🚀 go-pugleaf

In article <248480004wnr@xabcs.demon.co.uk>, J A F Bradbury
<Julian@xabcs.demon.co.uk> writes

>Who sells:
>- "certified" g.quality internal F-S-2 cables, 6/7 head?   (& ~ price?)

Make your own ?


>- 50-pin ?Micro-D? Active-Terminators?                     (& ~ price?)

SMC. Price ?  If you need to ask ?  The cost is silly but they were the
only UK vendor I could find for micro D.


>I've had a few quirks with a PD drive + NCR-815 + W95  <sigh>
>- It first pauses on reads (not writes: slower? @125-175KB/s) after >50MB
>  has been TR since power-on to either SCSI or EIDE h/d, then within 30secs
>  W95 hangs solid. PD busy-light goes out. A 5-10sec pauses is ~ right for 
>  a SCSI bus reset.
>
>  I'm assuming the PD drive (like a Tosh 3501, ie, CD) should not have 
>  sync TR mode enabled in fudge-95? It's already doing snail TR mode.
>
>
>Other devices (5) are ok.
>- I've not got NT on to see if the logs would say if it's a SCSI problem.
>  Noticed that even with the PD present-or-not Linux still throws a tantrum
>  during bootup, and it's stored on an *EIDE* h/d. That might have been due
>  to Norton Utilities trying to "fix" its partitions however :(

NT can be quite smart about scsi problems. I've seen 3.51 get into
problems and reinitialise the bus then continue as though nothing had
gone wrong !

The one thing it won't do is send start unit commands so if you switch a
device off by accident, it won't get to spin up again...


>It does sound like a SCSI lead problem, rather than anything else?

Good place to start.  Might be firmware if you have a very bad day.  :)


>- The Bus is Active at one end, Passive at the other. The lead is 1.5-1.75m
>  long, so with 5x internal devices wiring (=<10cm?) that's getting quite
>  close to the ideal-conditions limit of F-S-2.
>
>No-ones got a SCSI bus analyser they don't want? ;)

No, but if you find one I might share it...  :)


Cheers, J/.
-- 
John Beardmore