Aged Code Studio

Is there an AQUABYSS player with an A3000T (tower, not desktop) out there?
One of our fellow captains experiences very strange issues with the game on his A3000T. There are vertical bar glitches on the screen, just see it for yourselves:



As you can imagine, the game is unstable and crashes. He also tried running it without startup-sequence but the result was the same.
The game - from the very same archive - works perfectly fine on his Amiga 600.

If you have an A3000T and can test the game on it then please do so and let us know. Thanks in advance.

More detailed information from the machine is helpful. Which OS version, CPU, MMU Tools used, GFX card, Picasso version etc.

Here you go:

- A3640 CPU
- 16 MB ZIP Ram
- 2 MB chip RAM
- Big Ram Plus 256 MB
- EGS Spectrum 2MB
- Ariadne II
- Freeway Pro USB Card
- ROM 3.2.1
- AOS 3.2.2.1
- Zulu SCSI as replacement of Hard Disk with a 32GB SD.

- SCSI chip 08
- Super Buster 11
- DMAC 02
- Ramsey 04

Seems to be standard config except the special Spectrum GFX card. Which P96 Version is used? I know that in V3.3.2 is a fix for that card. -> http://wiki.icomp.de/wiki/P96

And i think that the player „Chritoph“ plays the game on an A3000T but with a ZZ9000 without probs. But iam not shure.

Hi all,

Yes, my current favorite Amiga is called "Sandy" and has the following specs:

- Amiga 3000T, DMAC2, Ramsey4, Buster 11, Agnus 3872B, 16MB Zip FPM 60ns for Fast, 2MB Zip 80ns for Chip, Amber as usual
- Cybervision 64, looping in Amber output externally, P96 3.3
- Prelude+ mpegit, looping in Paula output externally
- Triton (full) USB, only activating Prelude/Paula through ROM, Poseidon never active during bootup, starting on real usage only
- Ariadne2, Roadshow TCP/IP Stack
- CPLDICY, steering the CPU cooler
- Cyberstorm MK2 040/40MHz, 128MB Edo 60ns RAM
- Zulu-SCSI RP2040 internally, 128GB SD card, 32GB image used on it, 3.7MB/s in SysInfo with FFS
- Kick ROM 3.2.2.1
- OS 3.2.2.1
- Plextor PX20-TS CD-ROM
- 1 HD floppy, 1 DD Floppy
- everything, incl PSU, drives, ... recapped

100% stable, but that took me 1 year to achieve.

The A3000Ts busses are a mess, there is less radio stations in the world than overshootings and disturbances each on the A3000Ts mainboard.

Want an example?
- BFG stops Zorro from working completely
- Picasso4 not working
- Matze FliFi showing broken pixels like your screenshots, if not completely losing sync
- Triton Lite not working
- WarpEngine unstable, crashing and locking up
I got no oscilloscope, but I am sure the busses are a mess.

Rule that out, first. Remove all extension cards, try again.
Works? Add 1 card, try again.
Stops working? Dump the card.
Next.

3640 is said to be working in 3000T only with DMAC4/Ramsey 7.
Got those? Lucky you!
Not? Reconsider, please. Follow steps above.

Other than that... Is that the Amber output/onboard VGA?
Check Amber and Line Buffers, clean or fix socket.

Is that RGB output? Check Chip RAM, Denise, Agnus.

Is that some other flicker fixer? Go Amber.

Good luck and may the God of your choice be with you. You will need it!

As I can not edit anymore...

Recapped? Not? Do it!

Battery area - leaked? You sure it's okay? All lines - especially address lines - okay? You sure? Under the FPU? Photo or it didn't happen!

Oh, and while recapping... Remember, one of the capacitors close to the PSU connector was installed wrong way around. AFAIR, it decouples -12V vs GND, but Commo connected -12V to it's + pin....
But in that scenario, + goes to ground. Schematics, PCB silklayer and actually all boards I heard of/saw have it wrong, like the one on http://amiga.resource.cx .

You can find Chris Edwards blowing it up after recap on YouTube.

This is how it should be, believe me or not. Please measure yourself. No, this is not causing the problem above, but do not blow up one of these magnificent machines while recapping - and dry caps may be part of your problem.

There is no Amiga more steampunk than the 3000T, I still search for the lid to open to shovel the char coal in.

Hahahahaha, that is a very comprehensive reply Chritoph, thank you
I knew about a few issues with the A3000T but it turned out to be just the tip of the iceberg
If Violo hasn't read it, I'll let him know in our private correspondence.

Yeah,
I guess the 3000T is rare enough so I am sure to only know the tip of the iceberg either.

Although, the nightmarish position of the clock battery harmed most boards quite heavily. I wonder if a board in mint condition would do better - but finding one is mostly impossible.

If I ever get hold of a multi channel oscilloscope I will do some measurements and publish them. I guess these would be the worst of all Amiga signals. In fact, I regularly wonder how Commodore ever got that thing to work at all...

This happens to me sometimes.
My configuration:
AA3000+, CV64, MC68030+FPU, 2+16MB Ram, DSP, OS3.2, game is offline (copy from oryginał A3000, there is no problem).