| Using An A500 V2+ In An A600? | |
|---|
|
|---|
| | 
Mo Retro
Posts 241
13 Apr 2018 19:04
| Way back in december 2015, I raised the question over at Amibay if it was possible to use the A600 V2 in an A500.
Majsta answered that it wasn't possible due to reasons of a technical nature .
I never got to know what these reasons are.
http://www.amibay.com/showthread.php?71144-Vampire-600-V2-Amiga-600-FPGA-accelerator-Pre-order&p=741611&viewfull=1#post741611I would like add a new question on that subject: Is it possible to use an A500V2+ in an A600 by means of a socket adapter? I hear some say why would you want to do that?
1) Just for fun, to it's possible.
2) The A500 V2+ has 2 plus points compared to the A600 V2: actually the expansion header and the fast IDE. :) Can someone elaborate on that please? :)
| |
| | 
Leigh Russ
Posts 151
13 Apr 2018 20:15
| Theoretically is should work, however the adaptor would need to be more than just linking it up pin for pin, it would also need the required circuitry to disable the on board 68000 CPU. I doubt anyone will ever design such a thing, though it would be nice. Something similar would be needed to use the V4 on an A600, I believe when the V4 was announced there was rumours of an adaptor coming from Kipper2K but now hes not dealing with Vampire stuff I doubt this will ever come to see the light of day
| |
| |
Mo Retro
Posts 241
14 Apr 2018 11:01
| Yes indeed I remember now they were talking about an A600 adapter for the V4.
I hope that some one else take the challenge to make that adapter. So the A600 wouldn't be left out :)
| |
| | 
Igor Majstorovic
(Apollo Team Member)
Posts 407
14 Apr 2018 12:01
| From a technical point of view...
First V500 don't have bus arbitration logic signals(BR,BG and BGACK) so that card used on A600 could never disable old MC68K, it could never takeover the bus. In that situation Apollo core wouldn't start and everything could result in damaging original MC68K on A600 and probably level shifters on V500 card. Apollo core will start only in certain conditions regarding BG,DTACK and AS. Since MC68K on A600 don't get BR who don't exist in V500 it can't reply with signals like BG, then it can't finish cycle reporting it with DTACK and AS high so new CPU(Apollo) can take over. So because of lack of 2-wire or 3-wire bus arbitration this can never work. Also some other parts of Amiga chipset on A600 could die within seconds. Do you know that I destroyed 24 Amiga 600 before making working Vampire card.
There are also other things like E,VMA and VPA, signals who are integrated in V500 and not on V600 because on A600 those signals are generated by Gayle who don't exist in A500 as you know. In A600 replacement for VMA is also done by Gayle Even_CIA, Odd_CIA controlling their CS after address decoding. Also, partially emulated Gayle in V500 could conflict with that Gayle who exist on A600.
As much cards looks nearly the same they are not. If I say that I could do all of this with few additional wires and internal coding of Apollo-core it is possible but makes no sense since end user don't have access to Apollo core files to modify what's needed for this to work. I wouldn't recommend modifying anything because I m 100% certain that if will endup with dead Amiga and Vampire card. There is lot more here to explain why this won't work but I think that this is more than enough. And anyway A600 will have more Vampire models in future years to cover all things suggested by users so far. Making better card will be our ultimate goal.
| |
| | 
Damian Kowalski
Posts 2
14 Apr 2018 15:56
| Yes it's true. Build the adapter V600 to A500 is much easier.
| |
| |
Mo Retro
Posts 241
14 Apr 2018 18:19
| Thanks Majsta for taking the time to answer. This explains a lot. I hope only that the V4 adapter will be eventually made for the A600.
| |
| |
Igor Majstorovic
(Apollo Team Member)
Posts 407
14 Apr 2018 20:16
| @Damian Kowalski as I tried to explain you can't build adapter like that. It is possible maybe with some middle logic CPLD or something but again I wouldn't try that.
@Mo Retro making any accelerator FPGA based by using principles shown on Vampire V1 proves that this is trivial thing. Our only problem right now is that we are still too small for big things to happen. Eventually we will progress in future years. To tell you the truth I had very hard time to accept to manufacture 2 models same time.
| |
| |
Damian Kowalski
Posts 2
14 Apr 2018 23:11
| I know that this is possible because I built such an adapter from V600 to A500 and it works. of course, additional logic is needed for this. but it is not so difficult.
| |
|
Documentation about the Vampire hardware