64 mb on A4000 mobo

[Crumb]

if that was true, all would have done that...

It's true

He also reprogrammed the IDE GAL chap so it gives 5MB/s with a CSMKII. I know at least two people in Spain with that trick. It only involved reprogramming the GAL chip. It was done cleanly, without overclocking anything. He optimized the code that managed the transfers of the IDE port of the GAL found in old A4000 revisions. He was in contact with David Haynie to do this.

The modifications required for 64MB support were more complex and involved more than reprogramming a chip. You had to sent him the motherboard.

He also overclocked the AGA chips (One of my best friends also does this with his A2000 from time to time so he watches 100Hz/31Khz pseudoPAL screenmodes without scandoubler), the ZorroIII bus, his 3640 board...


Just because you haven't seen it, it doesn't mean it doesn't exist.

Just to clarify: AFAIK the IDE trick could only be done with old board revisions, and the 64MB trick involved more work than changing a single chip like the IDE speedup trick...

@FrankBrana

Gure gasiosie! ;-)  :pint:

[Crumb]

I have to see it to believe 64MB on 4000

I was going to sent him my A4000T motherboard but as he only did the trick with desktop ones and I already had a CSMKII fully populated with 128MB I thought... why would I want more ram? Some years later (now for example) I realized it could have been quite useful when you have 2 A4000 and just a CSPPC... :-/ But yes, I know people who has this modification. It involves reprogramming some chips (that maybe A4000T or latest A4000 board revisions don't have) and adding some wires (and maybe cutting others). It's definitely not easy and I seriously doubt you could reproduce the trick without having access to the original sources of the programmable chips.

and as for overclocked chipset... overclocking??? chipset???

Yeah, my friend has an overclocked ECS chipset that gives him 100Hz/31Khz pseudo PAL screens. :-) He also did a small circuit to reset his agnus chip when he resets his A2000 as CV3D sometimes isn't recognized by P96 (on A2000 with Blizzard 060) when you make a reset. That small circuit fixed that. I really wish he was in contact with "Bacterius" as he could made a few changes in my two A4000... standard motherboard IDE running at 5MB/s... not bad, uh? :-) The downside of the 64MB of RAM in the motherboard is that these are as slow as the 16MB have been always :-/

[Crumb]

@lopos

AFAIK the GAL source code used in various chips in the first A4000s is not property of Bacterius even if he rewrote some critical loops or modified some code to make it accept more memory. So he may get in trouble if he released it.

He made all the modification by himself (you had to sent him your motherboard) and only charged a little (24€ for the IDE hack and 60€ for the 64MB hack). He also made the modifications directly in the parties of the south of Spain. AFAIK he *never* passed the sources to anyone. If somebody has the GAL sources it will probably be Dave Haynie.

It's not impossible to optimize again the A4000 GAL so maybe we could ask Dave Haynie if he could publish the GAL source code of the chip that controls the IDE interface.

We may ask him also about the ram interface. Maybe Bacterius submited back his changes in the source code... but I don't think he would free the code as he's not the real owner. I have doubts about the code not being lost.

BTW last time I heard about Bacterius he was no longer interested in Amiga stuff and he was no longer a technical repair service. I had his phone number in a pile of papers at my parents' house, but I guess it ended up in the trashcan when I went to live alone 3 years ago.