First VAX-11/730 VMS Boot! (was: Re: VAX-11/730 %BOOT-F-Unexpected Machine Check)

Mouse mouse at Rodents-Montreal.ORG
Sun Jun 14 14:38:02 CDT 2015

> That might be a good approach.  The DELUA end of the cable has a Berg connec$

(Please try to avoid paragraph-length lines.)

I once had a machine in a case it wasn't designed for, leading me to
want to extend a DA-15 AUI connector.  I experimented with plain
untwisted, unshielded, un-ground-planed 15-pin ribbon cable between the
machine and a UTP transceiver.  I found that a foot or two of ribbon
cable made no visible difference.  As an experiment, I then tried about
thirty feet of ribbon cable and got ludicrously high error rates.

On calculating, I find that, at 10Mb, bits are about 65 feet long (at
that sort of frequency, signals in copper travel by skin effect at
roughly 2/3 c), and, since it's Manchester encoded, the underlying data
rate is twice that, meaning a transition potentially every 30ish feet.
A cable two to three feet long is short enough for everything to settle
in plenty of time even if the impedances are mismatched, but one thirty
feet long is about the size of the underlying bits and it's entirely
understandable that it would cause trouble.

So, if you can hang a transceiver off the back of the machine with
little-to-no drop cable length, you can probably get away with a few
feet of pretty much anything, but if you want to run any significant
length of drop cable, you may well need to treat it as a transmission
line and worry about characterstic impedance and shielding and such.
But it's worth experimenting with anyway; it's possible the cable you
have has the right impedance or some such!

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse at
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B

More information about the cctalk mailing list