[rescue] Re: Re: Re: OSM-2000/RSM214

[Jonathan C. Patschke]

On Thu, 18 Apr 2002, Robert Novak wrote:

> Thanks for the reminder and the input on this. First one is up and
> running. Hmmm, wonder if it's insane to run a diff-to-se converter on this
> (so it goes SE to HVD to SE). :-)

I couldn't get it to work.  There's enough delay to cause signal skew.  I
could get to the rightmost four devices in the shelf fairly reliably, but
accessing the leftmost three would -always- cause a bus reset.  It may
have been that I didn't have a swift-enough bus converter.

However, just for giggles, pull the interface card out of there.  Why,
look at that.  It's a Unisys SE<->DIFF bus converter.  Isn't that special.
I'd -love- to get the pinouts of that PGA connector on the backplane. That
way, I could run SE all the way to the port, making the device much more
useful to me.  The latency incurred by that converter, BTW, is part of the
reason for that big "do not use on an UltraFast & Wide bus" label on the
back.

Personally, I can't understand why Sun put that converter in there,
anyway.  The runs to the control module weren't that long, and I wouldn't
suspect that there'd be any big problems with interference.  The end
result is the shelf is less useful as a standalone component.

--Jonathan