[rescue] Netra 1100/1400 => 1105/1405?

[James Lockwood]

On 22 Aug 2001, Gregory Leblanc wrote:

> Ick.  What happened to computers getting DC current from the power
> supply?

The voltages being supplied are DC, but the regulation on some of them is
externally adjustable.  I suspect if the details of how the core voltage
is set were available it would not be hard to piggyback a voltage
regulator onto an ATX supply (though something monolithic like an LM317
wouldn't cut it, you need more amperage).

The feedback lines are labeled "sense +/-", which leads me to believe that
they differentially signal the delta between the current voltage and the
"ideal" voltage and should read 0V across them when it's spot on (at which
point the CPU's get powered).  If this is the case then it should be easy
enough to rig something, though I wouldn't bet a U60 motherboard on it!

> And while we're on the topic, does the SGI Indy do something like this?
> I'd much rather get some nice switching power supplies and hack new
> connectors onto them than have to take apart and repair power supplies
> from these beasts.

I think the Indy just outputs +5/+12/-5/-12 but am not sure.  There must
be pinouts out there somewhere.

> Any idea what a typical draw on one of these machines is likely to be?
> I'd make this a less general question, but I don't know enough about
> these machines to know what "reasonable" (not maxed, not bare bones
> either) would be.

I'd guesstimate "typical" 48VDC consumption at around 4A.  You want a
comfortable safety margin so I'd use a supply capable of at least 8A. The
nice thing about supplying DC is that you don't have any power factor to
worry about, and power backup is a breeze.

As a general rule, most telco equipment is pretty tolerant of
over/undervoltage.  I wouldn't push it too far but some telco gear can
actually operate with anything from 24V on up.

-James

--
James Lockwood
Guy on Summer Vacation (dot-com bombed)
http://www.foonly.com/resume/