Re: SCSI errors

From: Oliver Skelton <>
Date: Mon, 09 Aug 1999 15:54:16 +0900

Hi Michael,

Just to give you my experience in case it is of help.

My Falcon was fitted with Nemesis and AB040 by Titan Designs a few years
I was using Blow Up Hard 1 to expand the screen resolution.
I have been using scsi all along (mainly Jaz 1GB and DAT).
I run HDDRIVER 6.13 (some problems with removables on the AB040 with the
newer one I tried... I think it was 7.05).

Sometime ago I discovered scsi was unreliable... it turned out that in 2
colour mode I had no problems but in 256 and TC modes (especially as
I had expanded screen res. with Blow-Up) I was getting errors...
detected with tree check as you have seen.

I now have a Nova and have not seen any disk corruption for some
time.... basically Nova is offloading the bus.

There has been quite a lot of talk in the UK about the scsi DMA buffer
patch... which can need tweaking. It is also different with the Nemesis.
Some people also swear by various extra earths on Nemesis and using
shielded coaxial cable from Nemesis? to the DMA. I believe one company
decouple this shielding somehow when they fit Nemesis.
One thing about the DMA patch is that you are between a rock and a hard
place. Adjusted one way and you have trouble booting (black screen and
no boot... you just have to press reset again.. maybe again and again
etc.!). Adjusted the other way you get scsi errors.
On my machine it is quite common to get the black screen and failure to
boot... but usually the second time will bring it up ok.

Actually I planned to fit the shielded coax sometime.. but as everything
seems fine with my Nova I am in the 'don't mess with it if it works'

I think I saw a blue-black screen sometime when I first fitted the
Nova.. and I think it was caused by part of the Nova shorting to the
Falcon case or being shorted by the Falcon case. However, I am not at
all sure about this, it could have been bad connection of AB040 card or
Nova card. Anyway, no harm checking.

Received on ma. aug. 09 1999 - 10:53:56 CEST

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