.
Thanks will contact majista to see how to sort it out.
-
An A2000 with a vampire and a 486 bridgeboard would be kinda awesome.
But a scsi card or 16 bit memory or even an Ethernet card is not quite like the issues a bridgeboard can have.
An A500 with a 486 pc emulation in the vampire FPGA would be awesome too. Although you probably couldn't fit both at the same time.
-
An A500 with a 486 pc emulation in the vampire FPGA would be awesome too. Although you probably couldn't fit both at the same time.
I think they already added commands to inprove emulation speed, one emulator is in development I think
-
I think the only non-autoconfig part of the Vampire is the RAM addition. And the reason it's done that way is because the chipset functionality is being slowly but methodically absorbed into the SAGA core. Would anyone care for 128 Megs of Chip RAM?
Seriously, https://vimeo.com/206380147 shows the V500 working with a Picasso IV graphics card in an A2000. https://vimeo.com/206998554 shows a Vampire system running Chocolate Heretic using a Prisma Megamix Z2 sound card. Memory cards work with it too, as do SCSI controllers and other old Z2 expansions.
-
I think they already added commands to inprove emulation speed, one emulator is in development I think
I think they were talking about hardware emulator boards like a Bridgeboard.
-
I think they were talking about hardware emulator boards like a Bridgeboard.
why on earth should anyone reimplement 486er in a Vampire FPGA? That is even more senseless than reimplementing PPC in FPGA
-
I think the only non-autoconfig part of the Vampire is the RAM addition.
Actually autoconfig does works with Vampire fastmem if you update expansion lib on kick from 32bit Amiga. e.g. "expansion_40.2(A500-600-2000) -> expansion_40.2(A1200)".
-
Would anyone care for 128 Megs of Chip RAM?
Ideally, all the RAM should be ChipRAM.
"UMA" is the most current buzz TLA for this, iirc :laughing:
-
Actually autoconfig does works with Vampire fastmem if you update expansion lib on kick from 32bit Amiga. e.g. "expansion_40.2(A500-600-2000) -> expansion_40.2(A1200)".
That is *not* autoconfig RAM. It is motherboard RAM (on the A1200) with a custom detection logic build into this particular version of expansion, though it is not using the autoconfig mechanism. In particular, there is no configdev for this RAM, no vendor ID and no product ID. It is similar to the "ranger memory" in this respect which is neither detected by an autoconfig type mechanism, but rather a detection logic that is hard-coded into the ROM.
-
That part of expansion is at least identical over all versions, i.e. decoding of the Z-III extended flags in the config defs. It would not make much of a difference anyhow as the A2000 does not have the ability to address anything beyond the 24-bit space on its Zorro.
Well, since there are different model specific versions of expansion.library it certainly implies that the model specific memory maps are an issue. But even if this is not an issue for the latest versions of expansion.library it could be for older library versions.
... and even though the A500/A2000 motherboards are limited to the ZorroII 24 bit address space accelerators with 32 bit CPUs can access 32 bit local bus hardware which could be implemented in ZorroIII format.
-
I think they were talking about hardware emulator boards like a Bridgeboard.
Yeah, I was referring to an actual hardware bridgeboard like a A2386sx. I am lucky to have one with a 486slc and a couple of ISA cards. It would be interesting if the bridgeboards were compatible with the vampire. They, the bridgeboards, can be a bit finicky with some non-standard accelerators and zorro cards, it usuallly is a problem with overlap in the 24 bit address space or DMA issues, which should not be a problem with the vampire.
Maybe a couple months from now when I get my vampire 500 I can check it out and post it here.
-
why on earth should anyone reimplement 486er in a Vampire FPGA?
why not?
That is even more senseless than reimplementing PPC in FPGA
Emulating a 486dx100 in an fpga would be much more doable than a 300mhz ppc. So less senseless.