Re: ToolKit 4.09 source code

From: Michael Grove <>
Date: Thu, 11 Mar 1999 08:37:55 -0800

Jo Even Skarstein wrote:
> On Mon, 8 Feb 1999 10:48:53 +0100 (MET), Seb wrote:

> I use 5.07, and occasionally boot into TOS. It doesn't happen very often
> though, perhaps once or twice a year...
> > And am I alone : when I make a floppy access without any
> > floppy disk in the drive, the computer makes at best three bombs and
> > at worse resets !!!

I don't remember this occurring with TOS, but under MiNT and
Thing, as someone stated before, Thing generally quits to
the NAES manager. No data loss or anything, just start Thing
A rule of thumb I have found on my A/B system is to always
have a disk inserted. I generally leave the A/B boot disk
inserted, but most anyone will do. I have always used TK

> This has never happened to me. I have never used anything else than 5.07,
> even with MagiC, and it's generally very stable.
> > And when I make a SCSI access in Nemesis Low Mode, the computer resets !!
> Nemesis and Afterburner are not the ideal partners that Titan claims ;-) After
> a *lot* of tinkering I've managed to get my AB/Nemesis combo stable in 40MHz,
> but any SCSI-access fails. Usually it just locks my Falcon, but quite often
> it destroys data as well. Because of this I've almost stopped running my
> Falcon in 40MHz, simply because I tend to forget it and try to access my
> SCSI-drive...

I have set my A/B up on two of the Falcons I have without
problems at 40Mhz. I can run at 42Mhz, but without SCSI.
Same with my Mighty Sonic. It craps out at 42Mhz, but runs
well at 40Mhz. It was much harder to setup the 030 board
than it was the A/B. I did all the shielding that was
discussed early on within this mailing list, and have about
7 ground points for the Nemesis and A/B. I installed a
shielded line for the DMA clock-in, all clock-in's are run
to the alternate (troublesome) points as listed in the
Nemesis installation manual. No FPU is installed, and the
FPU clock pin (socket) is isolated from the DMA line and
motherboard (the Mighty has the FPU installed, but the
clock-in is from it's own oscillator). The DMA resistor
value is what came with the Nemesis. The Mighty, funny
enough, uses the Nemesis buffer, but without the cap or
resistor for best performance. Both machines use the
Wizztronics ST RAM memory upgrade card. The A/B uses (it's
in a rack case) a PC powersupply and a DEKA keyboard

But...... This works best with the NOVA, whether it be TOS
or MiNT. Without the NOVA, the machines are still stable,
but SCSI timeouts are frequent in T/C modes. What I did to
help this out a bit was not to use the Nemesis for Videl
acceleration, but feed a higher speed clock into the video
port (external). I didn't try this on the A/B, but with the
Mighty, SCSI performance in T/C modes improved. I use an
EZ135 on the SCSI, with a back up, so if a problem does come's no problem. I think I used a 25Mhz clock into
the Videl with the Nemesis running at 20Mhz, controlled with
Videlity. Of course this only helps if your using a VGA
multisync monitor.....

The third Falcon I have is a REV K. It flat out refused to
work reliably with either accelerator installed. The other
machines are older main boards.

Received on to. mars 11 1999 - 17:37:55 CET

This archive was generated by hypermail 2.3.0 : ti. nov. 03 2015 - 20:07:54 CET