| From: | richs |
| Date: | 16 Jun 2000 at 15:45:30 |
| Subject: | Re: Blizzard2060 and 3_5 (was Re: Oddities with MTec 020i in A2000 with OS3.1/3.5) |
Hello Kilian
You asked
> seems that my two hd's dont like each other. I have a IBM DCAS-32160 2.1GB
> and a Quantum trailblazer 820MB.
Hmm, I have Fujitsu 1.2G on 3.1 and Seagate Barracuda 9.2G on 3.5.
Soon I will remove the Fujitsu, so I can boot without holding down
mousebuttons to select which one I want.
Fuj is ID5, Barra is ID4.
Can you run both of your drives separately on their own?
Your total address space is inside the 4G limit, so you do not need
to think about any of the 3.5 fiddles. Did they work OK together
when you had 3.1?
Are they both 50pin connection? I would guess so being that small.
If one was a wide SCSI 68 then perhaps the upper address range was
confusing things.
Are they both set as single ended on the bus? Remember the 2060 is
not a differential bus.
Do you have autoselection on for one and off for the other? This
might be a strangeness, but I can't see why.
How many terminations are there on the SCSI bus? At the relatively
low transfer rates of the 2060 it is very tolerant of abuse, but
there comes a point where it gets unreliable.
There are no stubs on the SCSI bus are there? Same comment as the
above para.
How long is the SCSI bus? Mine is actually too long, 1.5M with the
two external boxes, but the cables are the best, the terminations
are correct and the longest stub is inside the CD/DAT box, about
8cms. Uh-oh the 9.2G is in the DAT slot at the moment, the 1.2G
has its own terminator, but the 9.2G does not. This will be a
problem when I do the final swap. :^(
> I tried everything from changing IDs, uncommenting 2060scsi.device from
> NSDPatch.cfg, setpatch noNSD waitforvalidat, but everything causes
> crashes when both drivers are on the chain.
Hmmm. I don't remember doing anything non standard there. I'll
have a look tonight and see how it is set. I am using the
2060scsi.device, but I do not remember it being patched, it can
already handle big drives, just the FFS needed the changes.
> But when disabling the caches in the bootmenu everything works fine. When
> only the IBM-drive is attached everything works fine, too.
Disabling caches means that you are effectively in 68000 mode.
This makes me think you are using the wrong 68040.library etc.
When I first got the 2060 I could not get the system to run in any
other way. But Ralph helped me out by pointing out the relevant
lines in the docs! Make sure you have Cyberpatcher and his friend
the Cybersnooper in your startup sequence, along with cpu (and NOT
the other tool CPU060). 3.5 trampled my S-S so I had to redo it
all based on the old one.
Please check you have all the correct phase5 SCSI softs in all the
usual places exactly as for 3.1.
> My dealer (where I bought the Bliz) has no idea either, and DCE-Computer
> did not answer my mails.
>
> ps: How did you install the HD-busy-LED? Each drive one LED, directly
> connected to the drive?
That would be nice ;^)
I used the header on the 2060 card. The lead to the A2000 front
panel HDrive LED needed to be extended by an extra 30cms or so.
This is now effectively a bus busy LED although AFAIR the signal is
more like Motor On. It lights when the scanner or the CDROM or the
DAT are accessed as well as the HDs.
Also the way the HD is mounted the LED on the end of the drive is
just visible close to a small hole in the box. ;^)
Kind regards
Richard Stevens
richs@rd.bbc.co.uk
These are my words and nothing to do with my employer.
------------------------------------------------------------------------
Make new friends, find the old at Classmates.com:
http://click.egroups.com/1/5530/3/_/644178/_/961166828/
------------------------------------------------------------------------